From skeptic to builder: why I changed my mind about AI for biology

This week’s Expert View comes from Albi Celaj, head of machine learning research at Deep Genomics, who reflects on how far AI-driven biology has come since its early, idealistic days. Drawing on a decade in the field, he traces the journey from skepticism to foundation models — and what it takes to build real drug discovery platforms from code.

In 2014, I sat in a room at a “Genetic Networks” conference listening to Brendan Frey present his talk titled Decoding the Genetic Determinants of Human Disease. Brendan presented data from his 2015 paper that AI had “decoded biology”, solved part of the splicing problem and could help to replace many experiments - we just needed to go out and apply it. I remember the optimism, but thought that there’s no way that these efforts had involved the realities of wet-lab work, let alone the more difficult constraints of drug discovery.  

Part of me wanted to dismiss the talk, but I couldn’t shake the feeling that there was something compelling underneath.  More than a decade later, I can definitely say that this moment was part of a broader movement in AI - what sounded like provocation in 2014 was just an early signal.

Before Deep Genomics: a practical skeptic