This number was first proposed in the 1990s by Robin Dunbar, a British anthropologist who found a correlation between primate brain size and average social group size.[3] By using the average human brain size and extrapolating from the results of primates, he proposed that humans can comfortably maintain 150 stable relationships.

oh okay

Also very disappointed to learn that Ian Pace, a pianist I admired, has been oneshotted by social media and gone down the racist transphobe 24 hour tweeting monster chute.

I love it when Americans get to find out that being a big racist is actually illegal in most of Europe.

this blog post is so badass

My favourite error, and it appears repeatedly, is that instead of an object passing from one character to another, a copy of the object suddenly finds itself in the second person’s hands. It is shockingly bad at portraying any kind of action! It’s never going to get better!

As usual this sort of message actually makes me worried about what she might do to herself. :(

I want any LLM I use to choose the very best, most precise words at every single decision point

lmao no you don’t

Oh boy, this is the most straightforwardly delulu lesswrong post I’ve ever seen.

If you see issues in the YouTube video, rest assured that those issues survived even more than two generation attempts.

lol

What I find delightful about this is that I already wasn't impressed! Because, as the paper goes on to say

Moreover, although the UBE is a closed-book exam for humans, GPT-4’s huge training corpus largely distilled in its parameters means that it can effectively take the UBE “open-book”

And here I was thinking it not getting a perfect score on multiple-choice questions was already damning. But apparently it doesn't even get a particularly good score!

From Re-evaluating GPT-4’s bar exam performance (linked in the article):

First, although GPT-4’s UBE score nears the 90th percentile when examining approximate conversions from February administrations of the Illinois Bar Exam, these estimates are heavily skewed towards repeat test-takers who failed the July administration and score significantly lower than the general test-taking population.

Ohhh, that is sneaky!

I love the way these idiots keep incrementing the number on their ChatGPT fantasy as if it's a sufficient image of the future and it's going to get everyone on board. Complete failure of imagination, don't try to picture any actual use for it or anything, just make it... more.

Oh well done, you added noise to a line going up!