Five years ago, the AlphaFold 2 breakthrough on the 50-year-old protein folding and structure prediction problem provided a remarkable demonstration of AI’s ability to advance science in profound ways. The world took notice: in 2024, my colleagues Demis Hassabis and John Jumper were awarded a Nobel Prize in Chemistry for this work.
Perhaps more importantly Demis said this on protein folding more than 5 years ago (and it’s captured in the documentary film "The Thinking Game" - worth a watch!):
"We just fold everything. And then give it to everyone in the world. Who knows how many discoveries will be made from that?"
To date, the freely available AlphaFold Protein Database (launched in 2021) has been used by 3.3 million researchers across more than 190 countries. 1 million of that number are researchers from low and middle income countries, many working on neglected diseases and so much more.
With so much happening every week in AI, I continue to think the progress at the intersection between AI & Science is under-told. AlphaFold may be the most well known example, but notable AI-enabled advances in science are revolutionizing other fields, including genomics, neuroscience, material science, physics and more, offering us the chance to solve complex problems important for humanity’s progress.
Almost a year ago, Demis and I wrote the following on the possibilities of AI to advance scientific discovery:
https://lnkd.in/gCuBW4qd
So much more has happened in the last year that has appeared in science journals and some of which we’ve posted here -- and yet its still so early, still so many limitations to overcome, and so much more still to come.
For now to mark 5 years of AlphaFold’s impact, my colleagues Demis Hassabis, John Jumper, Pushmeet Kohli, and Anna Koivuniemi have written a blog post below to more fully tell the story and illustrate just some of the incredible work AlphaFold is enabling: https://lnkd.in/geu7Twkn