David Fajgenbaum
speaker
1,017 appearances
2 recordings
2 series
first heard Sep 2025
last heard 26 Nov
David Fajgenbaum’s voice in public audio — every appearance, attributed to the second.
Trend
recordings per month · last 12 monthsRecordings per month over the last 12 months — 1 in all, peaking in Nov 2025 with 1.
Appearances
Right.
to come up with predictions on how likely drugs are to work in new diseases.
It's really mind boggling and just like to sort of put some numbers around it.
When we built our first AI platform about two and a half years ago and ran it for the first time, it took us 100 days to compute 75 million scores because there's 4,000 drugs and there's 18,000 diseases.
So if you tried every drug for every disease,
It would be 75 million possibilities.
It took 100 days.
Now it takes about 17 hours to compute the same 75 million scores.
The scores are really accurate.
They're telling us that drugs like lidocaine can help potential patients with breast cancer, things that our brains would have never just gone to immediately.
But what's so interesting also is that a lot of these insights are based on research that
We humans have already done, you know, we've studied this drug in the lab for this disease and maybe showed promise.
But then we moved on to that drug for that disease because this wasn't a profitable opportunity or maybe this was a better opportunity for someone's career.
And so there's all these breadcrumbs that have sort of been spread all over the medical research system.
And artificial intelligence is so good at picking up those breadcrumbs, making the connections.
And then it's really up to us.
And like you, we won an Audacious Project award this past year.
And now it's up to us to leverage the power of artificial intelligence to look across 75 million possibilities.
But then for us humans to say, this drug for this disease really needs to be studied in the lab.
This one needs to be studied in clinical trials.
Showing 901–920 of 1,017 · page 46 of 51
← Previous
Next →