AI has already fueled breakthroughs in biotechnology—but now, further advances in AI are poised to fuel pharmaceutical discoveries as well. Sarah Guo sits down with Joshua Meier and Jack Dent, co-founders of Chai Discovery, whose newly launched Chai-2 designs bespoke antibodies that bind to their targets at a jaw-dropping 20% rate. Jack and Joshua talk about the implications for Chai-2’s success rate at discovering antibodies for the pharmaceutical industry, how structure prediction is pivotal in making the model work, and future potential for using the model to optimize other molecular properties. Plus, they talk about what they believe bioscientists should be learning to best utilize Chai-2’s technology. 

Sign up for new podcasts every week. Email feedback to [email protected]

Follow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @_jackdent | @joshim5

Chapters:

00:00 – Joshua Meier and Jack Dent Introduction

01:09 – Genesis of Chai Discovery

06:12 – Chai-2 Model

10:13 – Criteria for Specifying Targets for Chai-2

13:12 – How the Chai-2 Model Works

16:12 – Emergent Vocabulary from Chai-2

18:15 – Hopes for Chai-2’s Impact

20:33 – Reception of the Chai-2 Model

22:16 – Future of Wet Lab Screening and Biotech

27:08 – Optimizing Other Molecule Properties

31:37 – Where Chai Invests From Here

36:20 – What Bioscientists Should Learn for Chai-2

40:23 – How Jack and Josh Oriented to the Biotech Space

43:38 – Platform Investment and Chai-2

46:53 – Scaling Chai Discovery

48:21 – Hiring at Chai Discovery

49:09 – Conclusion