Open Thread 230

...

This is the weekly visible open thread. Post about anything you want, ask random questions, whatever. ACX has an unofficial subreddit, Discord, and bulletin board, and in-person meetups around the world. 95% of content is free, but for the remaining 5% you can subscribe here. In this week’s news:

1: Solenoid_Entity, who runs the ACX Podcast, is expanding his empire to run a Less Wrong Curated Podcast and Metaculus Journal Podcast. Good luck, Solenoid!

2: More volleys in recent AI conversations: Cameron Bucker (“debates in deep learning are now repeating the same mistakes as comparative psychology”), Edwin Chen (how do humans do on the same questions Gary Marcus asked GPT?), and a new paper, Emergent Abilities Of Large Language Models (formalizing the insight that as models scale up, they can do completely new types of tasks, not just the old tasks better).

3: Related: does anyone reading this have access to the new Parti-20B image model from Google? I would like to check whether I have won my bet with Vitor about image model progress. If yes, I will write a post about it and give you good publicity. Please contact me at scott[at]slatestarcodex[dot]com

4: In Obscure Pregnancy Interventions, I mentioned a Mongolian study found that air purifier use during pregnancy increased birth rate, but that nobody had tested yet whether it increased later-life IQ. Those results are now in and it appears that it does. Caveats (some from here) only marginally significant, your area is probably less air polluted than Mongolia, normal air conditioners might work as air purifiers already. Update: Emil K says it seems p-hacked.