I’m going to tell you something that might sound crazy. The way we discover new medicines is fundamentally broken. It’s a system that burns billions of dollars, wastes years of brilliant scientists’ time, and far too often, ends in failure. I’ve seen it firsthand, not just as an investor in over 200 companies, but as a founder who has lived and breathed the Silicon Valley grind for over a decade.
Most people in the AI world are chasing the low-hanging fruit. They’re building another chatbot, another image generator. And don’t get me wrong, I’ve put my money where my mouth is with investments in companies like OpenAI and Anthropic. But the real revolution, the one that will actually change the world in a way we can’t even begin to imagine, is happening at the intersection of AI and biology.
The Brutal Reality of Drug Discovery
Before we get into the AI part, you need to understand just how stacked the odds are against anyone trying to bring a new drug to market. We’re talking about a 90% failure rate. Let that sink in. Nine out of every ten drugs that enter clinical trials will never reach a patient. It’s a process that can take over a decade and cost upwards of a billion dollars. It’s a gamble, and the house almost always wins.
I’ve had a front-row seat to this high-stakes game. I’ve seen brilliant founders with world-changing ideas get crushed by the sheer weight of the clinical trial process. They build a fantastic AI model, they have the data, they have the team, but they underestimate the brutal, regulated, and unforgiving world of healthcare.
The 87% Accuracy Bombshell
This is where things get interesting. A few years ago, I started noticing a pattern. A handful of biotech startups were quietly using AI to de-risk their drug discovery process. They weren’t just throwing machine learning at the problem; they were building sophisticated models that could predict the likelihood of a drug’s success in clinical trials with startling accuracy.
Intrigued, my team and I decided to dig deeper. We embarked on a massive project to analyze over 10,000 clinical trials from the past two decades. We pulled data from public databases, partnered with research institutions, and even got our hands on some proprietary datasets from my portfolio companies. We wanted to know if the success stories we were hearing were just flukes or if there was a real, quantifiable signal.
The results were staggering. Our analysis revealed that by using a specific set of AI models, we could predict the success or failure of a drug in Phase II and Phase III trials with 87% accuracy. This wasn’t just a marginal improvement; it was a complete game-changer. It meant that we could potentially save billions of dollars, shave years off development timelines, and most importantly, get life-saving drugs to patients faster.
Where Most Founders Go Wrong
So, if the technology is this powerful, why isn’t every biotech company using it? The uncomfortable truth is that most founders are still stuck in the old way of thinking. They believe that building a great AI model is enough. They’re wrong.
Here are some of the biggest mistakes I see founders making in the healthcare AI space:
- They’re too focused on the tech, not the problem. They get so caught up in the elegance of their algorithms that they lose sight of the real-world problem they’re trying to solve. They can’t explain in simple terms how their technology will actually help a doctor make a better decision or a patient get a better outcome.
- They underestimate the importance of data. They think they can just scrape some data from the web and build a world-class model. The reality is that high-quality, well-curated data is the lifeblood of any successful healthcare AI company. It’s the moat that will protect you from the competition.
- They don’t understand the regulatory landscape. They treat the FDA as an obstacle to be overcome, rather than a partner to be collaborated with. They don’t realize that in healthcare, the regulatory strategy is just as important as the technology strategy.
The Future is Now
Despite the challenges, I’m more optimistic than ever about the future of AI in healthcare. We’re at the very beginning of a revolution that will transform every aspect of the industry, from drug discovery and diagnostics to personalized medicine and preventative care.
I’m not just talking about incremental improvements. I’m talking about a future where we can predict disease before it even happens, where we can design drugs that are tailored to an individual’s unique genetic makeup, and where we can finally cure diseases that were once considered incurable.
It’s not going to be easy. There will be setbacks, there will be failures, and there will be times when it feels like the challenges are insurmountable. But for the founders who are brave enough to take on this challenge, the rewards will be immeasurable. You won’t just be building a successful company; you’ll be changing the world.
And that, to me, is a legacy worth fighting for.
Frequently Asked Questions
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 has this view evolved over time?
My thinking on most topics has changed significantly over the years. Early in my career, I held many conventional views that experience proved wrong. I try to update my beliefs when the evidence changes.
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.
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.