andai

26 ms·

751.
by andai 3mo ago
Oops, all Claudes!
752.
by andai 3mo ago
This story was adapted into a short animated film in 1987: https://www.youtube.com/watch?v=5LNHYz89sNc (Russian, with subtitles.)
753.
by andai 3mo ago
Step 1: Develop human-like minds Step 2: Dehumanize them constantly
754.
by andai 3mo ago
Would be nice if they also tested other models instead of several old versions of GPT.
755.
by andai 3mo ago
How do you build intuition about something you cannot remember?
756.
by andai 3mo ago
The Zed Shaw method!
757.
by andai 3mo ago
Stairway to gulag for you!
758.
by andai 3mo ago
This seems to point in the direction of a best-of-n approach?
759.
by andai 3mo ago
That's a great point. It's like deepfakes. You only know it's AI when it's not convincing; when the deception has failed. When you're fooled, you don't know it's AI. You just think it's a video.
760.
by andai 3mo ago
I had them at the Giraffe in Leuven. So I was already in Belgium. I expected to take the train back home at 9. Instead I regained consciousness at five in the morning. Except I was already fully awake, when I "woke up", walking ba
761.
by andai 3mo ago
Well here's a fun one. If the outside is simulated, why not the inside?
762.
by andai 3mo ago
Did they ask Claude to do the job interview too?
763.
by andai 3mo ago
Don't they just have 3x as many video game nerds? Those numbers sound pretty similar per capita.
764.
by andai 3mo ago
> They're word-association mechanisms with no embodiment and no way to associate the text vectors they manipulate with real-world phenomena. Doesn't most of this also apply to a guy living in The Matrix?
765.
by andai 3mo ago
It kinda looks like we can only train AI on dead people's data. I wouldn't be too upset about that. They write better anyway.
766.
by andai 3mo ago
Yeah, a lot of people have reported significant gains from "model fusion" lately, which I think the literature just calls ensembles. You just ask a bunch of models to do a thing and then you combine the results. Research tasks ben
767.
by andai 3mo ago
I found it excerpted in Getting Things Done, where the author presented it as the ideal for how one should approach their life, both professionally and personally.
768.
by andai 3mo ago
I don't think posts should be flagged for being AI (they often contain interesting or important ideas as well), but I do think it should be pointed out.
769.
by andai 3mo ago
I heard a story about a guy who was trying to get Claude to implement some feature. It said it would be too hard, it would take weeks. Eventually convinced it to try, and it one shotted it in 30 seconds.
770.
by andai 3mo ago
At that point I'd just get out and take the stairs. Race ya!
771.
by andai 3mo ago
Original title: Manual: •.,:;…!?· (Except printed in a really cool font!)
772.
by andai 3mo ago
> but when running multiple "Please look for issues" iterations it would never have spotted the issue by itself. Nah that's old meta, you're supposed to generate candidate issues with an agent swarm running at high te
773.
by andai 3mo ago
Human doesn't write blog post, that's still Claude's job...
774.
by andai 3mo ago
https://xkcd.com/1035/
775.
by andai 3mo ago
I had 10 Jaeger bombs one time and I woke up in Belgium.
776.
by andai 3mo ago
There seem to be heart issues involved in most of the energy drink related deaths. More specifically, people finding out they had heart issues, by dying. So "those people just shouldn't drink caffeine then" would be the stand
777.
by andai 3mo ago
I know someone who doesn't have ADHD, but when she drinks coffee it makes her sleepy. Her whole family drinks coffee before bed!
778.
by andai 3mo ago
I think the actual threshold is when the withdrawal misery falls below the noise floor of the ambient misery.
779.
by andai 3mo ago
>A considerable body of research suggests that incidents like these aren’t just chance or circumstance. Mixing alcohol and energy drinks was linked to higher rates of binge drinking, sexual abuse and drunk driving, according to a study c
780.
by andai 3mo ago
>They're so, so incomprehensible because the LLM has a super limited theory of mind for readers. They always assume that external readers have access to the full context and history of decisions in the project development The transf

More ›