MIT Technology Review is hosting a live Roundtables conversation that will explore an emerging and provocative question in biology and artificial intelligence: can machine-learning systems help invent entirely new forms of life?
The discussion centers on work by Samuel King, a Stanford University PhD candidate in bioengineering, who in 2025 used a generative AI model to produce proposed genetic “blueprints” for microscopic viruses. The output is not, at least so far, a confirmed case of AI bringing a living organism into existence. Instead, it represents an early step: using AI to suggest plausible viral genomes that researchers could, in principle, study and test. Still, the project offers a preview of what may come if generative models move from drafting sequences on a screen to guiding experiments that create novel biological entities in the lab.
King will be interviewed by James O’Donnell, MIT Technology Review’s senior reporter covering AI. The conversation will also touch on King’s recognition as one of the publication’s Innovators Under 35, and on how computational tools are reshaping the way scientists conceptualize biology—shifting from observing what evolution has already produced to imagining what might be possible to build.
The event is scheduled to stream live on October 16 at 18:30 BST, corresponding to 1:30 p.m. EDT and 10:30 a.m. PDT. Alongside O’Donnell and King, the session will reflect King’s affiliations with Stanford University and the Arc Institute.
The Roundtables program arrives during a broader debate about what today’s AI can and cannot do. A decade after AlphaGo’s landmark victory over Go champion Lee Sedol, MIT Technology Review points to a gap between that breakthrough and current AI systems: many modern models still do not consistently harness the kind of strategic machinery and structured exploration that powered that win. In the publication’s view, AI “agents” remain limited in their ability to independently pursue truly novel, open-ended research agendas—particularly the kind that would amount to genuine scientific innovation rather than incremental variation.
That skepticism forms part of the backdrop for the interview. Public conversations about artificial general intelligence and sweeping new capabilities have accelerated, but the outlet argues that many such claims weaken when examined closely. Against that noisy landscape, the Roundtables event positions King’s work as an example of younger researchers pushing at the boundaries of what AI might contribute—especially in domains like bioengineering, where the step from digital design to real-world consequence is unusually direct.
MIT Technology Review notes that its broader programming includes additional coverage, event announcements, and subscription offers for readers following these developments.
Why it matters
If generative AI can reliably propose viable genetic designs, it could accelerate parts of bioengineering—from basic research to medicine—while also raising new safety and governance concerns. Researchers, regulators, and the public have a stake in how quickly “in silico” designs move into living systems. The next key test will be whether AI-generated viral genomes remain theoretical artifacts or become experimentally validated organisms, and what safeguards accompany that transition.