I’m going to tell you a story. A story about a 5-year-long, soul-crushing, and ultimately failed journey to build an AI diagnostic tool. A journey that cost me millions of dollars, countless sleepless nights, and a good chunk of my sanity. And I’m going to tell you the brutal, unvarnished truth I learned from it.
I’ve been in Silicon Valley for over a decade. I’ve had two successful exits, RemoteTeam which was acquired by Gusto, and MovieLaLa which was acquired by Gfycat. I’ve also been fortunate enough to be an angel investor in over 200 companies, including some of the biggest names in AI like Anthropic, OpenAI, Scale AI, and Hugging Face. I’ve seen the AI revolution from the inside, and I’ve never seen anything like the hype around AI in healthcare. Everyone is talking about how AI is going to revolutionize medicine, cure diseases, and save lives. And you know what? They’re not wrong. But they’re not telling you the whole story.
They’re not telling you about the brutal, regulated, and soul-crushingly slow world of healthcare. They’re not telling you about the years of work it takes to get a product to market, only to have it rejected by doctors who are used to doing things a certain way. They’re not telling you about the endless meetings with hospital administrators who are more concerned with budgets than with innovation. They’re not telling you about the sheer, unadulterated pain of trying to build a business in a field that is actively hostile to change.
I am. I’m going to tell you all of it.
The “Genius” Idea
It all started with a simple idea. A “genius” idea, I thought at the time. I was going to build an AI-powered diagnostic tool for mental health. The idea was to use machine learning to analyze patient data—everything from their speech patterns to their writing style to their social media activity—and identify early signs of depression, anxiety, and other mental health conditions. The goal was to create a tool that could help doctors diagnose these conditions earlier and more accurately, and get patients the treatment they needed before it was too late.
I was convinced I was going to change the world. I had the technical expertise, the funding, and the connections. I had a team of brilliant engineers and data scientists. We had access to a massive dataset of anonymized patient data. We had everything we needed to succeed. Or so I thought.
Building the “Perfect” Model
We spent the first two years building the AI model. We tried every algorithm in the book. We fine-tuned our models for months. We A/B tested everything. We were obsessed with accuracy. We wanted our model to be perfect. And you know what? We got pretty close. Our model could predict the onset of depression with 90% accuracy. It was a technical masterpiece. We were so proud of it. We thought the hard part was over.
We were so, so wrong.
The Hard Part: Reality Hits
Building a great AI model is the easy part. The hard part is everything else. The hard part is navigating the brutal, regulated world of healthcare. The hard part is dealing with the gatekeepers, the bureaucrats, and the luddites. The hard part is trying to sell a product to people who don’t want to buy it.
Here are just a few of the brutal truths I learned during those five years:
Data is a nightmare. Getting your hands on clean, labeled, and usable data in healthcare is next to impossible. We spent a year and a half just trying to get the data we needed. We had to deal with endless legal agreements, privacy concerns, and technical hurdles. And even when we got the data, it was a mess. It was full of errors, inconsistencies, and missing values. We spent more time cleaning the data than we did building the model.
Regulations will crush your soul. The healthcare industry is a regulatory minefield. HIPAA, FDA, IRB—it’s a never-ending alphabet soup of acronyms that will make you want to tear your hair out. We spent hundreds of thousands of dollars on lawyers just to make sure we were compliant. We had to go through a years-long FDA approval process that was so opaque and arbitrary it felt like a Kafka novel.
Doctors are not your friends. You would think that doctors would be excited about a tool that could help them do their jobs better. You would be wrong. Most doctors are overworked, underpaid, and deeply skeptical of new technology. They’ve been burned by too many empty promises from tech companies. They don’t want to change their workflows. They don’t want to learn a new system. They just want to be left alone to do their jobs.
The sales cycle is eternal. Selling to hospitals is a special kind of hell. The sales cycle is brutally long. It can take years to get a deal done. You have to go through endless committees, presentations, and pilot programs. And even then, there’s no guarantee that you’ll get the sale. We had one deal that was in the works for two years. We had a signed contract. We were ready to go. And then, at the last minute, the hospital’s IT department killed the deal because they didn’t want to integrate our tool with their ancient EMR system.
The Team That Burned Out With Me
I wasn’t the only one who poured my heart and soul into this company. I had a team of 15 brilliant, passionate people who believed in the mission as much as I did. They were the best and the brightest. They could have worked anywhere, but they chose to work with me. They worked nights and weekends. They missed birthdays and holidays. They sacrificed so much for a dream that never came true.
And when I had to tell them that it was over, that we were shutting down, it was the hardest thing I’ve ever had to do. I saw the disappointment in their eyes. I heard the tremor in their voices. I felt the weight of their shattered dreams. And I knew that I had failed them. It’s a feeling that will haunt me for the rest of my life.
A Glimmer of Hope: What I'd Do Differently
It’s easy to look back and see all the mistakes I made. It’s harder to admit them. But if I could go back in time, here’s what I would do differently:
I’d start with the problem, not the solution. I was so in love with my “genius” idea that I never stopped to ask if it was a problem that anyone actually wanted to solve. I was a solution in search of a problem. I should have spent more time talking to doctors, patients, and hospital administrators before I wrote a single line of code.
I’d build a smaller, simpler product. We tried to do too much, too soon. We should have started with a much smaller, more focused product that solved a single, specific problem. We should have gotten that product into the hands of users as quickly as possible and then iterated based on their feedback.
I’d hire a healthcare expert from day one. I thought I could learn the healthcare industry as I went. I was wrong. It’s a complex, insular world with its own language, culture, and rules. I should have hired someone who had been in the trenches, who knew the players, and who could have helped me navigate the regulatory minefield.
The Brutal Truth I Learned
After five years of banging my head against the wall, I finally gave up. I shut down the company. I laid off my team. I walked away with nothing but a mountain of debt and a broken heart.
So what’s the brutal truth I learned? It’s this: building a great AI model is the easy part. The hard part is everything else. The hard part is the data, the regulations, the doctors, and the sales cycle. The hard part is the human element.
And here’s another brutal truth: the “move fast and break things” mantra of Silicon Valley is a recipe for disaster in healthcare. You can’t just launch a product and see what happens. You can’t just iterate your way to success. You have to be slow, methodical, and deliberate. You have to build trust. You have to play by the rules. And you have to be prepared for a long, hard fight.
So, What Now?
I’m not telling you this story to discourage you. I’m telling you this story because I want you to be prepared. I want you to know what you’re getting into. I want you to have a realistic understanding of the challenges you’re going to face.
Because here’s the thing: I still believe in the power of AI to revolutionize healthcare. I still believe that we can use technology to solve some of the biggest problems in medicine. But I also believe that we need to be smarter about how we do it. We need to be more patient. We need to be more collaborative. And we need to be more humble.
So if you’re an entrepreneur who’s thinking about getting into the healthcare space, here’s my advice: don’t do it because you think it’s going to be easy. Don’t do it because you think you’re going to get rich quick. Do it because you’re passionate about solving a real problem. Do it because you’re willing to put in the hard work. And do it because you’re prepared to fail. Because in healthcare, failure is not just an option. It’s a rite of passage.
Frequently Asked Questions
What experience informs this perspective?
This perspective comes from over a decade of building companies in Silicon Valley, two successful exits (RemoteTeam to Gusto, MovieLaLa to Gfycat), and investing in 200+ startups including Anthropic, OpenAI, and Scale AI. I write about what I've lived.
Do all experts agree with this view?
No, and that's fine. The best ideas in business are often contrarian. I share my perspective based on my experience and data, but I encourage you to seek out opposing viewpoints and form your own conclusions.
How can I apply this thinking to my own situation?
Start by identifying the core principle behind the opinion, not the specific example. Then ask yourself: does this principle apply to my context? If yes, test it in a small, low-risk way before going all in.
What's the most common pushback you get on this?
People often push back by citing exceptions or edge cases. And they're usually right that exceptions exist. But building a strategy around exceptions rather than patterns is a losing game for most founders.