Sublime
An inspiration engine for ideas
1/ Is the idea of IQ legit or total B.S.? With the replication crisis in social science, it's worth asking this since a number of major psychology findings didn't hold up under scrutiny.
To find out, at Clearer Thinking, we ran a massive study...🧵
Spencer Greenberg 🔍x.com
This is a heck of a paper & summary graph
Who makes the best manager in a lab experiment? People who volunteer to lead do worse than a leader picked by lottery. Intelligence helps, but a game of assigning workers to tasks is more predictive. Race, gender & age aren't predictive https://t.co/cLSbRkywcB
Sex differences in work preferences among gifted men and women. The lower the number, the more men have the preference relative to women. (via @SteveStuWill) https://t.co/QOllpqxCi5
My favorite ideas from the episode with Charlie Songhurst https://t.co/QqdaXkrgJk
Will scaling reasoning models like o1, o3 and R1 unlock superhuman reasoning?
I asked Gwern + former OpenAI/DeepMind researchers.
Warning: long post.
As we scale up training and inference compute for reasoning models, will they show:
A) Strong... See more
How Tyler Cowen Uses AI
1) Grok: Quick and effective for fact-checking tweets.
2) o1 Pro: Best for queries. His primary daily tool.
3) Deep Research: Ideal for long-form (10-page) reports; extremely impressive but less practical for personal routine; occasionally... See more
David Perellx.comMy anecdotal evidence generally seems to support the idea that group sizes will usually plateau at a number lower than 150 participants. This comes from 20 years of doing facilitation both on and offline, running several software companies, and running various forums at America Online. In particular, many online communities provide good evidence... See more
Christopher Allen • The Dunbar Number as a Limit to Group Sizes
Work on these things - Marginal REVOLUTION
marginalrevolution.com
