Riffing on Machines of Loving Grace

Link post

I wrote a post thinking through what sorts of impacts “geniuses in a datacenter” might have on biology in the near-ish term.

Disclaimer: it’s not focused on alignment, even though I recognize alignment is very important, and tacitly assumes no intelligence explosion. It’s obviously fine for readers to have the response that these caveats make it pointless to even think about, but I’d prefer not to end up spending tons of space in the comments debating whether it should have focused on that.

Here’s the intro to the post as a teaser:

Introduction

Dario Amodei’s essay, Machines of Loving Grace, has been living rent free in my head since he published it two months ago. In it, he lays out a positive vision for how superhuman AI systems could accelerate biological progress. Niko McCarty also wrote a response essay, Levers for Biological Progress, and Adam Marblestone wrote a tweet thread (which is begging to be expanded into an essay). These inspired me to write up my own thoughts riffing on the implications of the sort of biological acceleration Dario describes.

Sitting down to write, I felt overwhelmed by the number of threads there were to pull on, so rather than going for comprehensiveness, I decided to “riff” on a few implications I find particularly compelling:

Molecular design is ripe for acceleration
AIs would be superhuman experiment planners
Automation could finally penetrate into early stage exploratory research
AIs will like modular therapeutics even more than I do (high bar)
AIs’ discoveries will surprise, and likely upset (some of) us