Episode 99
The cave: pharma's data problem
Data Science Leaders | 30:20 | June 03, 2026
Dave Johnson’s first pharma job wasn't in a lab. It was in a room, dark and cramped, stacked floor to ceiling with 30 years of decaying discovery chemistry data on every storage medium imaginable.
They called it “the cave.”
A physicist turned data leader, Dave has spent two decades solving the same problem: scientific data is generated for one purpose, poorly captured, and left abandoned.
Now co-founding Dash Bio after years leading data science at Moderna, Dave joins Thomas to make a case the industry would rather not hear: AI pilots keep failing because organizations can't understand their own data or tell whether anything they build improves how they work.
Listen in to hear:
- Why "the cave" is still an accurate description of how most pharma companies manage data
- Why clinical trial numbers mean nothing without the metadata behind them
- How the US needs to treat China’s clinical trial volume as a wake-up call
