sparks

Perfect is boring. Perspective is interesting.

The biggest bottleneck in building superintelligence is that AI agents are not as of yet very good at evaluating how they’re doing at a given goal. If they could better self-assess, they could self-improve. And digital self-improvement loops could lead to superintelligence. Making progress on importing particular human taste/judgement into LLMs

”Once you do one thing, if you have a modicum of success, and you think you can do a second, third and fourth thing, you’re wrong. You can’t.”

-Micheal Saylor, via Laser Eyes

Reading the Mind in the Eyes Test

s3.amazonaws.com

enjoyed this

i wish people celebrated the internet as much as they bemoan it. it’s literally a World Wide Web of humans, an unthinkable miracle that you and i can connect because i tap my fingers on a piece of glass for 30 seconds. let us not be numb to this!! bring back wonder & appreciation

Isabel x.com

“It’s not the thing it’s your relationship with it”

Human advantage will be caring about other humans: “humans care about other humans and also not at all about machines”

His most striking prediction: “Who in this room would say with conviction, I’m sure I’ll be smarter than GPT5? Not a single person.

Sam Altman bfast with JB.
(note: i want to write about intelligence)