Excellent @emilymbender essay outlining the difference between thinkers who are studying real, present dangers of AI -- like how it's being used to make horrid, GIGO-fueled decisions about people's lives ...

... and the realm of tech folks who ignore that and instead focus on fuzzily-defined, gauzily sci-fi prospects of AI taking over the world, using everything from Bell-Curve-era definitions of IQ to longtermist calculations (i.e. "pulling numbers out of their asses")

https://medium.com/@emilymenonbender/talking-about-a-schism-is-ahistorical-3c454a77220f

@clive I’m starting to wonder if this is a consequence in part of so few of the technical papers on LLMs getting peer reviewed? Like Reviewer #2 would definitely make this point.

@jetjocko

good question! I’d be interesting to figure out precisely how many papers on the existential risks of AI have ever been peer reviewed

@clive or even if the technical papers that mention risk have.