<?xml version="1.0" encoding="utf-8"?><feed xmlns="http://www.w3.org/2005/Atom" xml:lang="en"><generator uri="https://jekyllrb.com/" version="4.4.1">Jekyll</generator><link href="https://andreacoravos.com/feed.xml" rel="self" type="application/atom+xml"/><link href="https://andreacoravos.com/" rel="alternate" type="text/html" hreflang="en"/><updated>2026-11-16T20:27:06+00:00</updated><id>https://andreacoravos.com/feed.xml</id><title type="html">blank</title><subtitle>Andy Coravos researches, builds, and backs things at the boundary of science and the humanities. Venture Partner at Pillar. Co-founder of HumanFirst (acquired by ICON). </subtitle><entry><title type="html">The Human as Conductor</title><link href="https://andreacoravos.com/writing/2026/the-human-as-conductor/" rel="alternate" type="text/html" title="The Human as Conductor"/><published>2026-03-26T15:30:18+00:00</published><updated>2026-03-26T15:30:18+00:00</updated><id>https://andreacoravos.com/writing/2026/the-human-as-conductor</id><content type="html" xml:base="https://andreacoravos.com/writing/2026/the-human-as-conductor/"><![CDATA[<h4>My health insurance changed three times in one year. Drowning in life admin, I started tinkering with AI tools. One experiment led to the next. Would-be chores became… fun.</h4> <figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9-TupC-X_QwGNh7Gsj93Lg.jpeg"/><figcaption>Photo by <a href="https://unsplash.com/@gwundrig?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Manuel Nägeli</a> on <a href="https://unsplash.com/photos/music-group-performing-on-stage-with-empty-audience-seats-NsgsQjHA1mM?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure> <p>After selling our health tech company, HumanFirst, my COBRA insurance changed three times in one year as I moved back to the US from Europe. Normally, this would result in me spending a lot of time on the phone with my insurer, staring at EOBs, and trying to figure out what I owed and what they owed me. Instead, I started setting up new AI-driven workflows to handle every day tasks.</p> <p>Although ChatGPT by OpenAI might have more brand recognition for the public, <a href="https://claude.ai/new">Claude</a> (in its various forms) by Anthropic has been the clear winner for me. I have a new sense that I am a ‘conductor’ in my life rather than the do-er. My version of Claude and I co-developed this analogy:</p> <blockquote>A conductor doesn’t play every instrument.<em> They hold the score, understand how the parts fit together, and create the conditions for each player to do their best work. The trumpet player is still the best trumpet player in the world. The conductor just knows when to call on them, and makes sure they have what they need.</em></blockquote> <blockquote><em>That’s the shift I’ve been making with my own healthcare. I’m not replacing my doctors. Rather, I’m arriving with context already assembled. My labs are consolidated and reviewed, historical patterns flagged, and questions drafted for when we meet. My clinician can’t always do their best work when half of the relevant information lives in a portal they can’t access, a conversation with a specialist they’ve never met, or a family history I couldn’t remember off the top of my head. I couldn’t access my data easily either. Electronic Health Record (EHR) systems were built for providers and billing, and clinical data is an afterthought. Everyone has been working from an incomplete picture.</em></blockquote> <blockquote><em>AI changes that for me, but not by directly putting “in charge” either the patient or the doctor. Using AI doesn’t cause a power flip — at least, not yet. Rather, it enables </em>a third way<em>: patient and doctor seeing each other more clearly, because the context that used to get lost in the gaps can now be surfaced. I show up with my picture assembled. My doctor confirms, challenges, builds on it. We both do better work than either of us could alone.</em></blockquote> <blockquote><em>That’s what these experiments taught me.</em></blockquote> <p>With AI, I feel like I have a jetpack… and it’s easier to set up than you might think. If you’re feeling behind or too late, this graphic by <a href="https://x.com/damianplayer/status/2025234388137468387">@damianplayer</a> is for you:</p> <figure><img alt="" src="https://cdn-images-1.medium.com/max/970/1*a_ll5kqIDl6ld-4MHCzKKg.png"/><figcaption>Image by @DamianPlayer: <a href="https://x.com/damianplayer/status/2025234388137468387">https://x.com/damianplayer/status/2025234388137468387</a></figcaption></figure> <p>The tech has changed dramatically in the past few months, and it means that everyone can be ‘technical’ now. Capabilities that used to be only available to engineers now are available through normal language. The main requirements to participate are curiosity and a chat interface.</p> <h3>The Experiments</h3> <p>Most of these took less than a day to set up and ran in the background while I was doing other things:</p> <ol><li>Find missing insurance reimbursements and get $$</li><li>Build a reusable family health history from a conversation with my mom</li><li>Make sense of labs from a bajillion different sources</li><li>Check my supplements for interactions when adding a prenatal</li><li>Improve my body composition: build muscle 💪</li><li>Attempting to consolidate my vaccine records</li></ol> <p>Feel free to jump around.</p> <p>A few things worth naming before you dive in:</p> <p><strong>On “jigs” versus “tools” (or how-to).</strong> Someone in my favorite AI community shared a useful analogy: in woodworking, a “jig” is a personal gadget you create to help you to do <em>your</em> work, but it is never intended for packaging and distribution to others. Right now, a lot of us have a growing toolchain of software jigs that probably couldn’t (&amp; shouldn’t) be shared with others since they’re so custom fit. With that caveat in mind, I’ll share with you some of my ‘jigs’. They aren’t out-of-the-box, product-level ready, but they should help you get started.</p> <p><strong>AI can only optimize the system it’s given.</strong> And there are a lot of broken incentive systems in healthcare. Companies often make more money when you’re sick than when you’re healthy, almost by definition. As one Goldman Sachs analyst memorably <a href="https://www.cnbc.com/2018/04/11/goldman-asks-is-curing-patients-a-sustainable-business-model.html">asked</a>: is curing people a “sustainable business model?” (No).</p> <p>The bigger fixes require policy (not prompts). While those progress slowly, I’d rather more people experiment than let others build the future for them.</p> <p><strong>On when to use AI.</strong> A lot of people ask “<em>can</em> I delegate this to AI?” Sure. AI is getting better at tasks all the time. But <em>should</em> you?</p> <p>If you’ve had the opportunity to work with a stellar executive assistant (EA), you know they <em>could</em> do a larger set of tasks than their agreed upon scope. The relationship takes time. Together, you figure out how to divide the work, what needs the exec’s discernment, what doesn’t.</p> <p>Before this year, I’ve never had this type of relationship with a non-human intelligence. There are a few quirks that I’m learning about when interacting with AIs, but it’s somewhat similar to how you get to know someone’s personality at work. Except this time, I can also more strongly influence (or program) that personality.</p> <p>My personal litmus test is ”do I have a <em>motivation gap </em>to work on a task, and how can AI break it down for me to help me get started?”</p> <p>Momentum begets momentum. Small wins become bigger wins. I’m an efficiency junkie — so it’s fun to check things off. It’s also surprising to me how I feel a strange sense of connection with a non-human intelligence, as we move from one successfully completed experiment to the next.</p> <p>The following examples are my stories, and not canonical how-tos. With time you’ll develop your own discernment for how you want to conduct your experiments. This is what today’s internet likes to call ‘<em>taste</em>’.</p> <h3>Experiment #1: The Flexpa Insurance Recovery</h3> <p><em>𝐓𝐋𝐃𝐑: My health insurer owed me hundreds of dollars and never alerted me. I used </em><a href="https://www.linkedin.com/company/flexpa/"><em>Flexpa</em></a><em> + </em><a href="https://www.linkedin.com/company/anthropicresearch/"><em>Anthropic</em></a><em> + </em><a href="https://www.linkedin.com/company/getduckbill/"><em>Duckbill</em></a><em> to find it and get checks reissued in ~24 hours. Without these tools, this life admin task would have taken weeks and a lot of Aetna hold-music.</em></p> <p>Last month I had a sinking feeling that I haven’t gotten all my reimbursement checks from Aetna over the past 12 months. I called them, sat on hold forever, and got them to void and reissue one check from a claim I knew about. But I had no way to know how many more were missing.</p> <p>I’m an angel investor in<a href="https://www.linkedin.com/company/flexpa/"> </a><a href="https://my.flexpa.com/">Flexpa</a>, which got me early access to their app. I downloaded my entire medical record (a huge JSON file that would have taken me forever to review), ran it through Claude, and found a bunch of claims flagged “owed to patient.” Hundreds of dollars I suspected were out there and couldn’t confirm until now. Plus some denied claims I’d always had questions about.</p> <p>Then I built a DIY medical billing advocate. A “medical defender” if you will:</p> <p><strong>Step 1: </strong><a href="https://my.flexpa.com/">Flexpa</a> to pull my records</p> <p><strong>Step 2: </strong><a href="https://claude.ai/">Claude</a> to analyze claims, flag what’s owed, and draft scripts/emails</p> <p><strong>Step 3: </strong><a href="https://getduckbill.com/">Duckbill</a> to call Aetna for me (h/t<a href="https://www.linkedin.com/in/libbylh/"> Libby Leffler</a>)</p> <p>After signing some PHI authorization forms, my Duckbill agent got on the phone (a few times) with Aetna, found checks sent to an old address, and had them voided and reissued. They also explained why certain claims were denied and how to fix them by submitting new info, plus which payments went to providers instead of me so I know who to follow up with.</p> <p>I did all of this in ~24 hours between other tasks. The AI-tasks flowed in the background, which was the first time I clearly felt the <em>human-as-conductor</em> experience. It’s cool that this AI workflow saved me days/weeks, frustration, and $$. (h/t<a href="https://www.linkedin.com/in/tim-schneider1290/"> Tim Schneider</a> from BHI who inspired me to write this experience up publicly.)</p> <p>I asked<a href="https://www.linkedin.com/in/andrewarruda/"> Flexpa’s CEO Andrew Arruda</a> if I could share the link, and apparently anyone can now pull their own data (yay):<a href="https://lnkd.in/edcwCc4Z"> </a><a href="http://my.flexpa.com">my.flexpa.com</a></p> <p>A few tips for exploring your insurance data:</p> <ul><li><strong>First, review denied claims.</strong> Check if the same provider got approved for the same procedure before. If so, ask your AI to draft an email: “You approved this in the past. What’s missing now?”</li><li><strong>Look for inconsistent reimbursements.</strong> My acupuncture visits were reimbursed at different rates for the same procedure. I had my AI prep scripts for the Duckbill agent to ask about it.</li><li><strong>Give yourself time to poke around.</strong> The first time I looked at my claims data, I didn’t know what I was looking for. Patterns surfaced anyway.</li><li><strong>Find a community.</strong> Pick 2–5 people who like tinkering and host a casual S<em>how and Tell</em>. It can be easier to grasp what’s possible when you see someone else do it.</li></ul> <p>Let’s zoom out for a moment to the real issue: AI won’t fix a system with dark patterns. Why is Aetna still mailing checks in 2025? Why are there no alerts when checks go uncashed? The better solution for me isn’t improved AI workflow for checks. It’s offering direct deposit for out-of-network reimbursements. I’m looking at you, Aetna.</p> <p>So your insurer knows they owe you money and they just… don’t tell you. The good news: your medical records know too, and now your AI tools can surface what insurers won’t.</p> <p>This first experiment unlocked a fire in me about the “personal health records 2.0” era. Let’s try a few more.</p> <p><em>If you’d like a sample “jig” (prompt) to work on this problem yourself, you can find that </em><a href="https://docs.google.com/document/d/19f7aJtssFLLfSIQ5BJsbCcRXJOy8os0b9V-L-fpTZyQ/edit?tab=t.0"><em>here</em></a><em>.</em></p> <h3>Experiment #2: Mom’s Oral History + Impact on Exercise Routines</h3> <p>I often dreaded going to a new doctor. Each clinic asks for different intake forms, and somehow I end up filling out the same 12-page packet twice (once online, and again on a clipboard in the waiting room). I’ve always wanted a “common app” for family health history.</p> <p>Turns out LLMs are good at restructuring data if you have the ground truth. So, I scheduled a one-hour Zoom interview with my mom about our family health history, and ran the transcript through Claude to extract conditions, ages of onset, which relatives, and patterns to watch for. Now I have a document I can remix for any new doctor, and I also shared it with my sisters and extended family.</p> <p>Even better: I use this document as context in my <a href="https://www.sarahtavel.com/p/getting-started-with-claude-projects">Claude project</a>. A Claude “project” lets me add my own files and data, and the model references them every time I ask a question.</p> <p>Two examples of helpful content from my mom: My grandmother died two months short of her 85th birthday from a heart attack. Her identical twin died at 60, also from a heart attack. As family lore goes, her twin refused to take her blood pressure medication and didn’t exercise. Both twins had the same genetics, and my grandmother got to live 25 years longer.</p> <p>A second tragic storyline in my family was how Alzheimer’s ravaged my grandfather, one of the first people to work on fiber optics out of MIT Lincoln Lab.</p> <p>Now when I ask Claude to help with my exercise routine, it reminds me:</p> <blockquote>“Ruth vs. Lorraine : Both had the same genetics, 25-year lifespan difference. Ruth exercised and took her meds. Be Ruth.”</blockquote> <blockquote>Or: “Brain protection: builds mitochondria, clears amyloid, protects against the family Alzheimer’s pattern.”</blockquote> <p>These nudges might not be your cup of tea (you can design your own!). For me they’ve driven more consistency in my workouts than anything since the subtle social pressure of my college sports team. And it feels great.</p> <p><em>If you want to try this for yourself, here’s </em><a href="https://docs.google.com/document/d/19f7aJtssFLLfSIQ5BJsbCcRXJOy8os0b9V-L-fpTZyQ/edit?tab=t.0"><em>a “jig” to build your own family health history.</em></a></p> <h3>Experiment #3: Labs Consolidation</h3> <p>Over the past few years, I’ve gotten labs from a bajillion channels: Labcorp &amp; Quest Labs (ordered by various clinicians), <a href="https://www.parsleyhealth.com/">Parsley Health</a>, <a href="https://www.functionhealth.com/">Function Health</a>, <a href="https://framer.levels.com/">Levels Health</a>, a full body MRI from <a href="http://ezra.com/">Ezra</a>, a Coronary Artery Calcium (CAC) scan from <a href="https://www.lumiheart.com/">LumiHeart</a>, genetic tests from <a href="https://www.natera.com/">Natera</a> and <a href="https://www.orchidhealth.com/">Orchid Health</a>, DEXA scan for bone density and body composition.</p> <p>(Why a full body scan? A friend had headaches, and her doctors couldn’t figure it out. She paid out-of-pocket for <a href="https://prenuvo.com/">Prenovo</a> and found a brain tumor. She’s got treatment and is now in remission.)</p> <p>My lab results sit in various portals, different formats, unclear reference ranges. So I downloaded summaries from each, uploaded them to a Claude project, and asked for my trends over time, out-of-range metrics flagged, and anything to discuss with my doctor at OneMedical.</p> <p>Now I ask my Claude project for the “one thing” I can do each week. I’ve been daisy-chaining minor changes like nutrition tweaks and sleep adjustments. Those small wins unlocked bigger ones. I even found the motivation to tackle in physical therapy a running-related IT-band issue I’ve been ignoring for 20 years.</p> <p>What feels different this time around is that I programmed my Claude project to respond to how my brain works. I like structure, and also novelty. The recommendations match my patterns. I’m hitting PRs in the gym, my sleep and HRV have increased, and I’m spending <em>less</em> total time than before on these tasks. I feel like this experience is a remedy to the attention economy: my progress is no longer measured by “time spent”.</p> <p><em>A note before the jig: AI can hallucinate, and none of this replaces a conversation with your doctor. What I’ve found is that staying curious and then confirming with a clinician is a reasonable division of labor. </em><a href="https://docs.google.com/document/d/19f7aJtssFLLfSIQ5BJsbCcRXJOy8os0b9V-L-fpTZyQ/edit?tab=t.0"><em>Here’s a “jig” to get started</em></a><em>.</em></p> <h3>Experiment #4: Medication Overlap Check</h3> <p>I order most of my supplements through <a href="https://fullscript.com/">Fullscript</a> and track everything in a Google Sheet (e.g., supplement, dosage, frequency), which I share with my doctor periodically.</p> <p>I was embarking on an egg freezing journey and needed to add a prenatal to my routine. I wanted to figure out the right brand (e.g., Ritual, WeNatal, Needed, Modern Fertility, Perelel, others) and make sure nothing in my current supplement stack would interact with it.</p> <p>Three years ago this was a massive undertaking. Every brand includes different ingredients at various doses. I built a spreadsheet in 2021 that got unwieldy. Then I tried Apple’s HealthKit medication feature, but you have to add each supplement one by one. It was a tedious UX, and I gave up.</p> <p>(The only person I know who had the fortitude to build this system prior to AI was Natasja Nielsen — <a href="https://natasja.notion.site/Zygote-Freezing-4ba81f991fb54efdb222fd0c4c182176?pvs=149">here’s her write up</a>. She’s a legend, and now runs the <a href="https://openlabelresearch.org/">Open Label Research Foundation</a>.)</p> <p><em>What I tried this year:</em> I uploaded a <em>screenshot</em> of my Fullscript order and <em>photos</em> of the leading prenatal nutrition labels to Claude, then asked: Are there any interactions? Overlaps? Which prenatal makes sense given what I’m already taking?</p> <p>Claude flagged that I was doubling up on magnesium, compared the pre-natals side by side, and I sent the output to my doctor. She confirmed it looked good. Total time: under 10 minutes. <strong>I’m now taking fewer total supplements and spending less money.</strong></p> <p>One observation on incentives: this would be a hard feature for a Fullscript product manager to advocate to build. Their business benefits when I buy more, not fewer. An AI that advocates on my behalf rather than the marketplace changes that dynamic entirely.</p> <h3>Experiment #5: Building Muscle (Finally!)</h3> <p>Earlier this year, I booked a DEXA scan to check bone density and body composition, which was partly triaged from my Zoom interview with my mom (Experiment #2, family history of osteopenia). Bone density came back fine. But my muscle mass percentile was shockingly low.</p> <p>I felt frustrated — how could that be? I’d worked out with a trainer for years.</p> <p>I loaded the results into Claude and walked through my routine. The underlying issue wasn’t how much I was exercising, it was what I was (not) eating. I wasn’t getting enough grams of protein. And a lot of what I thought was protein was collagen-based (bone broth, Vital Proteins powder). Collagen is an incomplete protein that doesn’t contribute much to muscle growth.</p> <p>So I finally learned about macros. Shifted to ‘complete’ proteins (whey, chicken, tofu) and cut back on collagen-based ones, and also started eating more. I’m hitting PRs now, exercising more consistently, in less total time.</p> <p>The AI caught a marketing fallacy: collagen protein ≠ muscle-building protein. It’s unlikely I would have hit my muscle targets with my previous diet no matter how much I worked out.</p> <p><em>This is </em><a href="https://docs.google.com/document/d/19f7aJtssFLLfSIQ5BJsbCcRXJOy8os0b9V-L-fpTZyQ/edit?tab=t.0#heading=h.9saufxpssb7l"><em>similar to the jig #3</em></a><em>. One thing I’d recommend: In Claude’s “</em><a href="https://support.claude.com/en/articles/10185728-understanding-claude-s-personalization-features"><em>personal preferences</em></a><em>” </em><strong><em>ask it to “default to scientific evidence” </em></strong><em>and</em><strong><em> “ask clarifying questions if unsure.”</em></strong><em> It applies to every conversation and makes the outputs noticeably more useful for me.</em></p> <h3>Experiment #6: The Vaccine Paper Trail (Failed Attempt)</h3> <p>Did you have one of these growing up?</p> <figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-KqMcXN1esOXQD1HQOXnXg.png"/><figcaption>Thanks mom! From the author’s childhood records.</figcaption></figure> <p>My vaccine record was nearly a sacred document. My mom and doctor updated it every year, and I carried it to school in a little plastic folder so it wouldn’t get wet.</p> <p>Fast forward to 2026: I was about to get a flu shot at Walgreens and they asked about my last vaccine. Hm, the only place I could find it was my Google Calendar. I never miss logging health appointments, so I can search “flu” or “TDaP” or “COVID” and see the date and location.</p> <p>Feeling ambitious after experiments #1–5, I thought: surely in 2026 I can get a digital version of that childhood record. All my vaccine data, searchable, not in Google calendar–perhaps a nicely formatted vaccine record.</p> <p>I only get vaccines from three places (Walgreens, CVS, and One Medical) and all three let me export PDFs. My issue was combining them. I tried linking Walgreens and CVS to Apple HealthKit. It showed the integration but wouldn’t pull all the data, <em>only</em> <a href="https://support.apple.com/en-us/102467">COVID-19 records</a> (sigh, which I’m guessing was an artifact from when we had to prove vaccination to get into restaurants).</p> <p>The one partial workaround: One Medical lets me upload records via their chat concierge. I sent them all my CVS and Walgreens PDFs and they added them to my OM record. Useful, but I still had to collect and send everything myself. Not so different from that plastic folder from childhood. This time, though, I also saved everything in my <a href="https://www.sarahtavel.com/p/getting-started-with-claude-projects">Claude project</a>, meaning I’m still in that conductor mode.</p> <p>(I’m sharing this because I’d love for someone to fix this — and perhaps a product person at any of these companies is reading! I’d be happy to road test this workflow with you…)</p> <h3>On keeping humans in the loop</h3> <p>I prefer to keep humans in the loop (Why? So an LLM doesn’t, say, <a href="https://x.com/AdrianDittmann/status/2025904299944063046">nuke my inbox</a>). Automation is not my end-game objective for this project.</p> <p>I enjoy being <em>The Conductor</em>: making the judgment calls, observing the flow, catching what the AI misses. My AI system handles the tedious parts. I handle what requires extra context or trust. I’ve also made deliberate tradeoffs on security:</p> <ul><li><strong>For personal data such as email and calendar,</strong> I’m cautious. I generally don’t want any tool that can edit or delete. I strongly prefer AI interactions that are <em>read-only, </em>and provide suggested actions.</li><li><strong>For health data,</strong> I’m much more permissive than younger me would have expected. My threat model has shifted. I used to worry about someone accessing my health information. Now I worry more about not getting care because the system made it too hard, or dark patterns, or I gave up. The risk of inaction feels higher than the risk of sharing.</li></ul> <p>As I design my second brain-health system, I want to keep <strong>memory</strong>, <strong>compute</strong>, and <strong>interfaces</strong> separate. This impacts how I think about my data (Flexpa JSON, lab PDFs) + my AI models (Claude, other) + my interfaces (could be Slack, email, browser, terminal). I don’t want to be locked into one system. I can remix each component as needed and as each part of the stack evolves.</p> <p><a href="https://meghanverenajoyce.substack.com/?sort=top">Megan Verena Joyce</a>, CEO of Duckbill, made an observation about interfaces I often come back to: “The ideal surface through which people will actually consume AI remains genuinely unsettled — app, API, voice, agent, [or] something we don’t have a name for yet.”</p> <h3>An arms race: what happens as more players use AI?</h3> <p>What if everyone deployed these experiments at scale? Aetna can handle me as an edge case. At scale, the insurer would need to change. And if every provider used AI to upcode, costs would rise. We’d find ourselves in an arms race with no referee.</p> <p>According to <a href="https://www.garnerhealth.com/blog/healthcare-cost-crisis-drivers-report">this January’s Garner Health report</a>,<em> </em>the healthcare sector has adopted AI at 2.2x the rate of the broader economy, with its total spend tripling to $1.4B last year. But 89% of that capital is going into back-office tools designed to maximize provider revenue, which fuels exactly that arms race: providers upcoding, payers denying faster, “severity inflation” driving a 1.7% increase in employer healthcare spend.</p> <p>The full picture of who’s using AI for what:</p> <ul><li><strong>Providers:</strong> upcode (charge more for the same visit)</li><li><strong>Payers:</strong> deny claims faster</li><li><strong>Pharma: </strong>optimize pricing</li><li><strong>Patients</strong> (like me): get through the health admin slog</li></ul> <p>Everyone is optimizing for their position; few are improving the system. Until we fix the underlying incentives, this is what we get: warring algorithms extracting rents from each other.</p> <p>Perhaps I’m too cynical, and this is why I hope we can agree as a society that <em>everyone</em> has the right to healthcare — not only the <em>poor</em> via Medicaid, the <em>old</em> via Medicare, or the (non-1099) <em>fully-employed</em> via employer plans. The person delivering your Uber Eats likely falls into none of those categories, and as a result, is far less likely to afford and have health insurance.</p> <p>Until that changes at a policy level, the best any of us can do is learn to work the system better than it works us.</p> <h3>The Patient as a Creator</h3> <p>A conversation with Thorsten Wirkes at Bertelsmann helped me name why this moment matters: these AI tools let anyone be a creator. Not in the influencer sense, but in a more true-nature human sense.</p> <p>I’m reminded of Meinrad Craighead’s poem:</p> <blockquote>Whether we are weaving tissue in the womb or</blockquote> <blockquote>weaving imagery of the soul — our work is the</blockquote> <blockquote>work of conception, gestation and birth . . .</blockquote> <blockquote>the mother has but one law — create,</blockquote> <blockquote>make as I do . . . transform one substance</blockquote> <blockquote>into another . . . transmute blood into milk,</blockquote> <blockquote>clay into vessel</blockquote> <blockquote>feeling into movement</blockquote> <blockquote>word into song</blockquote> <blockquote>egg into child</blockquote> <blockquote>fiber into cloth</blockquote> <blockquote>stone into crystal</blockquote> <blockquote>memory into image and</blockquote> <blockquote>body into worship</blockquote> <p>Creation is the human birthright. And AI, at its best, helps us express that.</p> <p>What’s happening today feels like a bigger-scale version of what Substack did for writers, Shopify for merchants, or Excel for analysts. The barrier to doing something that once required an institution (capital, or a lot of software engineers) dropped low enough that individuals (patients) can do it themselves.</p> <p>You’re a conductor now. Not because institutions handed you that power, but because the technologies got capable enough that you can take it. Put that jetpack on.</p> <p><strong><em>Thanks to:</em></strong><em> </em><a href="https://www.linkedin.com/in/thorstenwirkes/"><em>Thorsten Wirkes</em></a><em>, </em><a href="https://www.linkedin.com/in/tim-schneider1290/"><em>Tim Schneider</em></a><em>, </em><a href="https://www.linkedin.com/in/ginnyfahs/"><em>Ginny Fahs</em></a><em>, </em><a href="https://www.linkedin.com/in/juliajchou/"><em>Julia Chou</em></a><em>, </em><a href="https://www.linkedin.com/in/reshmakhilnani/"><em>Reshma Khilnani</em></a><em>, </em><a href="https://www.linkedin.com/in/johnwilbanks/"><em>John Wilbanks</em></a><em>, </em><a href="https://www.linkedin.com/in/andrewarruda/"><em>Andrew Arruda</em></a><em>, </em><a href="https://www.linkedin.com/in/rachellkatz/"><em>Rachel Katz</em></a><em>, </em><a href="https://www.linkedin.com/in/vincehartman/"><em>Vince Hartmann</em></a><em>, </em><a href="https://www.linkedin.com/in/caroline-reiner/"><em>Caroline Reiner</em></a><em>, </em><a href="https://www.linkedin.com/in/christina-bognet-90a87718/"><em>Christina Bognet</em></a><em>, and </em><a href="https://www.linkedin.com/in/john-weldon-517a52188/"><em>John Weldon</em></a><em> and many others for conversations that shaped this piece. And </em><a href="https://www.linkedin.com/in/adam-conner-simons/"><em>Adam Conner-Simon</em></a><em> for editing support!</em></p> <h3><em>Afterward</em></h3> <h4>Getting unstuck: learning to retire my old mental models</h4> <p>A few years ago, I wrote a piece on how<a href="https://blog.andreacoravos.com/software-enabled-clinical-trials-8da53f4cd271"> software was “eating” clinical trials</a>. Fragments from that work turned into a company called HumanFirst, which we sold to ICON (NASDAQ: ICLR) in 2024.</p> <p>After the sale, I took a sabbatical and tried to apply similar frameworks (“Jobs to be Done,” market maps) to understand how AI is changing care delivery. None of my drafts worked. Eventually I realized why: “Jobs to be Done” puts the doctor at the center of things with the linear flow I kept mapping (i.e. onboarding &gt; check-in &gt; visit &gt; orders &gt; billing). But patients don’t experience healthcare that way. We have multiple doctors, overlapping needs, things happening all at once. A linear framework (even a patient-friendlier one) wasn’t going to be the answer.</p> <p>Similar problem with how I was approaching personal health records (PHRs). I kept thinking of them as apps. But an app is a company-level view and reflects a product built <em>for</em> patients, not <em>by</em> them. My old frameworks and mental models were breaking in the age of AI. That felt disconcerting, and also freeing.</p> <p>The Flexpa experience was my big ah-ha. Decentralization isn’t inherently good or bad — it’s more of the current, unchangeable state of things. Years of interoperability policy and “one place for everything” efforts have been well-intentioned and largely haven’t worked for the patient experience. My HealthKit is separate from my labs, claims, supplement list and MRIs. But in 2026 that turns out not to matter as much as I thought. LLMs can synthesize across siloed sources.</p> <p>All these years, I’d been thinking about decentralization wrong. <strong>It’s less an <em>engineering problem</em> and more a <em>governance one</em>.</strong> Flexpa, Claude and Duckbill act as a coordination layer. If it works well, clinicians move up the stack into oversight and discernment, and stop being some of the most expensive data entry workers on earth (h/t to <a href="https://www.linkedin.com/posts/y0ussef_healthcare-20-is-emerging-openai-and-anthropic-activity-7417861595164340224-xwv1/?utm_medium=ios_app&amp;rcm=ACoAAADy7IQBa1nXPTxLmzQSTCxaoCptoGtHSsA&amp;utm_source=social_share_send&amp;utm_campaign=copy_link">Dr. Youssef Aboufandi</a> for that framing). This picture unlocked these concepts for me:</p> <figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*tKQCUiQG71lpWBYdAs-wvQ.jpeg"/><figcaption><a href="https://www.linkedin.com/posts/y0ussef_healthcare-20-is-emerging-openai-and-anthropic-activity-7417861595164340224-xwv1/?utm_medium=ios_app&amp;rcm=ACoAAADy7IQBa1nXPTxLmzQSTCxaoCptoGtHSsA&amp;utm_source=social_share_send&amp;utm_campaign=copy_link">Image by Dr. Aboufandi</a></figcaption></figure> <p>There’s more shared responsibility. The data and needs are <em>compiled</em> by the patient, and then <em>confirmed</em> by the doctor, rather than the other way around (as it’s traditionally been).</p> <p>I also like that my new system removes emotion from transactions that are usually loaded with it. My insurer doesn’t have to deal with a confused, frustrated patient on hold. They get a clear, properly formatted adjudication letter and some scripts from a dispassionate Duckbill human calling on my behalf. That turns out to be useful for everyone involved (I hope!).</p> <p>As Flexpa CEO Andrew Arruda put it: “Your health data shouldn’t live in an app. It should sit with you and plug into whatever you want.” It took me a while to understand what that meant in practice, but now I get it.</p> <h4>A ‘second brain’ for health (and why it only works in the past few months)</h4> <p>When I set out on this journey, I didn’t intend to build a next-gen personal health record. It organically emerged from the experiments. Each time I added a new data source (claims, labs, family history, supplements) the Claude project got a little more useful. After a few months, I had something that could not merely store information, but advocate on my behalf.</p> <p>A hot concept these days is the “second brain,” a personal knowledge management system that offloads memory and surfaces things when you need them. Popularized by <a href="https://www.buildingasecondbrain.com/">Tiago Forte</a> and expanded on recently by <a href="https://www.youtube.com/watch?v=0TpON5T-Sw4">Nate B. Jones</a>, it acknowledges a key insight: storage systems fail because they ask you to do cognitive work at the wrong moment, right when you just want to capture something and move on. Old-school patient portals had these problems, and then some.</p> <p>Here’s some differences between the old approach and the “second brain” I’ve been building:</p> <figure><img alt="" src="https://cdn-images-1.medium.com/max/570/1*xS0Lkj5LXZlOfTXyvVNsSQ.png"/></figure> <p>In practice, my new system is simple: I dump everything into one project. I don’t build folder structures or differentiate between a bill, lab result, or post-appointment voice memo. My AI models do the routing, and I ask for specific outputs, not open-ended reviews. (Think: less “analyze my claims,” more “what does Aetna owe me and what’s the script to get it?”)</p> <h3>Reflections from founders</h3> <p>To prepare this piece, a few conversations have lived in my head rent free.</p> <h4><strong>The Abridge View: The Conversation Is the Unit</strong></h4> <p>From a conversation with <a href="https://www.linkedin.com/in/johnwilbanks/">John Wilbanks</a> (formerly of Astera) and <a href="https://www.linkedin.com/in/juliajchou/">Julia Chou</a> (COO of<a href="https://www.abridge.com/"> Abridge</a>):</p> <blockquote>“The atomic unit of the patient experience is the conversation: patient and clinician. That conversation gets transformed into downstream artifacts: chart notes, care plans, referrals.”</blockquote> <p>Scribes used to be passive note-takers. Now they’re verifying chronic conditions, checking prior authorization, and supporting billing. The conversation contains more context than the note. Scribes are building an intelligence layer that captures what the EHR can’t.</p> <p>Wilbanks asked a question I keep thinking about: <em>“How do you get from ‘bullshit’ to ‘holy shit’? How do you build toward a trusted conversational relationship?”</em></p> <h4>The Abstractive View: Patient Intelligence</h4> <p>From a whiteboard session with Vince Hartmann and Caroline Reiner at<a href="https://www.abstractivehealth.com/"> Abstractive Health</a>:</p> <p>EHRs captured the record, and often lost the context. Scribes can capture more context but it’s often locked in transcripts. The next layer, what Vince calls <em>Patient Intelligence</em>, stitches together encounters, claims, pharmacy, devices, and social context into one view. No one does this well (yet). That’s what I’ve been doing manually with my second brain and the experiments above, and the team at Abstractive does even better at an enterprise-level.</p> <h4>The Medplum View: Headless EHRs and Patient-Built Stacks</h4> <p>In a conversation with <a href="https://www.linkedin.com/in/reshmakhilnani/">Reshma Khilnani</a>, CEO of Medplum (note: I’m an advisor), I noticed that what she’s building as an open-source company and what I’ve been building as a patient follow similar logic, just from opposite directions.</p> <p>Companies like Medplum, Welkin and Healthie are building “<a href="https://elion.health/resources/headless-ehrs">headless EHRs</a>,” which separate the data layer from the interface layer, so clinics have more control over patient experience without being locked into Epic, which holds records for roughly 300 million Americans and bundles everything together: database, workflows, billing, scheduling, patient portal (MyChart).</p> <p>My stack follows the same architecture without the name: Flexpa for data/ PDFs, Claude for compute, and then whatever interface I need that day. The Epic monolith is getting decomposed from both ends. No one can compete with it head-on, but lots of people are building around it: <em>patients</em> building their own stacks, <em>clinics</em> choosing flexible infrastructure.</p> <p>How those two movements meet is one of the more interesting questions in health tech right now in my view.</p> <h3>The Experiments: A Cheat Sheet</h3> <figure><img alt="" src="https://cdn-images-1.medium.com/max/638/1*8U6hDUVS8dKLRObzU-lZ0Q.png"/></figure> <p><img src="https://medium.com/_/stat?event=post.clientViewed&amp;referrerSource=full_rss&amp;postId=8459e2bfe0f1" width="1" height="1" alt=""/>&lt;hr&gt;&lt;p&gt;<a href="https://blog.andreacoravos.com/the-human-as-conductor-8459e2bfe0f1">The Human as Conductor</a> was originally published in <a href="https://blog.andreacoravos.com">Andrea’s Blog</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.&lt;/p&gt;</p>]]></content><author><name></name></author><category term="technology"/><category term="ai"/><category term="ai-in-healthcare"/><category term="everyday-life"/><category term="healthcare-reform"/></entry><entry><title type="html">The founder’s guide to funding health and science organizations</title><link href="https://andreacoravos.com/writing/2025/the-founders-guide-to-funding-health-and-science-organizations/" rel="alternate" type="text/html" title="The founder’s guide to funding health and science organizations"/><published>2025-06-20T22:44:18+00:00</published><updated>2025-06-20T22:44:18+00:00</updated><id>https://andreacoravos.com/writing/2025/the-founders-guide-to-funding-health-and-science-organizations</id><content type="html" xml:base="https://andreacoravos.com/writing/2025/the-founders-guide-to-funding-health-and-science-organizations/"><![CDATA[<h4><em>New collaboration with Astera Institute on funding your idea</em></h4> <p>A few months ago, <a href="https://www.linkedin.com/in/cate-hall-9a81a35/">Cate Hall</a> asked me: “where’s the fundraising handbook that guides scientists through how the game actually works? And not the sanitized version VCs publish on Twitter.” The information asymmetry between founders and investors is real. VCs negotiate term sheets daily. You might do it a few times in your career. We’re flipping that script. Our 50+ page handbook was born from Cate’s question.</p> <p>My co-author <a href="https://www.linkedin.com/in/rachellkatz/">Rachel Katz</a> and I have had time to reflect since we individually raised venture funding and sold our respective companies to public companies. We interviewed top entrepreneurs and nonprofit leaders to bring these whisper network lessons directly to you.</p> <figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*UKMjilZ0JW3oycbIiu8EGA.jpeg"/></figure> <p>The things that somehow never make it into polite fundraising guides:<br/>- A 101 guide to entity types and sources of capital for today’s environment <br/>- Tools for building momentum, even if you’re running out of runway<br/>- The hierarchy of investor introductions<br/>- Whisper network tactics on negotiating term sheets<br/>- Working back from your exit: how investors will get paid (and how employees can too)</p> <p>This work stands on the shoulders of giants. For invaluable fundraising conversations and perspectives that shaped this handbook, big thank you to <a href="https://www.linkedin.com/in/emcohen/">Elliot Cohen</a>, <a href="https://www.linkedin.com/in/mtesquivel/">Michael Esquivel</a>, <a href="https://www.linkedin.com/in/lmarrone/">Lisa Marrone</a>, <a href="https://www.linkedin.com/in/sumnernorman/">Sumner L Norman</a>, <a href="https://www.linkedin.com/in/wes-panek-61a3b143/">Wes Panek</a>, <a href="https://www.linkedin.com/in/bcppatterson/">Brian C. Patterson</a>, <a href="https://www.linkedin.com/in/wgpatrick/">Will Patrick</a>, <a href="https://www.linkedin.com/in/conorrussomanno/">Conor Russomanno</a>, <a href="https://www.linkedin.com/in/zachweinberg/">Zach Weinberg</a>, <a href="https://www.linkedin.com/in/carolyn-witte-03b50a26/">Carolyn Witte</a> (and many more founders who shared their insights on background) for your thoughtful engagement with these ideas.</p> <p>The editorial refinement of this work would not have been possible without <a href="https://www.linkedin.com/in/christina-agapakis-1017b5/">Christina Agapakis</a> and <a href="https://www.linkedin.com/in/ramyamahalingam/">Ramya Mahalingam</a>, whose careful attention to detail and clarity significantly enhanced the manuscript, and <a href="https://www.linkedin.com/in/michela-caffrey-designer/">Michela Caffrey</a> for skillful layout design. Any shortcomings remain entirely our own.</p> <p>I’m grateful to <a href="https://www.linkedin.com/company/astera-institute/">Astera Institute</a> for supporting open-access work that challenges the status quo — and for giving me a chance to reflect on the past few years and pay it forward.</p> <p>We wanted to give you tools, not rules. Our goal: help you get funding efficiently so you can focus on getting world-shifting health and science innovations to market faster. Keep going. We’re rooting for you.</p> <p><strong>Access the handbook</strong>: <a href="https://astera.org/inside-baseball-the-founders-guide-to-funding-health-and-science-organizations/">https://astera.org/inside-baseball-the-founders-guide-to-funding-health-and-science-organizations/</a></p> <p><img src="https://medium.com/_/stat?event=post.clientViewed&amp;referrerSource=full_rss&amp;postId=952afdaaa446" width="1" height="1" alt=""/>&lt;hr&gt;&lt;p&gt;<a href="https://blog.andreacoravos.com/the-founders-guide-to-funding-health-and-science-organizations-952afdaaa446">The founder’s guide to funding health and science organizations</a> was originally published in <a href="https://blog.andreacoravos.com">Andrea’s Blog</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.&lt;/p&gt;</p>]]></content><author><name></name></author><category term="fundraising"/><category term="health"/><category term="open-access"/><category term="entrepreneurship"/><category term="science"/></entry><entry><title type="html">Angel Investing in the Arts, Part II: Broadway Edition</title><link href="https://andreacoravos.com/writing/2025/angel-investing-in-the-arts-part-ii-broadway-edition/" rel="alternate" type="text/html" title="Angel Investing in the Arts, Part II: Broadway Edition"/><published>2025-06-16T15:13:06+00:00</published><updated>2025-06-16T15:13:06+00:00</updated><id>https://andreacoravos.com/writing/2025/angel-investing-in-the-arts-part-ii-broadway-edition</id><content type="html" xml:base="https://andreacoravos.com/writing/2025/angel-investing-in-the-arts-part-ii-broadway-edition/"><![CDATA[<h4>If you believe in backing founders, you can back storytellers too. This series shares how you can angel invest in the arts and the creation of new worlds.</h4> <p><em>A companion piece to </em><a href="https://blog.andreacoravos.com/angel-investing-in-the-arts-film-and-movies-9934da81ce0f"><em>Part I: Angel Investing in Film</em></a></p> <figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*D5n9XJxlbhe-RN4P"/><figcaption>Source: <a href="https://www.google.com/url?q=https://www.newyorktheatreguide.com/show/24495-suffs-musical-tickets&amp;sa=D&amp;source=docs&amp;ust=1749772754688364&amp;usg=AOvVaw1F8iMC0mrG93f31JlJIrAs">https://www.newyorktheatreguide.com/show/24495-suffs-musical-tickets</a></figcaption></figure> <p>In my last post, I wrote about why some of us in the tech world are investing in independent documentary films: the production timelines, creative updates, and exit models can feel surprisingly familiar to startup/tech investing. There’s another corner of the creative world that also fits neatly into this portfolio mindset: <em>Broadway</em>.</p> <p>This past year, I had the grand experience of investing in <a href="https://suffsmusical.com/"><em>Suffs</em></a>, a musical about the women’s suffrage movement written by Shaina Taub. It opened at the Music Box Theatre, and earlier this year, we found out it was <em>nominated for six Tony Awards and won two</em>. A pinch-me moment, to say the least for my first Broadway investment.</p> <p>What made it even more special was that I invested through <a href="https://www.nvpbroadway.com/">Nothing Ventured Productions (NVP)</a>, a Broadway investment collective founded by my college friend Kathryn Minshew, along with Alexandra Cavoulacos and Christina Wallace, all successful tech entrepreneurs who brought the angel investing playbook into theater production.</p> <p>Kathryn and I go way back to our days at Duke where we hilariously took a ‘wilderness skills’ class together. She’s also known for founding The Muse, a career platform used by over 70 million people to research companies and careers. Now she’s extending that entrepreneurial DNA to the world of live performance, and bringing her friends along for the ride.</p> <p>NVP is building a portfolio of Broadway projects, just like an angel syndicate might invest in a batch of seed-stage startups. They use a roll-up vehicle model to pool checks from accredited investors, often with lower minimums than a traditional Broadway investment would require. It’s democratizing access to a historically closed world.</p> <h3>Why Broadway?</h3> <p>Unlike film or tech, Broadway shows offer relatively clear and immediate cash flow visibility. You know exactly how many tickets are sold, what the merchandise numbers look like, and what’s coming in the door week by week. To me that transparency is rare and frankly refreshing. Here’s how one recent NVP investor update broke it down:</p> <p><strong>Wrap</strong> = ticket sales value that day<br/><strong>Gross</strong> = ticket value from attendees that day<br/><strong>Advance</strong> = value of tickets sold for future performances<br/><strong>Groups not yet paid</strong> = confirmed group sales pending cash receipt</p> <p>You can think of these metrics a bit like MRR or bookings in a SaaS business, signals of demand and health, plus the added bonus of nightly applause and regular critical theater reviews.</p> <h3>Broadway as an asset class</h3> <p>Like film, Broadway offers a spectrum of opportunities across genre and scale. Here’s a quick primer:</p> <ul><li><strong>Original Musicals </strong>(like Suffs): high risk, high reward, often mission-driven</li><li><strong>Revival Musicals:</strong> more predictable, often tied to nostalgia or recognizable IP</li><li><strong>Plays:</strong> generally lower budgets, but sometimes smaller commercial upside (similar to musicals that new plays are higher risk/reward, establish plays lower risk/reward)</li><li><strong>Limited Engagements/Touring:</strong> opportunities to recoup via short runs or roadshows</li></ul> <p>For the investors the terms are generally equity-style participation; some offer recoupment waterfalls and royalty participation, plus the occasional invite to Suffs-related parties and performances.</p> <h3>The Suffs Model</h3> <p>From a capital perspective, <em>Suffs</em> followed a fairly standard budget structure for a Broadway musical of this scale, raising just around $19 million to get to opening night (<a href="https://www.nytimes.com/2023/10/18/theater/suffs-broadway-hillary-clinton-malala.html">NYTimes</a>). That included “hard costs” (like set construction, cast and crew salaries, rehearsal space, and administrative overhead) as well as reserves for previews, marketing, and contingencies.</p> <p>Once a show opens, Broadway becomes a weekly cash flow business. Broadway musicals operate with a recurring cost structure that typically ranges from $400K-$1M per week, depending on the scale of the show, the venue, and marketing needs. That includes everything from salaries and rentals to advertising and theater expenses. You can think of it like a startup’s monthly burn, and break-even is tied to consistently hitting a certain occupancy threshold at premium ticket pricing.</p> <p>For investors, this opens up a unique kind of financial visibility. You’re not waiting years for a liquidity event. You see gross ticket sales every week, along with real-time metrics on advance sales (future demand), merchandise revenue (non-ticket upside), and even line-item profitability estimates. Some investor updates even include “recoupment scenario” charts showing weeks to break even at various levels of capacity (e.g. 70%, 80%, 90% full houses), much like a SaaS model might show customer acquisition cost payback or time to profitability.</p> <p>Of course, recoupment isn’t guaranteed. But with strong press, savvy pricing strategy, and energetic audience response, the upside can move quickly and in the case of <em>Suffs</em>, momentum only grew post-opening though it ended its Broadway run earlier than we had hoped. It’s now shifting to a National Tour later this fall.</p> <p>As with angel investing, risk is real. Many shows never recoup. But there are also scenarios, especially with strong press, merchandise traction (Suffs sold thousands of merch opening night), and big-name backers (the <em>Suffs</em> producers include Hillary Clinton and Malala Yousafzai), where things can tip from break-even to upside.</p> <h3>The investor experience</h3> <p>One of the most fun parts of being involved with <em>Suffs</em> was the community from early access to dress rehearsals, to special merch collabs, to getting a behind-the-scenes look at how the show was evolving in previews. Investor updates included real-time box office strategy (“flash sales” to drive demand), press updates (e.g., New York Times, The New Yorker, Playbill), and performance recaps. The investor comms read like a blend of a portfolio company update plus a fan letter.</p> <figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*U20zd9oNNpDEqx_T"/></figure> <p>When I finally saw the show live with friends, it was electric. Here’s a photo of me holding the program in the Music Box Theatre. We were beaming. Felt special to watch it knowing we were some of the financial backers to make this story possible.</p> <h3>Closing curtain</h3> <p>If you’re a tech investor with a love of storytelling, community, and creative exposure, Broadway might be your next unexpected asset class. Your investment might not always win a Tony, but you might just help bring something beautiful into the world, and learn a new world along the way.</p> <p>Organizations like <a href="https://www.nvpbroadway.com/">Nothing Ventured Productions</a> are making Broadway investing more accessible, and there are other groups and individual producers creating similar opportunities for those interested in supporting live theater. Drop them a line if you’re interested in upcoming shows.</p> <p><strong><em>Note:</em></strong><em> This is not investment advice. This is me on sabbatical, having fun with the time I am on this planet, supporting people doing brave things and telling stories that I’m excited to see in the world. I get zero financial benefits from sharing this post, except that perhaps more of my friends will start supporting the arts if they didn’t know it was within reach.</em></p> <p><em>Lmk if you decide to jump in.</em></p> <p><img src="https://medium.com/_/stat?event=post.clientViewed&amp;referrerSource=full_rss&amp;postId=0514c36e6407" width="1" height="1" alt=""/>&lt;hr&gt;&lt;p&gt;<a href="https://blog.andreacoravos.com/angel-investing-in-the-arts-part-ii-broadway-edition-0514c36e6407">Angel Investing in the Arts, Part II: Broadway Edition</a> was originally published in <a href="https://blog.andreacoravos.com">Andrea’s Blog</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.&lt;/p&gt;</p>]]></content><author><name></name></author><category term="arts-and-culture"/><category term="how-to"/><category term="angel-investors"/><category term="broadway"/></entry><entry><title type="html">Angel Investing in the Arts: Film and Movies</title><link href="https://andreacoravos.com/writing/2025/angel-investing-in-the-arts-film-and-movies/" rel="alternate" type="text/html" title="Angel Investing in the Arts: Film and Movies"/><published>2025-06-04T13:18:51+00:00</published><updated>2025-06-04T13:18:51+00:00</updated><id>https://andreacoravos.com/writing/2025/angel-investing-in-the-arts-film-and-movies</id><content type="html" xml:base="https://andreacoravos.com/writing/2025/angel-investing-in-the-arts-film-and-movies/"><![CDATA[<h4>If you believe in backing founders, you can back storytellers too. This series shares how you can angel invest in the arts and the creation of new worlds.</h4> <figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/0*lT3wK3BUIv-nfaCA"/></figure> <p>In the summer of 2021, Linda Xie <a href="https://x.com/ljxie/status/1415377341413347328">tweeted</a> a Mirror.xyz link to crowdfund a new documentary that she was co-producing named <em>Vitalik: An Ethereum Story</em>, a character-driven documentary film on the challenges and rewards of building a new world.</p> <p>In just <a href="https://deadline.com/2021/07/optimist-ethereum-gamestop-meme-stocks-vitalik-buterin-1234797317/">over 48 hours, the film team raised $2 million worth of Ethereum</a> for the production of the documentary. Some donors received producer credits and NFTs created by digital artist pplpleaser. On April 15, 2025, the film premiered on Apple TV and Amazon Prime Video.</p> <p>I was inspired by this exposure because historically my only alternative investments had been in VC-backed start-ups and I didn’t realize film was an asset class I could participate in.</p> <p>After Linda posted the 2021 tweet, I went to <a href="https://www.goodside.xyz/">Sundance</a> and met <a href="https://en.wikipedia.org/wiki/Jennifer_8._Lee">Jenny 8 Lee</a>, documentary producer who runs <a href="https://www.goodside.xyz/people">Goodside</a>, a community of film lovers who convene at festivals like Sundance, Cannes, and SXSW. Part of what I love about entrepreneurship is not only building new products, but also building communities and (I hope!) improving the way societies exist and evolve. Perhaps most importantly, the evolution of any new world originates with storytelling.</p> <p>As a newcomer to the world of investing in the arts, I wanted to share my journey so if other folks in tech wanted to jump in, you don’t have to start from zero.</p> <p>So far, I’ve invested in a documentary via a grant mechanism, and I’m evaluating an equity or grant investment in a second film. I’ve also made a grant into a short story platform that’s used as lead gen for filmmakers.</p> <p><strong><em>For context:</em></strong><em> I’m the former co-founder and CEO of HumanFirst, acquired by ICON plc (NASDAQ: ICLR) earlier this year, and in a past life I worked at KKR, a private equity firm. Over the past 10+ years, I’ve made 65+ angel investments– mostly small checks in start-up companies, as well as in a few VC and PE funds.</em></p> <h3>Decision point: narrative or documentaries (‘docs’)</h3> <p>If you’d want to extend the analogy from start-up company to film roles, here’s my favorite cast on how to think about the positions:</p> <figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*xn8n0K_krr1aOIVK"/><figcaption>Source: <a href="https://farcaster.xyz/linda/0xc23f804d">https://farcaster.xyz/linda/0xc23f804d</a></figcaption></figure> <p>At a high level, there are two types of <em>independent</em> films: <strong>narrative</strong> and <strong>documentary</strong>.</p> <p>Independent films and those developed directly by distributors (like Apple, HBO, or Netflix) follow very different funding and production models. Unlike studio films, indie films typically don’t have distribution secured at the outset. That said, a streamer or distributor may acquire or license an independent film later in its lifecycle. In fact, that’s often the goal.</p> <p>Here’s a simplified comparison to help you think through how to support each type of independent film.</p> <figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*PIbsTfopNLa7tw3nj7srBw.png"/></figure> <h3>Case Study: Two documentaries I’m supporting</h3> <p>The introductions to both documentaries were organic. Linda introduced me to Carrie Weprin, who produced the Vitalik film. Then I met Jessie Seng, when she was matched as my roommate at Sundance last year in the Goodside house.</p> <figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Roc6KVXA2luHyScmhUmO-A.png"/></figure> <h3>Decision point: equity or grant investments</h3> <p>When investing in film, you can choose to give grants or invest for equity (or do a blend).</p> <ul><li><strong>Grants</strong> are typically made to a 501(c)(3) fiscal sponsor. Your support is philanthropic. These are often tax-deductible and aligned with impact goals. You can use a <a href="https://www.fidelitycharitable.org/giving-account/what-it-costs.html">Donor-Advised Fund (DAF)</a> to facilitate this.</li><li><strong>Equity</strong> means you’re treated as an investor in the film’s LLC. If the film makes money, you get a return based on a waterfall structure (after the film recoups production costs).</li></ul> <p>For example the <a href="https://strangedesignfilms.com/#holds-you-tight">*<em>holds you tight*</em> team</a> is raising both:</p> <ul><li><strong>Equity Investments </strong>for individuals looking for financial return. Investors are entitled to recoup their principal investment plus a premium and a share in the film’s backend profits. The *<em>holds you tight*</em> filmmakers are capping the amount of equity raised to maximize the potential for return on investment (ROI). Equity investments start at $50k.</li><li><strong>Grants &amp; Donations</strong> (via Women Make Movies as fiscal sponsor). Grants and donations receive preferential credit placement to equity investors. (Update on June 5, 2025: You can make a grant directly via their <a href="https://strangedesignfilms.com/#holds-you-tight">Tribeca Film Festival landing page</a>.)</li></ul> <h3>Communication &amp; investor updates</h3> <p>The investor updates are what I’ve enjoyed the most. In start-ups, investor updates are bullet-pointed, objective, and often formulaic:</p> <ul><li>Revenue</li><li>Product shipped</li><li>Customers won/lost</li><li>What we need help with</li></ul> <p>But in film, the updates are creative expressions themselves.</p> <p><strong>Jessie Seng</strong>, director of <em>FTNL</em>, sent one of the most moving early updates I’ve received. She shared not just logistical updates, but also the emotional journey of what it means to fundraise, create, and stay independent as a first-time indie filmmaker after having big production experience at HBO Max. Her email mixed reflections about Cambodian American history being “not a fit” for many traditional funders, the cost of staying independent to center survivors’ stories instead of sensationalizing perpetrators, and small victories- like receiving a grant after filming a powerful interview with a survivor.</p> <p>Jessie’s updates are raw, emotional, and deeply personal… a reminder that investing in an independent documentary film is often about supporting a creator’s lived experience, not just a final product.</p> <p>As another example, <strong>Carrie Weprin</strong>’s updates for <em>*holds you tight*</em> are structured, polished, and production-driven. Her update included:</p> <ul><li>A detailed phase-by-phase <strong>budget</strong> and <strong>timeline</strong> overview (development, production, post-production)</li><li>Clear fundraising<strong> targets</strong> tied to each phase</li><li>Different<strong> contribution tiers </strong>(starting at $5,000 for “Special Thanks” credit, with higher tiers offering Co-Executive Producer credits)</li><li>A <strong>planned cadence of quarterly updates</strong> with milestones (shooting, editing, grant outcomes, festival submissions)</li></ul> <p>Carrie’s approach reflects operational rigor almost like a seed-stage start-up update, while maintaining a creative and mission-driven heartbeat. Her materials made it easier for me to onboard to the terminology and expectations of film development. That’s part of my goal: I hope that by financially supporting these projects, I’m also creating a playground to learn a new industry.</p> <p>All to say, both styles are brilliant. Understanding the filmmaker’s approach will help you set your expectations as an arts investor and also give you a view into what stage you’d want to start participating in.</p> <p>So far with respect to timing, I have chosen to support film projects in the early stages. Investing is not my full time job, so I don’t have the capacity to do much diligence. My first checks are small — and often grants. I look for teams that will share regular investor updates, similar to what I get with my angel VC-backed portfolio companies.</p> <p>If the filmmakers are strong communicators and share regular updates, I may switch from grant support to equity investments later as the filmmakers have further developed the IP and as startup people like to say: ‘derisked’.</p> <figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*d_x0pU2WNp3KJTnZzWTYZQ.png"/></figure> <p>Timelines can vary dramatically — 18 to 36+ months is not uncommon for docs.</p> <h3>A changing landscape and cuts to public arts funding</h3> <p>Another important context to understand: traditional funding sources for independent film, like grants from the National Endowment for the Humanities (NEH), are <a href="https://www.washingtonpost.com/style/2025/04/07/neh-grants-canceled-cuts/">facing significant and historic budget cuts</a>. For filmmakers like Jessie Seng, this had a direct impact, as her previously-awarded California Humanities grant was <a href="https://www.kqed.org/arts/13974246/neh-funding-canceled-grants-california-humanities">canceled a few weeks ago</a>.</p> <p>Jessie and Kitty shared that as NEH and similar federal institutions reduce their arts funding, independent filmmakers are increasingly having to look for non-traditional funders, like individuals, private foundations, and alternative models including direct angel investments (could be me and you!).</p> <p>This creates both a challenge and an opportunity. The challenge is obvious: fewer institutional grants means more competition and more instability for critical storytelling projects. But the opportunity is that individuals like us, who may have historically focused only on start-ups, can now step in to support creators who might otherwise be left out of traditional funding pipelines.</p> <p>Supporting films like <em>FTNL</em> and <em>*holds you tight*</em> isn’t just about backing a project. it’s about keeping critical, human stories alive in a shrinking public arts landscape.</p> <h3>Lesson: ask for what you want</h3> <p>Just like in VC, you don’t get what you don’t ask for. The same applies here. If there’s a certain participation strategy or amount you’d like to invest… just ask.</p> <p>Both <em>holds you tight</em> and <em>FTNL</em> are actively raising funds and welcoming new supporters. If this inspires you to make your first angel investment in the arts, reach out to the producers:</p> <ul><li>📩 Carrie Weprin (<em>holds you tight</em>): <a href="mailto:strangedesignfilms@gmail.com">strangedesignfilms@gmail.com</a></li><li>📩 Kitty Hu (<em>FTNL</em>): <a href="mailto:kitty@shoesoffmedia.com">kitty@shoesoffmedia.com</a></li></ul> <p>Even if you’re just exploring, they’d love to connect. Ask for their pitch deck or one-pager to start a conversation.</p> <p>Kitty also generously offered to host informal “office hours” for new folks who want to learn more about the documentary funding landscape broadly, or what it’s like to support a film as a first-time backer. Send her a note if you’re interested.</p> <p><strong><em>Note:</em></strong><em> This is not investment advice. This is me on sabbatical, having fun with the time I am on this planet, supporting people doing brave things and telling stories that I’m excited to see in the world. I get zero financial benefits from sharing this post, except that I hope more friends realize that supporting the arts is closer and more accessible than they might have thought.</em></p> <p><em>Let me know if you decide to jump in.</em></p> <p><img src="https://medium.com/_/stat?event=post.clientViewed&amp;referrerSource=full_rss&amp;postId=9934da81ce0f" width="1" height="1" alt=""/>&lt;hr&gt;&lt;p&gt;<a href="https://blog.andreacoravos.com/angel-investing-in-the-arts-film-and-movies-9934da81ce0f">Angel Investing in the Arts: Film and Movies</a> was originally published in <a href="https://blog.andreacoravos.com">Andrea’s Blog</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.&lt;/p&gt;</p>]]></content><author><name></name></author><category term="investing"/><category term="movies"/><category term="angel-investors"/><category term="film"/><category term="documentary"/></entry><entry><title type="html">The Early Stage Go-To-Market Handbook</title><link href="https://andreacoravos.com/writing/2025/the-early-stage-go-to-market-handbook/" rel="alternate" type="text/html" title="The Early Stage Go-To-Market Handbook"/><published>2025-05-09T14:22:31+00:00</published><updated>2025-05-09T14:22:31+00:00</updated><id>https://andreacoravos.com/writing/2025/the-early-stage-go-to-market-handbook</id><content type="html" xml:base="https://andreacoravos.com/writing/2025/the-early-stage-go-to-market-handbook/"><![CDATA[<h4><em>New collaboration with Astera Institute on selling software to pharma and biotech companies</em></h4> <p>In the early days of <a href="https://www.linkedin.com/company/gohumanfirst/">HumanFirst</a>, structuring commercial contracts was one of our biggest challenges — how should we set pricing, protect IP, and package our software and services products. I was fortunate to have founder friends and other execs share their MSAs, order forms, and customer agreement templates, which proved invaluable.</p> <p>During my sabbatical, I’ve had the opportunity to reflect on our early commercial decisions. Together with <a href="https://www.linkedin.com/in/abby-devito-60517b107/">Abby DeVito</a> (our Head of GTM at HumanFirst), <a href="https://www.linkedin.com/in/shu-h-b16a8122/">Shu Hu</a> (our counsel at Gunderson who advises the fastest growing life sciences startups), and <a href="https://www.linkedin.com/company/astera-institute/">Astera Institute</a>, we’ve created <a href="https://astera.org/pharma-sales-handbook/">the comprehensive guide</a> we wished we’d had from day one.</p> <figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*8u0qovDhQZPZCWRAg43zsg.jpeg"/></figure> <p>This handbook would have saved me months (if not years) of iteration to establish solid agreements — the true foundation of strong customer relationships. We’re now sharing these early-stage learnings as an open-access resource for founding teams navigating pharma and biotech SaaS partnerships for the first time: <a href="https://astera.org/pharma-sales-handbook/">https://astera.org/pharma-sales-handbook/</a>.</p> <p>This work stands on the shoulders of giants. For invaluable commercial conversations and perspectives over the years that turned into this handbook, big thank you to <a href="https://www.linkedin.com/in/cbenko/">Chris Benko</a>, <a href="https://www.linkedin.com/in/aswin-chandrakantan/">Aswin Chandrakantan MD</a>, <a href="https://www.linkedin.com/in/rohanganesh/">Rohan Ganesh</a>, <a href="https://www.linkedin.com/in/davidbhamilton/">David Hamilton</a>, <a href="https://www.linkedin.com/in/brad-hirsch/">Brad Hirsch</a>, <a href="https://www.linkedin.com/in/elizabethiorns/">Elizabeth Iorns</a>, <a href="https://www.linkedin.com/in/shaankamal/">Shaan Kamal</a>, <a href="https://www.linkedin.com/in/kevin-obana-7aa5801b9/">Kevin Obana</a>, <a href="https://www.linkedin.com/in/ashoka-rajendra-4b1aa41b/">Ashoka Rajendra</a>, <a href="https://www.linkedin.com/in/spearlman/">Samantha Pearlman Reed</a>, <a href="https://www.linkedin.com/in/talrosenberg/">Tal Rosenberg</a>, <a href="https://www.linkedin.com/in/jay-rughani/">Jay Rughani</a>, <a href="https://www.linkedin.com/in/courtney-webster-ops/">Courtney Webster</a>, <a href="https://www.linkedin.com/in/sajithw/">Sajith Wickramasekara</a>, <a href="https://www.linkedin.com/in/johnwilbanks/">John Wilbanks</a>, and <a href="https://www.linkedin.com/in/yunancy/">Nancy Yu</a> for their thoughtful engagement with these ideas.</p> <p>The editorial refinement of this work would not have been possible without <a href="https://www.linkedin.com/in/christina-agapakis-1017b5/">Christina Agapakis</a> and <a href="https://www.linkedin.com/in/adam-conner-simons/">Adam Conner-Simons</a>, whose careful attention to detail and clarity significantly enhanced the manuscript. For their creative vision and artistic contributions, we thank <a href="https://www.linkedin.com/in/oliviareaney/">Olivia Reaney-Hall</a> and <a href="https://www.linkedin.com/in/eliza-tehan-66262bb5/">Eliza Tehan</a>, whose images and graphics brought our concepts to life, and <a href="https://www.linkedin.com/in/michela-caffrey-designer/">Michela Caffrey</a> for skillful layout design. Any shortcomings remain entirely our own.</p> <p>If you’re an early-stage founding team selling SaaS into pharma or biotech, check out this open-access resource. Our goal: help you avoid months of painful iteration so you can focus on getting life-saving innovations to market faster.</p> <p>Access the handbook: <a href="https://astera.org/pharma-sales-handbook/">https://astera.org/pharma-sales-handbook/</a></p> <p><img src="https://medium.com/_/stat?event=post.clientViewed&amp;referrerSource=full_rss&amp;postId=e3b832b8e929" width="1" height="1" alt=""/>&lt;hr&gt;&lt;p&gt;<a href="https://blog.andreacoravos.com/the-early-stage-go-to-market-handbook-e3b832b8e929">The Early Stage Go-To-Market Handbook</a> was originally published in <a href="https://blog.andreacoravos.com">Andrea’s Blog</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.&lt;/p&gt;</p>]]></content><author><name></name></author><category term="sales"/><category term="saas"/><category term="startup"/><category term="biotech"/><category term="open-access"/></entry><entry><title type="html">The Neurotech Compendium: Reflections from the Frontier</title><link href="https://andreacoravos.com/writing/2025/the-neurotech-compendium-reflections-from-the-frontier/" rel="alternate" type="text/html" title="The Neurotech Compendium: Reflections from the Frontier"/><published>2025-04-03T17:14:49+00:00</published><updated>2025-04-03T17:14:49+00:00</updated><id>https://andreacoravos.com/writing/2025/the-neurotech-compendium-reflections-from-the-frontier</id><content type="html" xml:base="https://andreacoravos.com/writing/2025/the-neurotech-compendium-reflections-from-the-frontier/"><![CDATA[<h4>Brian Takeo Smith and I are dusting off the archives from our seven-chapter Harvard neurotech IP that has only grown more relevant as our society explores the frontiers of human evolution</h4> <p>Last year, <a href="https://www.statnews.com/2024/02/29/icon-acquires-humanfirst-digital-tools-clinical-trials/">we sold HumanFirst</a>, a company I co-founded to improve human flourishing, to ICON plc (NASDAQ: ICLR). I’ve stepped away from my operational role and entered a period of sabbatical. This transition offers an opportunity to share both published and unpublished notes from our early vision — these ideas have only grown more relevant as we advance toward AGI and enhanced human capabilities.</p> <p>HumanFirst’s genesis traces back over a decade, rooted in my fascination with brain/mind connections and the potential for technology to advance human capabilities. While our commercial journey at HumanFirst focused on addressing a critical bottleneck in clinical trials — specifically, developing better evidence for endpoint measurement to de-risk and streamline the process — <em>there was always a parallel vision that remained unexplored in the company.</em></p> <figure><img alt="" src="https://cdn-images-1.medium.com/max/778/1*zdLzLSraEExaoOUwSAL3Xg.jpeg"/></figure> <p>One of my first memories exploring human potential happened freshman year at college. Every Duke freshman is required to take a class called WRITING20 — and the subtype I picked was on “Authenticity and Artifice”. At nineteen, I was captivated by a book called “<a href="https://www.amazon.com/dp/B000FCK4PY/ref=dp-kindle-redirect?_encoding=UTF8&amp;btkr=1">Radical Evolution: The Promise and Peril of Enhancing Our Minds, Our Bodies</a>” which chronicled the human enhancements that researchers like DARPA were making in hospitals and labs with our military members.</p> <p>I explored what it was like to alter limbs, change mental acuity, and other capabilities that would make our species more advanced. While writing my capstone paper, I had no idea that I was <em>already included</em> in that set of augmented humans.</p> <h3>Authenticity and Artifice: Human augmentation hidden in plain sight</h3> <p>When I returned home for spring break my freshman year, my family shared with me that I was conceived via gamete intrafallopian transfer (<a href="https://en.wikipedia.org/wiki/Gamete_intrafallopian_transfer">GIFT</a>). “You are a test tube baby. A GIFT baby,” said my mom. Until that day, I didn’t know that my very existence was a product of scientific innovation — a technological advancement that once existed beyond societal acceptance but has since become normalized.</p> <p>Let’s look at the history: the first wave of research IVF resulted in a baby born in 1978, and by the time I was born in 1987, commercial assisted reproductive tech just started to become available. My mother told me that I was one of the first people in the country born via GIFT, which was a technological procedure available at the time to be a <a href="https://www.usccb.org/committees/pro-life-activities/begotten-not-made-catholic-view-reproductive-technology">acceptable</a> intervention within the Catholic church to overcome infertility (compared to ‘<a href="https://www.usccb.org/committees/pro-life-activities/begotten-not-made-catholic-view-reproductive-technology">immoral’ IVF</a>). Notably, this conversation is much less likely to happen today, as the Overton window has shifted.</p> <p>This newfound knowledge rocked my whole world: I wasn’t a ‘naturally’ created human. From then on I fundamentally shifted the way I both relate to humans and consciousness. I realized human dignity <em>trumps</em> human authenticity. So, you could say I became an accidental pioneer of human potential from birth.</p> <p>In 2015, nine years after that fateful WRITING20 class, I met <a href="https://www.linkedin.com/in/briantakeosmith/">Brian Takeo Smith</a>, a grad school classmate who led teams overseas in the Navy’s Special Operations/Explosive Ordnance Disposal units. We teamed up for an independent project under professor <a href="https://www.hbs.edu/faculty/Pages/profile.aspx?facId=122194">Shikhar Ghosh</a> to survey a range of players working on various aspects of brain advancements and human potential.</p> <p>Our goal was to get a glimpse into the future — while sharing our learnings with the broader community. We outlined seven posts that would span across the science, funding, regulatory, and ethics of neurotech and human potential. We submitted our work for graduation, and had great ambitions to release our compendium to our broader community.</p> <p><em>… and then I let this project sit in my drafts for nearly a decade.</em></p> <p>This past weekend, I was inspired to ‘get current’ and realized I needed to release this writing out into the world. Continuing to wait was indulging my own more perfectionist tendencies. In the spirit of Jack Conte’s “<a href="https://www.youtube.com/watch?v=5r5VIlMhftM&amp;themeRefresh=1">Work to Publish</a>” — hello world; here it is!</p> <p>Brian and I took on this neurotech independent project (IP) in grad school because we believed then and it’s still clear to us now that the next few decades will be a defining moment for the human understanding of the brain and body. Humans are going to develop better ways to look at the brain — and also approaches to improve and repair it. Making such great strides in brain science is going to require many partners working together. Of course, we’ll have to understand the science, but we’ll also need to have the funding, a favorable regulatory environment, and a deep connection to humanity that allows innovation to flourish.</p> <p>As we make scientific strides, we will have to face uncomfortable ethical questions. With any great technology advancement, we must be thoughtful about building the world that we want — and not an accidental one.</p> <h3>Here’s how we structured our NeuroTech IP:</h3> <h4>Phase 1: The Scientific Journey</h4> <p>In this phase, Brian and I drafted four blog posts. The first post looked at the<a href="https://www.neurotechx.com/blog/you-can-literally-turn-on-the-lights-using-your-mindmatilda-is-not-science-fiction-anymore"> </a><a href="https://medium.com/neurotechx/you-can-literally-turn-on-the-lights-using-your-mind-matilda-is-not-science-fiction-anymore-c5851b86e6fb">technologies that are being developed today to <strong>read</strong> the brain</a>.</p> <p>The next three posts were on emerging approaches to <strong>edit</strong> and <strong>repair</strong> the brain:</p> <ol><li><a href="https://medium.com/neurotechx/can-stimulating-your-nervous-system-reverse-parkinsons-reduce-depressive-episodes-and-help-you-4699b2bb3698">Neuromodulation</a> (e.g., stimulation of the brain)</li><li><a href="https://medium.com/@takeo.smith/d28f41645162">Optogenetics</a> (e.g., controlling the brain with light)</li><li><a href="https://www.wired.com/story/prescription-video-games-and-vr-rehab/">Digital medicine</a> (e.g., software as medicine)</li></ol> <h4>Phase 2: The Funding and Regulatory Environment</h4> <p>We believed — and still believe — one of the best ways to get a glimpse into the future is to understand the current financial incentives. In 2013, the White House launched a Grand Challenge called the BRAIN Initiative. We wrote this<a href="https://medium.com/neurotechx/the-brain-initiative-the-race-to-understand-the-human-brain-c2fddfea271e"> post on an overview of the initiative,</a> and as of March 2025 the NIH shared an <a href="https://braininitiative.nih.gov/news-events/blog/brain-initiative-research-leads-fda-approval-adaptive-deep-brain-stimulation">ground-breaking announcement</a> on how the BRAIN Initiative has evolved, including research that was key to an FDA-approved personalized deep brain stimulation (DBS) product for Parkinson’s.</p> <p>In 2018, a year after we completed our IP, I decided to dive deep into the regulatory environment by serving at the FDA in the Digital Health Unit as an Entrepreneur In Residence (more on the <a href="https://rockhealth.com/create-meaningful-change-in-healthcare-serve-a-tour-of-duty-in-government/">EIR role here</a>). If you’re questioning your next career move, I highly recommend serving a ‘Tour of Duty’. Our government needs more folks like you.</p> <h4>Phase 3: Deep Dive on Neurotechnology and the US Military</h4> <p>Nuclear power, air travel, the internet, and numerous other scientific advances with great societal impact were born from military use-cases. In our write-up on <a href="https://medium.com/@takeo.smith/42ee2e16fcc3">neurotechnology initiatives and the military</a>, we hypothesized that the more permissive research/testing/funding environment of the US military will be the first place that many of these advancements will be applied.</p> <h4>Phase 4: Ethics and Our Future</h4> <p>We wrapped up our work with a <a href="https://blog.andreacoravos.com/evolution-of-the-human-brain-the-past-the-now-and-our-future-a372bbe2fc80">discussion on the ethical questions</a> that came up during this research. Our goal for this section was to surface and explore the uncomfortable questions facing our society.</p> <h3>The Full List: Our NeuroTech Ventures Compendium</h3> <ol><li><a href="https://medium.com/neurotechx/you-can-literally-turn-on-the-lights-using-your-mind-matilda-is-not-science-fiction-anymore-c5851b86e6fb">You can (literally) turn on the lights using your mind — Matilda is not science fiction anymore</a> (published in NeuroTechX, March 2017)</li><li><a href="https://medium.com/neurotechx/can-stimulating-your-nervous-system-reverse-parkinsons-reduce-depressive-episodes-and-help-you-4699b2bb3698#.tsfp2fsxs">Can stimulating your nervous system reverse Parkinson’s, reduce depressive episodes, and help you learn the piano?</a> (published in NeuroTechX, March 2017)</li><li><a href="https://www.wired.com/story/prescription-video-games-and-vr-rehab/">The Rise of Software as Medicine: Swapping out Pills for Apps, Algorithms, and Signals</a> (published in WIRED, November 2018)</li><li><a href="https://medium.com/@takeo.smith/optogenetics-controlling-the-brain-with-light-d28f41645162">Optogenetics-Controlling the Brain With Light</a> (unpublished, in conversation with NeuroTechX)</li><li><a href="https://medium.com/neurotechx/the-brain-initiative-the-race-to-understand-the-human-brain-c2fddfea271e">The BRAIN Initiative: The race to understand the human brain</a> (published in NeuroTechX, April 2017)</li><li><a href="https://medium.com/@takeo.smith/applications-neurotechnology-and-the-military-42ee2e16fcc3">Applications: Neurotechnology and the US Military</a> (unpublished, in conversation with NeuroTechX)</li><li><a href="https://blog.andreacoravos.com/evolution-of-the-human-brain-the-past-the-now-and-our-future-a372bbe2fc80">The Evolution of the Human Brain: The Past, The New, and Our Future</a> (written in 2017, published in 2025)</li></ol> <p>After we graduated, I co-founded HumanFirst, guided by a consistent vision: to use technology to expand human potential while addressing real-world healthcare challenges. HumanFirst’s flagship platform, <a href="https://www.iconplc.com/solutions/outcome-measures/digital-health-technologies">Atlas</a>, eventually cataloged over <a href="https://medium.com/elektra-labs/digital-biomarkers-landscape-21681066324b">3,000 technologies</a> and 15,000 digital measures, resulting in <a href="https://medium.com/elektra-labs/driving-towards-measures-that-matter-43cab7c4a937">25+ ontologies</a> that <a href="https://medium.com/elektra-labs/tested-tried-and-true-why-digital-biomarkers-are-the-ai-ml-strategy-pharma-should-prioritize-ad0b9f9ff367">emerged through advances in AI/ML</a> — and 24 of the top 25 pharma companies used Atlas by the time we <a href="https://www.iconplc.com/solutions/outcome-measures">merged HumanFirst into ICON’s outcome measures service</a>.</p> <p>That growth continued to accelerate: by 2024, over <a href="https://medium.com/elektra-labs/digital-endpoints-widely-adopted-in-pharmaceutical-and-biotech-sponsored-clinical-trials-c14f5d4b02e8">130 pharma and biotech</a> organizations had relied on AI-powered digital biomarkers and sensor-derived clinical outcome assessments (COAs) to capture patient-centered endpoints.</p> <p>Our focus at HumanFirst enabled our product’s success, and at the same time, we had to leave some promising early concepts unexplored. After I left HumanFirst and got a break from operating, I wanted to openly publish our early vision and flag a couple of neurotech pathways that we didn’t have the capacity to commercialize (<a href="https://www.forbes.com/sites/greglicholai/2023/06/12/joy-is-next-frontier-of-precision-medicine/">like this one</a>). My hope is that our compendium can serve as an open foundation for folks to build upon — and as an invitation to collaborate as I re-emerge post-sabbatical.</p> <p>As society stands at the cusp of transformative advances in intelligence and human experience, I’m opening these archived notes that remain remarkably prescient. At its core the AI race towards AGI, and ultimately ASI, is a race for <em>intelligence</em>, and I believe that we will need to build thoughtfully technologies for our brain/mind, specifically neurotechnologies, to integrate these AI advancements together with human ones.</p> <p>The seven articles listed above comprise our <em>The Neurotech Compendium</em> — insights from nearly a decade ago that have only grown in relevance as we continue to explore the frontier of human-machine integration and cognitive enhancement.</p> <h3>Acknowledgments</h3> <p><em>I’d like to share a big thank you to the team at </em><a href="https://neurotechx.com/"><em>NeuroTechX</em></a><em>, who collaborated and published many of our pieces. To Shikar Ghosh who was our faculty sponsor of this project and Brian (</em><a href="https://twitter.com/TakeoSmith"><em>@TakeoSmith</em></a><em>) my co-conspirator on this IP. Big thanks to Adam Conner-Simons and Dan Goodwin for reviewing the pieces, and to Jess Cole, Ginny Fahs, Rachel Katz, Linda Xie and John Weldon who inspired me to resurrect this project out of my drafts and ship it!</em></p> <p><em>To follow our work and future posts, you can find me on </em><a href="https://bsky.app/profile/arc.bsky.social"><em>Bluesky</em></a><em>, </em><a href="https://warpcast.com/coravos"><em>Farcaster</em></a><em>, X (</em><a href="https://twitter.com/andreacoravos"><em>@andreacoravos</em></a><em>) and </em><a href="https://scholar.google.com/citations?user=CeiPwnIAAAAJ&amp;hl=en"><em>Google Scholar</em></a><em>. Would love to connect with fellow researchers, entrepreneurs, and enthusiasts working at the intersection of the brain, mind, and human flourishing.</em></p> <figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*bdlRHFzU6juYASqM"/><figcaption><em>Sofia, Andy, and Brian (2017) working on prototypes in grad school that would eventually become HumanFirst.</em></figcaption></figure> <p><img src="https://medium.com/_/stat?event=post.clientViewed&amp;referrerSource=full_rss&amp;postId=8fb043c9e9d9" width="1" height="1" alt=""/>&lt;hr&gt;&lt;p&gt;<a href="https://blog.andreacoravos.com/the-neurotech-compendium-reflections-from-the-frontier-8fb043c9e9d9">The Neurotech Compendium: Reflections from the Frontier</a> was originally published in <a href="https://blog.andreacoravos.com">Andrea’s Blog</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.&lt;/p&gt;</p>]]></content><author><name></name></author><category term="mind"/><category term="ai"/><category term="brain"/><category term="evolution"/><category term="neurotech"/></entry><entry><title type="html">Evolution of the Human Brain: The Past, The Now, and Our Future</title><link href="https://andreacoravos.com/writing/2025/evolution-of-the-human-brain-the-past-the-now-and-our-future/" rel="alternate" type="text/html" title="Evolution of the Human Brain: The Past, The Now, and Our Future"/><published>2025-04-02T16:41:45+00:00</published><updated>2025-04-02T16:41:45+00:00</updated><id>https://andreacoravos.com/writing/2025/evolution-of-the-human-brain-the-past-the-now-and-our-future</id><content type="html" xml:base="https://andreacoravos.com/writing/2025/evolution-of-the-human-brain-the-past-the-now-and-our-future/"><![CDATA[<h4>Humans are outpacing their own evolution. As scientific advancements accumulate, it can be easy to retreat to the comfort of “the science” or “the data,” overlooking our agency in shaping societal and ethical choices. The path forward demands not only technical progress, and also conscious stewardship.</h4> <figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*hcR3sBbxbeovUdlJK3iFRw.jpeg"/></figure> <p><em>Note: mostly written in 2017, with light edits over the years, published in 2025 as part of my open archives project — sharing work that has only grown more relevant with time.</em></p> <p>We cannot predict what lies ahead for neurotech. It may usher in transformative breakthroughs, present serious ethical challenges, or fall short of its promised potential. What remains certain is that the opportunity to improve countless lives compels us to proceed — boldly yet thoughtfully, with equal measures of determination and responsibility as we navigate this emerging frontier.</p> <p>In 2015, Brian Harris and Owen McCarthy founded <a href="https://medrhythms.com/">MedRhythms</a> after discovering music’s potential to help people recover from neurological injuries. Brian had implemented effective music-based interventions for language, movement, and cognition recovery at Spaulding Rehabilitation Hospital. Collaborating with Owen, they realized they could transform these clinical approaches into algorithms and software.</p> <p>Their musically-derived technology had wide-ranging applications, as music could target different parts of the brain with precision. Beyond repair, it could enhance neuroplasticity and essentially rewire brain function. Using software and remote wearables, they envisioned a world where they could reach patients anywhere. After securing venture capital funding, MedRhythms was off to the races.</p> <p><em>Editor’s note in 2025: Over the past decade, the company has achieved remarkable success, including </em><a href="https://www.prnewswire.com/news-releases/medrhythms-announces-fda-class-ii-listing-for-movive-to-support-gait-rehabilitation-and-motor-function-for-parkinsons-disease-302394002.html"><em>FDA Breakthrough Device designation</em></a><em>, approval as a primary endpoint in a clinical trial, </em><a href="https://medrhythms.com/science/#publications"><em>12 peer-reviewed publications</em></a><em>, and </em><a href="https://medrhythms.com/press-release/medrhythms-announces-new-cms-reimbursement-code-for-intandem-rehabilitation-system-for-chronic-stroke-gait-impairment/"><em>Medicare approval with a dedicated reimbursement code</em></a><em>.</em></p> <p>MedRhythms is among the pioneers in what would become known as digital therapeutics (DTx), part of the global <a href="https://www.fda.gov/medical-devices/digital-health-center-excellence/software-medical-device-samd">“software as a medical device” (SaMD)</a> sector regulated by organizations like the <a href="https://www.wired.com/2017/05/medicine-going-digital-fda-racing-catch/">FDA’s Digital Health Unit</a>. These technologies are increasingly being adopted as interventions alongside pharmaceutical drugs and traditional medical devices.</p> <h3>The Past: Discovering Our Plastic Brain</h3> <p>“Plasticity” is a common term to those who follow neurotechnology, but we may take for granted how new of a concept it is. The early 20th century was dominated by the belief that brain structure was fixed after childhood development. Santiago Ramón y Cajal, despite his pioneering work on neurons, famously stated in 1928 that “<a href="https://psycnet.apa.org/record/1928-10334-000">the nerve paths are something fixed, ended, immutable</a>.” <a href="https://www.nature.com/articles/nn1100_1166#:~:text=In%201949%2C%20Donald%20Hebb%20predicted,place%20in%20the%20scientific%20pantheon.">Donald Hebb’s 1949 work on synaptic plasticity</a> (often summarized as “neurons that fire together, wire together”) laid crucial theoretical groundwork. The term “<a href="https://en.wikipedia.org/wiki/Neuroplasticity">neuroplasticity</a>” itself began appearing in scientific literature in the 1960s and 1970s.</p> <p>Now, the prevailing view in neuroscience is that our brains are far more “plastic” than we formerly believed.</p> <blockquote>Neuroplasticity is the capacity for continuous alterations of the neural pathways and synapses of the living brain and nervous system in response to experience or injury (<a href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2562625/">NIH</a>).</blockquote> <p>Neuroplasticity and the ability to alter the brain isn’t inherently good — it can produce both <em>learning</em> and <em>unlearning</em> outcomes for a human. Yes, we could use plasticity to improve measures like learning rates, but also knowledge could be lost by destabilizing a system and artificially modulating plasticity <em>[editor’s note: thanks for Dan Goodwin and Sebastian Seung for this observation]</em>. Every day, scientists are building a stronger understanding of how to control and influence those outcomes.</p> <p>In the 1960s, Paul Bach-y-Rita was one of the first to seriously study the idea of neuroplasticity when he treated patients with neurological disabilities using sensory substitution. In the 1970s, Mark Rosenzweig (UC Berkeley) and Michael Merzenich (UCSF)<a href="http://www.annualreviews.org/doi/full/10.1146/annurev.psych.47.1.1"> started to uncover that experience and activity can change the brain’s circuitry</a>. In the past 50 years, a number of scientists and researchers have made great strides in understanding how to view, image, repair, reorganize and improve the brain.</p> <h3>The Now: Transitioning from nature-directed evolution towards a human-directed one</h3> <p>The communities that research and study the brain have started to diverge. In the US, we have a mature set of institutions that focus on how to <strong>decrease the rate of brain disorders and disease</strong> in our society (e.g., NIMH). On the other side, we also have another set of institutions that are aiming to <strong>accelerate human augmentation and enhancement</strong> (e.g., DARPA, IARPA).</p> <p>Let’s start with brain disorders. The technologies that we have in our toolkit to handle<a href="https://www.ted.com/talks/thomas_insel_toward_a_new_understanding_of_mental_illness"> brain disorders</a> are becoming more robust. Neuromodulation, the ability to stimulate portions of the nervous system, has already made great advances in decreasing the rates of Parkinson’s and treatment-resistant depression. Digital therapeutics, which could scale up as wide and as fast as the internet, are on the rise. Boston-based<a href="http://akiliinteractive.com"> Akili Interactive Labs</a> is developing clinically-validated cognitive therapeutics, assessments, and diagnostics that look and feel like high-quality video games. Soon, a patient with pediatric attention-deficit/hyperactivity disorder may be prescribed non-drug treatment — an iPad with Akili’s EndeavorRx therapy — rather than a pill like Adderall or Ritalin. [<em>Note: by 2020 Akili’s EndevorRx got</em><a href="https://www.fda.gov/news-events/press-announcements/fda-permits-marketing-first-game-based-digital-therapeutic-improve-attention-function-children-adhd"><em> FDA clearance</em></a><em>.</em>]</p> <p>In 2013, the White House launched the<a href="https://medium.com/neurotechx/the-brain-initiative-the-race-to-understand-the-human-brain-c2fddfea271e"> Brain Research through Advancing Innovative Neurotechnologies (BRAIN) Initiative</a>, which straddles both communities. The goal of the initiative is to (a) understand the brain better to both (b) enhance human potential and decrease the rate of disorders.</p> <p>BRAIN Initiative funds are already accelerating technological development. Additionally, many private companies are starting to invest in enhancing human potential. In 2017, a series of high-profile companies like Bryan Johnson’s Kernel, Elon Musk’s Neuralink, and Facebook’s F8 were seeking to better harness and adapt the brain’s power.</p> <p>Although we won’t know which companies and initiatives will be successful in the long-term, having access to funds and the best research minds is already advancing the field faster.</p> <h3><strong>The Future:</strong> Let’s build a future we want; not an accidental one.</h3> <p>Let’s dive into some of the questions that we’ll likely face over the next few decades. In the past, it was likely tempting to leave the ethical discussions to<a href="https://future.a16z.com/question/future-of-expertise/#andy-coravos"> experts</a> or regulators. But technology is moving faster than our political and legislative systems can proactively handle, especially for topics that are not accessible to the general public. This puts enormous pressure on both businesses, consumers and philanthropy to decide what type of world we want to live in.</p> <p>This means that it will be important for a greater number of members in our society to form an opinion on what this future should look like. Because if you don’t, someone will take a stand on your behalf.</p> <h4>How do we define “normal” in brain science / mental health?</h4> <p>We use lots of phrases to describe the brain and most are filled with charged connotations. In our society, helping something get back to normal is considered to be a good thing, but often helping someone advance beyond normal is deemed “bad.”</p> <p>For example, if someone is diagnosed with schizophrenia, our society generally will seek to find a way to provide life-long treatment. On the flip side, we are more cautious about how we handle human advancement. We want to help their children perform well in school and score above average on tests, but there are limits to how we get there. If a child is diagnosed with ADHD, then prescriptions are OK. Without a diagnosis, is it ethical to take Adderall? What’s the boundary to take the pill or not? Learning where to draw the lines is a life-long exercise, and those lines are regularly re-drawn (related: the Overton window).</p> <p>Our current medical system for brain-related disorders often use a binary classification for the end-state diagnosis:</p> <ul><li>“Yes, you are depressed.”</li><li>“No, you do not have Alzheimer’s.”</li><li>“Yes, you have ADHD.”</li><li>“No, you do not have PTSD.”</li></ul> <p>Some of the criteria can change depending on severity (e.g., major depressive disorder, MDD)- though a diagnosis or not is binary. There are a few conditions, like Autism, that have a spectrum based classification.</p> <p>These binary descriptions are intimidating. Once someone is a “yes” are they always a yes? Does that “yes” linger with them for ages after symptoms no longer display? What if there is no biomarker?</p> <p>Perhaps we have a binary classification because it is easier to code it in a medical system, but really all of these classification seem to be sliding scales.</p> <p>It doesn’t have to be a binary classification. Thomas Insel, former NIMH Director and then at Google Verily, was one of the pioneers of the<a href="https://www.nimh.nih.gov/about/directors/thomas-insel/blog/2012/research-domain-criteria-rdoc.shtml"> Research Domain Criteria (RDoC) system</a>. RDOC is an experimental approach to the classification of mental disorders that incorporates multiple dimensions: behavior, thought patterns, neurobiological measures, and genetics. This means that every classification is more of a sliding scale than a binary “yes/no” based on limited outcomes.</p> <p>As a society, even if we have the language for the brain-related outcomes we’d like to achieve, how do we think about classifying states of the brain? “Mental health” is often used in situations where someone has a deficiency compared to a standard of “normal.” Human augmentation is for some a Silicon Valley dream and for others a scary, science-fiction nightmare.</p> <p>But what and who truly defines this “normal” state of the brain? How will it change over time?</p> <h4>Will we need to rethink “human rights”?</h4> <p>Marcello Ienca and Roberto Andorno from the <em>Institute for Biomedical Ethics at the University of Zurich, Switzerland</em> posit that we will need to rethink rights in a new paper, <a href="https://blogs.biomedcentral.com/bmcblog/2017/04/26/new-category-human-rights-neurorights/">published in<em> Life Sciences, Society and Policy</em></a>. They posit:</p> <blockquote><em>In the play </em>Comus<em>, written by John Milton in 1634, a young noblewoman is abducted by a sorcerer named Comus and bounded to an enchanted chair. Despite being restrained against her will, the woman repeatedly refuses Comus’ advances and claims </em>“Thou canst not touch the freedom of my mind”<em>, confident of her capacity to protect her mental freedom from any external manipulation. This idea of the human mind as the ultimate domain of absolute protection from external intrusion has been increasingly outdated by advances in neuroscience and neurotechnology.</em></blockquote> <p>In the piece, the autors posit that the advances in neurotechonology to image, repair, and improve the brain pose a fundamental ethical, legal, and social challenge: “<em>determining whether, or under what conditions, it is legitimate to gain access to or to interfere with another person’s neural activity.”</em></p> <p>They advocate for a “reconceptualization of existing human rights” and, perhaps, a creation of “new human rights” that they call “neurorights”:</p> <blockquote><strong>Neurorights</strong>: the right to cognitive liberty, the right to mental privacy, the right to mental integrity, and the right to psychological continuity.</blockquote> <p><em>Cognitive liberty</em> is not something we can take for granted anymore. For example, with the rise of deep brain stimulation, we’ve been able to<a href="https://medium.com/neurotechx/can-stimulating-your-nervous-system-reverse-parkinsons-reduce-depressive-episodes-and-help-you-4699b2bb3698"> reverse the symptoms of Parkinson’s</a> by altering the circuits of the brain. Or MDMA-assisted therapy for <a href="https://www.nature.com/articles/s41591-023-02565-4">moderate to severe PTSD</a>. But those circuits will also be accessible to malevolent actors.</p> <p><em>Mental privacy </em>aims to “protect individuals against unconsented intrusion by third parties.” The authors note that Steven Spielberg’s <em>Minority Report</em> scenario has become an increasingly real risk. In April 2017, Facebook was<a href="http://mashable.com/2017/04/19/facebook-brain-interface/"> bragging at F8 about it’s brain-computer-interface research</a> and the ability to type directly from the brain. The lines between when we are plugged “in” versus not will be increasingly blurred.</p> <p>The <em>right to mental integrity</em> is already a recognized international law (Article 3 of the EU’s Charter of Fundamental Rights). The objective is to protect people against threats like “unwanted neurostimulation, malicious neurohacking and potentially harmful memory manipulation.”</p> <p>The right to <a href="https://www.britannica.com/topic/psychological-continuity">psychological continuity</a> is a right that has been discussed more recently with the rise of cognitively manipulative digital content. It includes suggestions about how to deal with:</p> <blockquote>invasive marketing strategies such as those in which advertising is purposively designed to bypass a person’s rational defenses and alter their preferences or behavior</blockquote> <p>With the rise of any great technology, there is both good and bad. The benefits for neurotechnology for our society is enormous, and we’ll want to have the legal and ethical frameworks to guide its development.</p> <h4>Who is responsible for guiding this technological advancement?</h4> <p>Sometimes it’s obvious when we introduce something into the world that it will shift the paradigm (e.g., nuclear bombs or the internet), and with these shifts they were both often funded by governments and guided by policy (e.g., with nuclear came nuclear pacts). With the internet, the power of nation states has been changing, and that means other organizations are picking up the slack — take for example the rise of AI related tech which is largely privately funded rather than publicly developed.</p> <p>Our government has been historically slow to think through and regulate emerging technologies, and independent citizens have come together to fill the gap. For example,<a href="https://openai.com/"> OpenAI</a> is a non-profit artificial intelligence research company that aims to carefully promote and develop friendly AI in such a way as to benefit humanity as a whole (<em>editor’s note: written in 2017</em>). OpenAI is supported by over US $1B in commitments by founders who are in part motivated by concerns about existential risk from AI intelligence.</p> <p>As the power of the democratic nation state changes, the power of tech organizations rises and these non-governmental organizations are increasingly setting standards. But these organizations do not have any obligation to “the people” — <a href="https://www.wired.com/insights/2015/01/google-and-elon-musk-good-for-humanity/">is this a world we want to live in</a>?</p> <p>We are already seeing this shift happen in the autonomous car industry, where Google, Apple, Uber among others are setting the standards and ethics around life-and-death decisions for the car and its surroundings. How do we get in front of this for brain advancements?</p> <h4>Is all technological advancement good?</h4> <p>The duality of technology has been a part of human history for ages. With fire, we could warm villages and burn down cities. Nuclear power can provide electricity for billions and destroy nations.</p> <p>Phillip Rogaway, Department of Computer Science, at UC Davis, wrote about technological optimism in his piece “<a href="http://web.cs.ucdavis.edu/~rogaway/papers/moral.html">The Moral Character of Cryptographic Work</a>”:</p> <blockquote>Technological optimists believe that technology makes life better. According to this view, we live longer, have more freedom, enjoy more leisure. Technology enriches us with artifacts, knowledge, and potential. Coupled with capitalism, technology has become this extraordinary tool for human development. At this point, it is central to mankind’s mission. While technology does bring some unintended consequences, innovation itself will see us through.</blockquote> <blockquote>Technological optimism animates everyone from school children to Turing Award winners. Accepting his 2012 Turing Award, Silvio Micali, said that</blockquote> <blockquote>Computer science is marking an epical change in human history. We are conquering a new and vast scientific continent. . . . Virtually all areas of human activity . . . [and] virtually all areas of human knowledge . . . are benefiting from our conceptual and technical contributions. . . . Long live computer science!</blockquote> <blockquote>If you’re a technological optimist, a rosy future flows from the wellspring of your work. This implies a limitation on ethical responsibility. The important thing is to do the work, and do it well. This even becomes a moral imperative, as the work itself is your social contribution.</blockquote> <blockquote>But what if computer science is not benefiting man?</blockquote> <p>Our brain-related advances have the same conundrum. What are the technologies that we want to introduce into our world, and which ones do we want to avoid?</p> <ul><li>There’s the promise that neuroscience could wipe out negative emotions, but should it?</li><li>We have more “evolved” humans — where is the line between being “authentically” human and does it matter?</li><li>And perhaps a more haunting question that we don’t know an answer to yet: <em>can software suffer</em>?</li></ul> <p>Many of the advancements are inevitabilities. If there is an incentive, it will happen. Change in policies and incentives can unleash massive changes in society (think Orphan Drug Act of 1983 and the impact on rare disease research). It will be useful to map out the likely inevitabilities for neurotech that we can identify today, and then really work through the incentives.</p> <p>There is real danger in passively sitting back and watching our future shape us. We will have a much better society if we actively work through these questions, form opinions, build organizations (like the non-profit OpenAI intended to be and has since evolved), and hold our governments and institutions accountable to shaping our future.</p> <p>Let’s lean in and build an <em>intentional future</em>; not an accidental one.</p> <p><em>Acknowledgements: Big thanks to </em><a href="mailto:Takeo.Smith@gmail.com"><em>Brian Smith</em></a><em> for kicking off this NeuroTech project with me many years ago, and for reviewing this post. To my editor Adam Conner-Simons, </em><a href="mailto:owenzmccarthy@gmail.com"><em>Owen McCarthy</em></a><em> for the MedRhythms example, </em><a href="mailto:greg@leppert.me"><em>Greg Leppert</em></a><em> and Dan Goodwin for reviewing very early drafts, and to my friends </em><a href="https://twitter.com/jessunscripted"><em>Jess Cole</em></a><em>, Ginny Fahs, Rachel Katz, Linda Xie, and John Weldon who inspired me to resurrect this project out of my drafts and ship it!</em></p> <p><img src="https://medium.com/_/stat?event=post.clientViewed&amp;referrerSource=full_rss&amp;postId=a372bbe2fc80" width="1" height="1" alt=""/>&lt;hr&gt;&lt;p&gt;<a href="https://blog.andreacoravos.com/evolution-of-the-human-brain-the-past-the-now-and-our-future-a372bbe2fc80">Evolution of the Human Brain: The Past, The Now, and Our Future</a> was originally published in <a href="https://blog.andreacoravos.com">Andrea’s Blog</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.&lt;/p&gt;</p>]]></content><author><name></name></author><category term="brain"/><category term="neuroplasticity"/><category term="ai"/><category term="ethics"/><category term="neuroscience"/></entry><entry><title type="html">Decentralized clinical trials</title><link href="https://andreacoravos.com/writing/2018/decentralized-clinical-trials/" rel="alternate" type="text/html" title="Decentralized clinical trials"/><published>2018-10-15T11:48:33+00:00</published><updated>2018-10-15T11:48:33+00:00</updated><id>https://andreacoravos.com/writing/2018/decentralized-clinical-trials</id><content type="html" xml:base="https://andreacoravos.com/writing/2018/decentralized-clinical-trials/"><![CDATA[<h4>A framework for “direct-to-patient,” “remote,” “site-less,” and “virtual” clinical research</h4> <p><em>A year ago, the </em><a href="https://www.elektralabs.org/"><em>Elektra Labs</em></a><em> team published a piece on “</em><a href="https://blog.andreacoravos.com/software-enabled-clinical-trials-8da53f4cd271"><em>Software-Enabled Clinical Trials</em></a><em>” which mapped software tools that support clinical research. Since then, a number of new clinical trial approaches have emerged, and in the following piece we’ll decrypt some of the buzzwords from “site-less” to “direct-to-patient” to “virtual” and beyond.</em></p> <p>We depend on clinical trials to bring safe and effective drugs and devices to market.</p> <p>However, many of our life-altering drugs and devices are developed on limited populations in strict laboratory-controlled settings, which may not match natural settings. Over <a href="https://blogs.deloitte.com/centerforhealthsolutions/digital-rd-four-ways-to-maximize-patient-engagement-in-clinical-trials/">70 percent of potential participants live more than 2 hours away</a> from the nearest study center — traveling to a physical clinical trial site is a burdensome, limiting engagement. Today, less than 5% of the US population participates in clinical research.</p> <p>Additionally, the data collected in clinical trials — the information we hang multi-billion dollar investments on and that patients are literally living and dying by — is such a tiny snapshot of the lived experience with a disease.</p> <p>Decentralized clinical trials (DCTs), which are conducted in a study participant’s home using digital tools, offer a way to make better informed decisions about the efficacy of new therapies. More sensitive, objective measures from digital technologies coupled with a greater density of information — continuously sampling multiple times a day, not just one a quarter — will help us fail faster and win cheaper (h/t <a href="https://medium.com/u/ce0ceef9fcb9">Jennifer Goldsack</a>).</p> <p>“Being healthy or sick isn’t a one-shot deal; it’s a continuous process that spans a lifetime,” <a href="https://www.wsj.com/articles/open-your-app-and-say-ahh-1537397802?emailToken=00f92c80ee3b359cdfc9478ee93b180fiqebRRTmTg0KSvOZz+etPncsnQQuxl0Vd2q/AZYE6swzRk0e6lOZJEFIwXrF3A2PJEP0isToJKr7jjxzSvGnJg%3D%3D&amp;reflink=article_copyURL_share">writes Sean Khozin, Director of the FDA INFORMED incubator in the WSJ</a>, and the data from these digital tools may be “more valuable” than the data from traditional clinical trials, which are collected at specialized medical centers rather than in a patient’s natural environment.</p> <p>Well-structured DCTs would use fewer resources to bring higher quality drugs and devices to the market at a faster rate. Our goal for this post is to provide (a) frameworks for evaluating DCTs and (b) definitions of supporting technologies — a lack of clarity and inconsistencies in language surrounding decentralized trials hinders scientific advancement.</p> <p>In taking a cue from the FDA/NIH:</p> <blockquote>Effective, unambiguous communication is essential for efficient translation of promising scientific discoveries into approved medical products. <strong>Unclear definitions and inconsistent use of key terms can hinder the evaluation and interpretation of scientific evidence</strong> and may pose significant obstacles to medical product development programs. (<a href="https://www.ncbi.nlm.nih.gov/books/NBK326791/">BEST Framework</a>)</blockquote> <p>Below is our first draft of frameworks and definitions for the community. If you have suggestions, send us a note on Twitter (<a href="https://twitter.com/elektralabs">@ElektraLabs</a>) or leave a comment.</p> <h3>What is a decentralized clinical trial (DCT)?</h3> <p><em>Note for crypto readers — “decentralized” as defined in this post is </em><strong><em>not</em></strong><em> blockchain-related. If you’re interested in healthcare blockchains, check out the </em><a href="http://biomedicalblockchain.org/"><em>Mt. Sinai’s Center for Biomedical Blockchain Research</em></a><em> (CBBR) </em><a href="https://www.statnews.com/2018/07/25/blockchains-biomedicine-health-care-buyer-be-informed/"><em>Op-Ed in STAT News</em></a><em> and this </em><a href="https://db.biomedicalblockchain.org/"><em>crowdsourced biomedical blockchain project list</em></a><em>.</em></p> <figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3imF28x-dHkPZPkpN0xEAw.png"/></figure> <h4>Two components: “where” and “how” patient data are collected</h4> <p>We proposed that two features of data collection that determine how “decentralized” a clinical trial is:</p> <p>(1) <strong><em>Where</em> are data collected?</strong> (e.g., how dependent is the trial on a “site” such as a clinic or hospital center?)</p> <p>Clinical trial sites serve both patients and sponsors (e.g., the biopharma or device manufacturer). Historically, the <strong>site</strong> is where drugs, medical devices and other therapies are tested on humans — often a clinic or research hospital. A new model has emerged called “<strong>direct-to-patient</strong>” or “<strong>remote</strong>” where patient data is collected in the home or in the study participant’s natural environment.</p> <p>This shift has recently accelerated as more digital tools are able to collect clinically-validated data (e.g., digital biomarkers and electronic outcomes assessments — eCOAs) and have improved operational capabilities (e.g., the ability to perform blood draws at home with mobile phlebotomists). Some clinical trials employ a <strong>hybrid </strong>approach<strong>, </strong>with some sites and some remote locations. <a href="https://twitter.com/CraigLipset?ref_src=twsrc%5Egoogle%7Ctwcamp%5Eserp%7Ctwgr%5Eauthor">Craig Lipset</a> at Pfizer refers to these as <em>“location-flexible</em>” trials. These types of trials can maximize patient-centricity by allowing the study participant more flexibility in how she or he participates in the research.</p> <p>As with any new technology, the industry is still figuring out the best language to describe emerging products and services. Some people refer to remote clinical as <a href="https://www.clinicaltherapeutics.com/article/S0149-2918(17)30200-X/fulltext">“<strong>site-less</strong>” clinical trials</a>, though this phrase is a misnomer. Even <a href="https://www.science37.com/">Science 37</a>, which specializes in this this type of trial, has a site: <em>itself. </em>Science37 <strong><em>is</em></strong> the site —<em> </em><a href="https://www.science37.com/">the “Metasite</a>”<em> </em>— which provides support and coordination for the study participants.</p> <p>(2) <strong><em>How</em> are the data collected?</strong> (e.g., does collection require an intermediary like a study team or phlebotomist to collect the data?)</p> <p>In the past, most clinical trial data were collected <em>via an intermediary</em> — someone from the study team would record the information and then record it in the EDC and/or case report form.</p> <p>The data could be used to test the waters and monitor a new study site or participant to inform future protocol design, or the data could be used to provide evidence to support the study endpoint. A clinical trial <strong>endpoint</strong> refers to “an event or outcome that can be measured objectively to determine whether the intervention being studied is beneficial” (NIH).</p> <p>As digital tools advance, we can collect more endpoint-supporting data at home via digital surveys and sensors, and study teams can “visit” patients at home via telemedicine conference calls. This means that more of the data is participant/<strong>patient-generated</strong>.</p> <p>The word “virtual” in clinical trials has a loose meaning — sometimes used to refer to “remote” studies as described above. In this case, we define <strong>fully virtual</strong> as the method for collecting clinically-validated patient data <em>without an intermediary</em>, and all the data are fully participant-generated.</p> <h4>Why are decentralized clinical trials attractive?</h4> <p>DCTs have a long list of benefits, many of which were published last week in <a href="https://www.statnews.com/2018/09/21/clinical-trials-reinvention-janssen/">STAT News First Opinion</a> by <a href="https://twitter.com/andreas_koe?ref_src=twsrc%5Egoogle%7Ctwcamp%5Eserp%7Ctwgr%5Eauthor">Andreas Koester</a>, M.D., the global head of Janssen Clinical Innovation. DCTs can recruit more people into trials, increase retention and engagement, collect more continuous data in natural settings, while shortening the study time (faster to market), and decreasing costs.</p> <p>Additionally, software-driven tools like <em>telemedicine</em>, <em>personalized and clinically validated digital measures</em> from wearable sensors (e.g., digital biomarkers) and surveys (eCOA), and <em>improved logistics and operations management</em> (e.g., coordinating phlebotomists at home, shipping drugs to the home) have accelerated the feasibility and adoption of DCTs in recent years.</p> <h4><strong>Should all studies be fully decentralized?</strong></h4> <p>No — “fully decentralized” does <strong>not</strong> make sense for all types of research. Some studies may require an MRI or CT scan, and these technologies are expensive and are maintained at a clinical trial site. Also, the study protocol may require a blood draw, which is challenging to do at home with personalized tools (RIP Theranos).</p> <figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gs6ruxnLfhvilzavLmmByw.png"/></figure> <p><em>The best study protocols mix and match centralized and decentralized tools based on study objectives. </em>For instance, observational post-market studies, often called Phase IV studies, are great candidates for a fully virtual approach, which is able to capture more continuous data in natural environments. In contrast, oncology studies will likely require more of a hybrid approach with more expensive site-based equipment.</p> <p>We asked <a href="https://twitter.com/CraigLipset?ref_src=twsrc%5Egoogle%7Ctwcamp%5Eserp%7Ctwgr%5Eauthor">Craig Lipset from Pfizer</a> about his thoughts and he replied that “a study that is truly ‘patient-centric’ would allow the patient to choose their own adventure.” If, for example, the patient is capable and wants to supply their own blood sample, then the protocol could be set up without an intermediary (more decentralized). In contrast, many people we asked preferred to have a trained phlebotomist draw their blood (less decentralized) rather than using a home-kit.</p> <h4><strong>Which organizations are supporting decentralized trials?</strong></h4> <p>Pfizer’s 2011 <a href="http://blogs.nature.com/news/2011/06/virtual_clinical_trial_could_h_1.html">REMOTE (Research On Electronic Monitoring of Overactive Bladder Treatment Experience) study</a> is widely referenced as the first clinical trial of an FDA-approved pharmaceutical, which used digital technologies to recruit and manage participants entirely from their homes. This study also pioneered one of the first versions of electronic informed consent (eConsent).</p> <p>Since then, a number of contract research organizations (CROs) and specialty vendors have developed products across the decentralized trial spectrum, many of which we outlined in our <a href="https://blog.andreacoravos.com/software-enabled-clinical-trials-8da53f4cd271">blog post last year</a> on software-enabled clinical trials.</p> <figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-_UBSxUo2Y-PRKXW-jU5RQ.png"/></figure> <p>Biopharma and medical device companies (“sponsors”) commonly work with a number of third party vendors during product development (e.g., contract research organizations / CROs). Some companies are more “full-service” in their approach, and will support the sponsor to design the protocol, recruit patients, set up and manage the site, and capture the patient data.</p> <p>Other companies focus in a single area, like high-quality data capture, because many digital tools for clinical trials have features that require special attention (e.g., continuous data capture from a wearable is harder to collect, store, analyze and validate than a blood test taken 1x/month for a year). With specialization, it’s more likely that generalist CROs will have to partner with a few “speciality” vendors — companies like Koneksa and Evidation — to successfully bring a product to market.</p> <p>The software-enabled clinical trials landscape is moving quickly, and as organizations develop deeper product offerings a one-page landscape map is limiting, because duplicating logos gets cluttered.</p> <figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9w6Qb-6uL9lAKlJp6cddpg.png"/><figcaption>Our landscape map updated September 2018. To see the most recent version, check out <a href="http://www.elektralabs.org/decentralized-trials">Elektra Labs</a>.</figcaption></figure> <p>To update the map over the past year, we asked organizations to add comments to our Medium post or send us an email with product update — two methods that are not scalable. So, we followed the lead of all the organizations on this map, and created a web-based version of the landscape, which we launched at DPharm this month. Check out the <a href="http://www.elektralabs.org/decentralized-trials">landscape map on the Elektra Labs website</a> to add and update company information.</p> <h4>Related Initiatives</h4> <p>Overtime, we hope to collaborate with other catalogs like <a href="http://edcmarket.appspot.com/about">eClinicalHealth’s EDCMarket</a> to create a more holistic view of company offerings to support software-enabled clinical trials. EDC Market’s data is becoming out of date and the team is open to finding partners to revive and expand the database.</p> <p>The Clinical Trial Transformation Initiative (CTTI), a public-private partnership co-founded in 2007 by Duke University and the US FDA, is currently working on a project <a href="https://www.ctti-clinicaltrials.org/projects/decentralized-clinical-trials">exploring the challenges of conducting decentralized clinical trials (DCTs) through telemedicine and mobile healthcare providers</a>.</p> <p>CTTI is digging into several topics including “telemedicine state licensing issues, FDA review division reception, and Good Clinical Practice-related issues — to help <a href="https://www.ctti-clinicaltrials.org/projects/mobile-technologies">advance widespread use of mobile technologies</a> in DCTs.” CTTI <a href="https://www.ctti-clinicaltrials.org/projects/decentralized-clinical-trials">released their work</a> this fall at <a href="http://theconferenceforum.org/conferences/disruptive-innovations-us/overview/">DPharm</a> in Boston last month.</p> <h4><strong>What’s next?</strong></h4> <p>We believe that many clinically-validated endpoints used in research will eventually transition into clinical care. Historically, measures that supported research (“endpoints”) and care (“outcomes”) were siloed, but we are seeing examples of these measurements merging. Many companies are working towards a universal vision of human digital measurement across the continuum of research and clinical care.</p> <figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*z8IDjbjJHHg5ID3f4_yglg.png"/></figure> <p>For example, <a href="https://evidation.com/">Evidation Health</a> views itself as new kind of health and measurement company that focuses on quantifying the relationship between everyday behaviors and health outcomes. They don’t see themselves as a CRO / clinical trials company, per se, though Evidation does a lot of clinical studies in their enterprise B2B business.</p> <p>Clinical research provides a practical use-case to link (or validate) everyday behaviors and outcomes. For instance, many companies like Evidation, Sage Bionetworks, and Medable that are developing digital biomarkers, “consumer-generated physiological and behavioral measures collected through connected digital tools” (<a href="https://rockhealth.com/reports/the-emerging-influence-of-digital-biomarkers-on-healthcare/">Rock Health</a>), validate these tools through clinical research <em>as a first step</em> towards what may eventually become a validated digital diagnostic (e.g., software that diagnosis) or a digital therapeutic (e.g., software that treats/intervenes).</p> <p>Similarly, a number of big tech companies are also developing digital measures for more clinical settings. Last week, the FDA <a href="https://www.accessdata.fda.gov/cdrh_docs/pdf18/DEN180044.pdf">cleared a “software as a medical device” (SaMD) for the Apple Watch that determines the presence of atrial fibrillation</a>, an abnormal heart condition. This clearance provides a regulatory pathway for Apple to create more advanced diagnostics and interventions for the patient, at home, decentralized.</p> <p>It’s in the early days for SaMDs and digital measures like biomarkers. We will need better tools to compare and select validated measures for both research and care. Our next post will dig into more “digital measures,” including sensors, digital biomarkers and electronic clinical outcome assessments (eCOAs). <em>Stay tuned.</em></p> <p><em>This work was supported by the Harvard-MIT Center for Regulatory Science. Thanks to </em><a href="https://medium.com/u/93854b681160"><em>Joe Dustin</em></a><em>, </em><a href="https://medium.com/u/3dc83f1de691"><em>Craig Lipset</em></a><em>, Chris Benko, </em><a href="https://medium.com/u/ce0ceef9fcb9"><em>Jennifer Goldsack</em></a><em>, Bill Wood, Deborah Kilpatrick, Elena Izmailova, Yusuf Ghadiali, </em><a href="https://medium.com/u/49aba4e8c9c2"><em>Franklin Yang</em></a><em>, </em><a href="https://medium.com/u/625d689f7bd0"><em>Ankit Gordhandas</em></a><em>, </em><a href="https://medium.com/u/926fce0ce255"><em>Sofia Warner</em></a><em>, </em><a href="https://medium.com/u/5d54086f1e76"><em>Michelle Longmire</em></a><em>, Marni Hall, Sean Khozin, </em><a href="https://medium.com/u/1b076477fed"><em>Adam Goulburn</em></a><em>, and many others for contributing to this piece.</em></p> <p>An aside on “<strong>real-world,</strong>” a buzzword we avoided in this piece but is important to define. The regulatory definition of real-world data is <a href="https://www.fda.gov/scienceresearch/specialtopics/realworldevidence/default.htm">the data collected outside of a randomized control trial (RCT)</a>. <em>Real-world </em><strong><em>evidence</em></strong><em> </em>is the evidence derived from <a href="https://www.sciencedirect.com/science/article/pii/S1098301517301717"><em>real-world </em><strong><em>data</em></strong></a>. If study participants contribute to some measurements at home, e.g. pain measurement via ePRO or step count wearing a sensor, in the context of an RCT, they do <em>not constitute real-world data</em> because the participants have been preselected for study entry by the inclusion and exclusion criteria of a given trial — and do not represent the overall population in a certain indication. This is why in on this piece on DCTs, we talked about the benefits of <em>health-related data collected in natural settings — </em>and not “real world” settings. In the context of drug development, the “real world data” a biopharma collects in <em>post-marketing</em> studies is not restricted by the inclusion/exclusion criteria. <em>(Adapted from a conversation with Elena Izmailova at Takeda)</em></p> <p><img src="https://medium.com/_/stat?event=post.clientViewed&amp;referrerSource=full_rss&amp;postId=e9dbde90ea95" width="1" height="1" alt=""/>&lt;hr&gt;&lt;p&gt;<a href="https://blog.andreacoravos.com/decentralized-clinical-trials-e9dbde90ea95">Decentralized clinical trials</a> was originally published in <a href="https://blog.andreacoravos.com">Andrea’s Blog</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.&lt;/p&gt;</p>]]></content><author><name></name></author><category term="biometrics"/><category term="clinical-trials"/><category term="digital-health"/><category term="health"/><category term="pharmaceutical"/></entry><entry><title type="html">Twitter’s design flaw that stunts healthy, civil, productive conversations</title><link href="https://andreacoravos.com/writing/2018/twitters-design-flaw-that-stunts-healthy-civil-productive-conversations/" rel="alternate" type="text/html" title="Twitter’s design flaw that stunts healthy, civil, productive conversations"/><published>2018-03-02T19:25:16+00:00</published><updated>2018-03-02T19:25:16+00:00</updated><id>https://andreacoravos.com/writing/2018/twitters-design-flaw-that-stunts-healthy-civil-productive-conversations</id><content type="html" xml:base="https://andreacoravos.com/writing/2018/twitters-design-flaw-that-stunts-healthy-civil-productive-conversations/"><![CDATA[<p><em>Last month, the Twitterverse stumbled upon an all-too-common situation that is worth discussing in more than 280 characters. Deleting a Tweet is the deficient but only available option to stop the viral spread of incomplete information as new context arises. Twitter has a product design flaw that stunts productive conversations, and it’s time to lobby Twitter to make a change.</em></p> <figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Qa-7T1rRp-IgxjdBMrcV0w.png"/></figure> <p><em>This is Part 2 in the series. In </em><a href="https://medium.com/andreas-blog/going-viral-new-responsibilities-in-the-attention-economy-d0e97a4b09f1"><em>Part 1</em></a><em>, we looked at the ‘side effects’ of an attention economy.</em></p> <p>Last month, I posted a Tweet that went viral and garnered 1.7M+ impressions (see <a href="https://medium.com/andreas-blog/going-viral-new-responsibilities-in-the-attention-economy-d0e97a4b09f1">Part 1</a> for details). As more people saw the tweet, I learned additional context — some confirming and some denying elements of the Tweet.</p> <p>The following two days were a frustrating experience, because I was not able to add any context into the conversation. Only the original Tweet goes viral.</p> <p>Twitter has a huge flaw in its product: the author of a post has only one suboptimal way to slow down the spread of information — <em>delete the Tweet</em>. This design flaw hinders our ability to engage in an open, civil, and public conversation around the world.</p> <p>I believe we need to be responsible for the things we say/put out into the world, and share how we’ve shifted our thinking over time as more information becomes available. I don’t want to create a culture where someone can say “shithole,” <a href="https://www.washingtonpost.com/politics/trump-acknowledges-tough-language-but-appears-to-deny-shithole-remark/2018/01/12/c7131dae-f796-11e7-beb6-c8d48830c54d_story.html">deny it</a>, and obfuscate truth. Deleting a Tweet obfuscates truth, and we lose the conversation and clarifications.</p> <p>Last month, some users crafted shame-inducing arguments stating that I kept the Tweet for popularity reasons.</p> <h3>Taylor Lorenz on Twitter</h3> <p>I don&#39;t believe in deleting tweets when I can continue to rake in the RTs by spreading false information.&quot; https://t.co/mCVftDlbn8</p> <p>I can feel the outrage. <strong>When something is broken, it is easy to shift blame to the individual instead of fighting for a system change </strong>(see <a href="https://en.wikipedia.org/wiki/Fundamental_attribution_error">fundamental attribution error</a>)<strong>. </strong>Ad hominem attacks are effortless; shipping paradigm-shifting product changes takes work.</p> <p>There are many product updates Twitter could ship to give power back to the individual and slow down the flow of information while a Tweet’s veracity is under review:</p> <p>(1) <strong>Allow the author of the Tweet to turn-off retweets</strong>.</p> <p>(2) <strong>Allow the author of the Tweet to cluster multiple Tweets. </strong>If a user wants to retweet the original post, s/he would have to retweet a bundle of Tweets. If this option were available, I would have added the Instagram CTO and Dopamine Labs clarifications. If Twitter is concerned about this feature clogging the feed, it could turn on after a certain level of virality.</p> <p>(3) <strong>Give the author the ability to “flag” a Tweet.</strong> This could be a general flag or spectific — perhaps “under review” or “read the full thread.”</p> <p>(4) <strong>Give the community the ability to “flag” a Tweet. </strong>What if I went away for the weekend and couldn’t append updates? If members in the community see an error, allow them to flag — or downvote.</p> <p>(5) <strong>Give the author the ability to prompt the user with a message before re-tweeting </strong>(e.g., “read the full thread before retweeting” or “new facts have emerged, retweet with caution”)</p> <p><em>I would have used any of these features last month, but they do not exist yet in the app.</em></p> <p>(6) While I don’t strongly support this option, Twitter could give the author the ability to <a href="https://www.theverge.com/2017/4/13/15278800/edit-tweets-twitter-hysteria-nihilism"><strong>edit the Tweet</strong></a><strong> </strong>(HT <a href="https://medium.com/u/b6e05447f2cf">Casey Newton</a>). If this option is selected, the Tweet should have an audit trail similar to Facebook post updates that include the history.</p> <p>If we all agree that fake news erodes trust in our society, <strong>deleting a Tweet</strong> is the worst option we have to stop the viral spread of information. We lose the ability to link back to the offending message, and the conversation and context becomes challenging to reconstruct.</p> <p>The criticism that struck me the most was from <a href="https://medium.com/u/78c5166b8328">Benedict Evans</a>, a senior partner at <a href="https://medium.com/u/df45fd4a749d">Andreessen Horowitz</a> (a16z), who posted <a href="https://twitter.com/BenedictEvans/status/952602590566432768">multiple</a> <a href="https://twitter.com/BenedictEvans/status/952600744007385088">Tweets</a> over the weekend pushing me to delete the post. Andreessen Horowitz is a VC firm well-known in the Valley for investing millions in crypto/blockchain technologies, which I also strongly support. The a16z team has made multiple arguments about how blockchains are powerful sources of truth because they have immutable, time-stamped records, removing the risk of having a rogue central system administrator <em>edit, modify, delete, or alter </em>records. In a blockchain-based future, it’s not possible to delete.</p> <p>In the Internet we have today, once something is said publicly, copied, and shared to multiple sources, the high fidelity option we have is to move forward and to append.</p> <p><em>As an individual</em>, I’m responsible for what I write publicly and for updating my position as more information becomes available.</p> <p><em>As a system platform</em>, Twitter, is responsible for providing tools that stop the spread of misleading content while preserving an audit trial so the community can discern truth.</p> <p>The platforms we use to communicate impact how information spreads and have consequences on elections, politics, society — and they have meaningful, visceral impacts on our individual behaviors and emotions. The platforms have tried to dodge responsibility by claiming a “neutral” platform, but product decisions impact the accessability of information, audit trails, and accuracy.</p> <p>Yesterday, Twitter’s CEO <a href="https://medium.com/u/ab69c5472679">Jack Dorsey</a> acknowledged that the platform needs to optimize for different metrics and released a <a href="https://blog.twitter.com/official/en_us/topics/company/2018/twitter-health-metrics-proposal-submission.html">request for proposal</a> to “increase the collective health, openness, and civility of public conversation.” This is a landmark moment for the company as it asks the public for ways to create a better system.</p> <p>If we are serious about <a href="https://www.wired.com/story/free-speech-issue-tech-turmoil-new-censorship/">“fixing” social media</a>, it’s time to step <a href="http://www.theodore-roosevelt.com/trsorbonnespeech.html">into the arena</a>, update the system, and fix product design failures. If you are committed to quickly verifying the truth and stopping the flow of misleading content, if you recognize that more facts will emerge over time, <a href="https://blog.twitter.com/official/en_us/topics/company/2018/twitter-health-metrics-proposal-submission.html">submit a proposal to Twitter</a> for how to ship a better product.</p> <p>I’ve submitted mine.</p> <p>#ForAHealthierTwitter</p> <p><em>Many thanks to </em><a href="https://medium.com/u/cd336ad4fe83"><em>Dan Romero</em></a><em>, </em><a href="https://medium.com/u/1d4fe842a881"><em>Diana Kimball Berlin</em></a><em>, </em><a href="https://medium.com/u/306b74344e83"><em>Renee DiResta</em></a><em>, </em><a href="https://medium.com/u/6053392f2d42"><em>Ellen Chisa</em></a><em>, </em><a href="https://medium.com/u/6a54160e13ba"><em>Seth Rosenberg</em></a><em>, </em><a href="https://medium.com/u/175f6ddf9dd9"><em>Yasmin Razavi</em></a><em>, </em><a href="https://medium.com/u/b044c3f9ff74"><em>veronicaosinski</em></a><em>, Addison Godine, </em><a href="https://medium.com/u/b947efe0a51a"><em>Fred Ehrsam</em></a><em>, </em><a href="https://medium.com/u/3b6ad741de83"><em>Caitlin Strandberg</em></a><em>, and three [redacted] others who talked with me about the Tweet and read through this series to provide feedback/ideas for how we can make our digital worlds better.</em></p> <p><img src="https://medium.com/_/stat?event=post.clientViewed&amp;referrerSource=full_rss&amp;postId=73628ecaa736" width="1" height="1" alt=""/>&lt;hr&gt;&lt;p&gt;<a href="https://blog.andreacoravos.com/twitters-design-flaw-that-stunts-healthy-civil-productive-conversations-73628ecaa736">Twitter’s design flaw that stunts healthy, civil, productive conversations</a> was originally published in <a href="https://blog.andreacoravos.com">Andrea’s Blog</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.&lt;/p&gt;</p>]]></content><author><name></name></author><category term="product-design"/><category term="social-media"/><category term="blockchain"/><category term="tech"/><category term="twitter"/></entry><entry><title type="html">Going Viral: New Responsibilities in The Attention Economy</title><link href="https://andreacoravos.com/writing/2018/going-viral-new-responsibilities-in-the-attention-economy/" rel="alternate" type="text/html" title="Going Viral: New Responsibilities in The Attention Economy"/><published>2018-03-02T19:24:56+00:00</published><updated>2018-03-02T19:24:56+00:00</updated><id>https://andreacoravos.com/writing/2018/going-viral-new-responsibilities-in-the-attention-economy</id><content type="html" xml:base="https://andreacoravos.com/writing/2018/going-viral-new-responsibilities-in-the-attention-economy/"><![CDATA[<p><em>In the arms-race for attention, our online communication platforms employ nudges, notifications, and other tools to change our behavior and increase our “engagement.” Like drugs, algorithms also have (unintended) side-effects — stoking human anger, outrage and insecurities. In response, a new movement is digging into how to realign technology to serve humanity’s best interests.</em></p> <p>Last month, I posted a quote and a question about Instagram’s emotional impact from <a href="https://www.theglobeandmail.com/technology/your-smartphone-is-making-you-stupid/article37511900/">an article in <em>The Globe And Mail</em></a>, a top newspaper in Canada.</p> <h3>Andy Coravos on Twitter</h3> <p>Wait. @instagram strategically *withholds* &quot;likes&quot; from users that they believe might disengage hoping they&#39;ll be disappointed and recheck the app?! Harvesting painful insecurities. This is so messed up. https://t.co/tXs9R1T1zK</p> <p>Over 72 hours, the original Tweet garnered 1.7M+ impressions, Dopamine Labs and the CTO of Instagram both provided clarification, and I was faced with a series of choices about how to handle events as they unfolded over the weekend.</p> <p>As I learned more information — some confirming and some denying elements of the Tweet, I had questions: How do I provide an update? As new information arises, how do I provide context to the original Tweet?</p> <p>Twitter has a huge flaw in its product: the author of a post has only one suboptimal way to slow down the spread of information — <em>delete the Tweet</em>.</p> <p>In a world where anyone can share information that could be consumed by millions of people <em>within hours</em>, we are faced with a new set of social norms. I’ll dig into two issues:</p> <ol><li>How should we evaluate the veracity of the Tweet as new information arises?</li><li>How can Twitter give us better tools to manage updates when something is going viral?</li></ol> <h3>Part One: Welcome to The Attention Economy</h3> <p>As <a href="https://medium.com/u/7ae2822fb046">Albert Wenger</a> writes in his new book, <a href="https://www.google.com/url?hl=en&amp;q=https://worldaftercapital.gitbooks.io/worldaftercapital/content/&amp;source=gmail&amp;ust=1483671496922000&amp;usg=AFQjCNHaGtcwqsQ1cX2VkMhshIsir4-Yxw"><em>A World After Capital</em></a><em>: </em>with the onslaught of content, attention has become the most scarce and valuable human good.</p> <p>In “<a href="https://www.samharris.org/podcast/item/what-is-technology-doing-to-us">What is Technology Doing To Us?</a>,” Sam Harris and Tristan Harris, a former Design Ethicist from Google, discuss the arms race for human attention. In an ad-based economy where “engagement” is the holy-grail metric, content that sparks anger increases “time spent on site”. Why? Because when we are outraged, we re-engage with the platform. The lines between user engagement, addiction, and behavioral manipulation are hazy.</p> <iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fw.soundcloud.com%2Fplayer%2F%3Furl%3Dhttp%253A%252F%252Fapi.soundcloud.com%252Ftracks%252F317724861%26show_artwork%3Dtrue&amp;url=https%3A%2F%2Fsoundcloud.com%2Fsamharrisorg%2F71-what-is-technology-doing-to&amp;image=http%3A%2F%2Fi1.sndcdn.com%2Fartworks-000217828973-dsl77i-t500x500.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=soundcloud" width="800" height="166" frameborder="0" scrolling="no"><a href="https://medium.com/media/8c628afe1512e3d6d9650ce5474a0448/href">https://medium.com/media/8c628afe1512e3d6d9650ce5474a0448/href</a></iframe> <p>A <a href="https://www.wired.com/story/free-speech-issue-tech-turmoil-new-censorship/">national discussion</a> has started to emerge. Last month, two major Apple Shareholders pushed <a href="https://www.bloomberg.com/news/articles/2018-01-08/jana-calpers-push-apple-to-study-iphone-addiction-in-children">for a study of iPhone addiction in children</a>. As <a href="https://medium.com/u/b6e05447f2cf">Casey Newton</a> editor at <em>The Verge </em>argues<em> “</em><a href="https://www.theverge.com/2017/10/27/16552620/facebook-trust-survey-usage-popularity-fake-news">Facebook has a trust problem</a>.” Yesterday, Jack Dorsey issued a <a href="https://blog.twitter.com/official/en_us/topics/company/2018/twitter-health-metrics-proposal-submission.html">request for proposal</a> to “increase the collective health, openness, and civility of public conversation.” A new conversation is emerging to make social media products healthier.</p> <p>The Tweet I posted hit a nerve in the middle of this discussion. The following day, <a href="https://medium.com/u/7aa8a43f8aa1">Mike Krieger</a>, the CTO of Instagram, replied to my Tweet.</p> <h3>Mike Krieger on Twitter</h3> <p>@AndreaCoravos @instagram To be super clear, we don&#39;t do this.</p> <h3>Andy Coravos on Twitter</h3> <p>@mikeyk @instagram Thanks @mikeyk -- could you clarify by what you mean by &quot;this&quot;? Meaning that Instagram will always update the # of likes instantaneously? How are notifications handled? These often seem bundled on my app. What&#39;s the underlying algorithm/trigger points?</p> <h3>Mike Krieger on Twitter</h3> <p>@AndreaCoravos @instagram this = strategically withhold likes. replication lag/etc may mean things aren&#39;t instantaneous but not intentionally so. and notifications we try and strike a balance of being timely + not over-sending notifs. UI shows our latest/best count once you&#39;re in the app</p> <p>At this point, Mike has made it clear that any delayed “likes” inside the Instagram app are not intentional. <a href="https://twitter.com/sarahmei/status/952667861838778368">Sarah Mei</a> and <a href="https://twitter.com/JediJeremy/status/952394829140443136">JediJeremy</a> both tweeted explanations on how technical constraints could delay “likes” in the app.</p> <p>But what about outside-of-app (e.g. iOS/Android) push notifications? Mike notes that “notifications” are bundled and <strong>not</strong> real-time. And so we are left with a grey zone. Multiple people, myself included, have reported receiving iOS notifications saying someone has “liked” a photo <em>hours</em> after the “like” showed up in the app. <em>Is this a bug or a feature? </em>What metric does Instagram use to optimize the timing of its notifications?</p> <p>I reached out to Dopamine Labs, the team quoted in the article, and they stand by their statement, writing that it was based on <a href="https://twitter.com/UseDopamine/status/952683603527217152">public statements of Facebook leaders</a>. As the time of publishing this post, the Canadian newspaper cited in the Tweet <a href="https://www.theglobeandmail.com/technology/your-smartphone-is-making-you-stupid/article37511900/">updated the article with the CTO’s response</a>, but did not redact the original quote.</p> <h3>Dopamine Labs on Twitter</h3> <p>@AndreaCoravos @instagram @mattmayberry (2/5) . . IG doesn&#39;t send push notifications for likes when they happen. They pool them. To figure out when to send those bundles, Instagram uses a policy algorithm. It is well know industry best practice to A/B test push notification policies like the one used by Instagram .</p> <h3>Dopamine Labs on Twitter</h3> <p>@AndreaCoravos @instagram @mattmayberry (3/5) It&#39;s also best practice to set the &quot;success criteria&quot; used to pick between different test policies as a mix of short term and long term engagement metrics. Sean Parker has said &quot;consume as much of your time and conscious attention as possible. . .</p> <p>As an engineer, I’m assuming it would be difficult (or impossible) for Instagram to disclose their algorithm for push notifications for each user. However, without a public statement about what’s being optimized, it’s unclear what’s happening. Like drugs, algorithms have (unintended) “side effects,” and emotional impact may still occur even if unintentional.</p> <blockquote>Drugs have (unintended) side-effects, and algorithms do, too.</blockquote> <p>Thankfully, there’s a movement to make a change, even within Twitter itself with the (launched yesterday) <a href="https://blog.twitter.com/official/en_us/topics/company/2018/twitter-health-metrics-proposal-submission.html">Health Metrics Proposal Submission</a>, asking the public to propose better optimization/health metrics. Similarly, Mark Zuckerberg made a public announcement that his 2018 goal is to <a href="https://www.facebook.com/zuck/posts/10104380170714571">“Fix Facebook</a>,” and has been testing different Newsfeed algorithms <a href="https://www.theguardian.com/technology/2018/jan/11/facebook-news-feed-algorithm-overhaul-mark-zuckerberg">in favor of ‘meaningful social interactions’</a>.</p> <p>Other organizations, like <a href="https://medium.com/u/c806a873d503">Tristan Harris</a>’s <a href="http://humanetech.com/problem/">Center for Humane Technology</a>, which launched a few weeks ago, are reversing the digital attention crisis and realigning technology with humanity’s best interests.</p> <p>Nonetheless, until these changes are widely implemented, I still have a Tweet going viral. How can I add the context updates to the original Tweet ASAP? Read a short-term proposal in <a href="https://blog.andreacoravos.com/twitters-design-flaw-that-stunts-healthy-civil-productive-conversations-73628ecaa736">Part 2</a>.</p> <p><img src="https://medium.com/_/stat?event=post.clientViewed&amp;referrerSource=full_rss&amp;postId=d0e97a4b09f1" width="1" height="1" alt=""/>&lt;hr&gt;&lt;p&gt;<a href="https://blog.andreacoravos.com/going-viral-new-responsibilities-in-the-attention-economy-d0e97a4b09f1">Going Viral: New Responsibilities in The Attention Economy</a> was originally published in <a href="https://blog.andreacoravos.com">Andrea’s Blog</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.&lt;/p&gt;</p>]]></content><author><name></name></author><category term="attention-economy"/><category term="behavioral-health"/><category term="instagram"/><category term="social-media"/><category term="twitter"/></entry></feed>