قد يعجبك
there is much talk of ai-assisted moral reasoning, but when ais begin to deeply diverge from our moral intuitions, how will we know whether to trust it?
many wish for a future shaped by a large coalition of augmented humans rather than a few hyperscalers but decentralization of power mustn’t be good inherently while it lowers variance, im yet to be convinced more “participation” will lead us to a future with more flourishing
right after playing social deception games, i notice that my automatic responses outside of the game are more likely to be deceptive, and i have to watch myself and stop myself from lying lol
the new uber ui is genuinely unusable please — i do not care about nearby restaurants — tell me where my driver is
this seems incredibly underdiscussed a lot of people are focused on extending cots, but the cots themselves look bafflingly inefficient
But the chain of thought here uses the exact string "they parted disclaim marinade—they parted illusions" 99 times, often over and over right after one another! It's very hard to tell a story in which doing this is somehow an effective way to solve problems.
me and the boyz blasting chapel roan in a shitty suv going down an empty freeway down the middle of nowhere see u at neurips
if you are open to changing the world without reaping the benefits, you may be able to change the world more than you would otherwise
eval prevails.
This is one of my favorite things about Anthropic culture as well. Dario kind of embodies this, but there's more generally a kind of "essay culture", a shared practice of open, intellectual debate, with norms that expect a kind of seriousness and earnestness.
Dario's essays and long debate slack threads are one of my favorite parts of Anthropic's culture. They're open, detailed - and incredibly raw. Everyone at the company ends up having a good sense of how the company is making decisions and what matters. Its the kind of thing that…
i feel so much better when i update up on long(er) timelines it just gives me more of a chance to actually do things in the world i do wonder if this biases me towards thinking timelines are longer than they are, because it’s pleasant (or the other way via overcorrection)
I wonder what govt intervention in agi development will look like in practice
i’m curious how much model-output reasoning correlate with CoT reasoning when you eg ask gpt 5 to explain its reasoning
United States الاتجاهات
- 1. Cowboys 67.2K posts
- 2. Cowboys 67.2K posts
- 3. LeBron 87.7K posts
- 4. Gibbs 17.2K posts
- 5. Pickens 12K posts
- 6. Brandon Aubrey 6,699 posts
- 7. #OnePride 9,248 posts
- 8. Ferguson 10.2K posts
- 9. #DALvsDET 5,956 posts
- 10. CeeDee 9,694 posts
- 11. Goff 8,058 posts
- 12. Eberflus 1,983 posts
- 13. Bland 8,189 posts
- 14. DJ Reed N/A
- 15. Shang Tsung 16.5K posts
- 16. Al Michaels N/A
- 17. Schotty 1,634 posts
- 18. Austin Reaves 13.1K posts
- 19. Turpin 2,082 posts
- 20. #LakeShow 4,265 posts
قد يعجبك
Something went wrong.
Something went wrong.