Generative ML in chemistry is bottlenecked by synthesis

4.4k words, 21 minutes reading time

Arguments Generative ML in chemistry is bottlenecked by synthesis 4.4k words, 21 minutes reading time Abhishaike Mahajan Sep 16, 2024 49 16 6 Share Note: I am not a chemistry expert. Huge shout-out to Anand Muthuswamy , Gabriel Levine , and Corin Wagen for their incredible help in correcting my misunderstandings here! But some may remain, please DM me or comment if you see one! Also, PostEra is a sponsor on this post! They are one of the few small-molecule ML startups that specialize in helping address synthesizability challenges in generative models. This is distinct from the ‘synthesis probl

saved by

• Sudarsh K
• Sri Nandan Gondi
• Curtis Chong

related reading

• A primer on why computational predictive toxicology is hardsubstack.com
• A primer on why computational predictive toxicology is hardowlposting.com
• Using Self-Correcting Search to Accelerate Materials Discoverygoodfire.ai
• diffuse.onediffuse.one
• Evolutionary Scale · ESM3: Simulating 500 million years of evolution with a language modelevolutionaryscale.ai
• Rafael Gomez-Bombarelli: The bittersweet lesson of scaling in AI for materialsbidmap.berkeley.edu
• Making Claude a chemist \ Anthropicanthropic.com
• The ML drug discovery startup trying really, really hard to not cheat (Leash Bio)owlposting.com
• Teaching generative models to hallucinateblog.escalante.bio
• Generative AI: A Creative New World | Sequoia Capitalsequoiacap.com
• AquaGen_RXRX.pdfvalencelabs.com
• Program Synthesis with Large Language Modelsarxiv.org