ftxbro

47 ms·

901.
by ftxbro 3y ago
maybe robots made them
902.
by ftxbro 3y ago
> "There is no claim about the IQ at all" > "This has nothing to do with the IQ" sorry i forgot that people on the internet don't understand analogies Maybe it would have been clearer if I would have said some
903.
by ftxbro 3y ago
I'm not trying to hate on Anand but this has big '200 IQ research mathematician says it's a myth that you have to be a genius to become a successful research mathematician' energy.
904.
by ftxbro 3y ago
There are rumors that the 'code interpreter' plugin in particular is different, like it uses more advanced training possibly even a more advanced pre-trained model and that it's better understood as like GPT-4.5. I personally
905.
Cinematic Mindscapes: High-Quality Video Reconstruction from Brain Activity(arxiv.org)
1 points by ftxbro 3y ago|1 comments
906.
by ftxbro 3y ago
Thanks that one is very interesting. I especially like how GPT4 performs worse (zero shot greedy decoding) than ChatGPT on those questions. I wonder have you thought about how it relates to 'inverse scaling' ( https://gi
907.
by ftxbro 3y ago
bill gates got honeypotted oh no anyway
908.
by ftxbro 3y ago
effective accelerationism lets gooooooo!
909.
by ftxbro 3y ago
> "Definitely finding negative results like this is tough by just doing experiments, since it could be that the model might have succeeded if you had just trained it for more time/made it bigger/etc." This is interest
910.
by ftxbro 3y ago
is this the internet of things
911.
by ftxbro 3y ago
our ai is going to make robots that could replace humans, but don't worry robots don't have humanlike eyeballs so we are also making an eyeball scanner that scans your eyeballs everywhere you go so you will be identified as a huma
912.
by ftxbro 3y ago
> "If by intelligent you mean" the meaning of intelligence is up to the poll responder and it's an implicit part of the poll
913.
by ftxbro 3y ago
is this some class warfare
914.
by ftxbro 3y ago
the machine spitted some machine oil onto your sandwich when you didn't tip
915.
by ftxbro 3y ago
If some method becomes popular (like transformers in this case) there are so many waves of papers where people 'interpret' the new thing by saying how it's really [their own research project / grant / area of stud
916.
by ftxbro 3y ago
This isn't it. Look at Figure 4 in https://arxiv.org/abs/2303.08774 those test results aren't like reading your horoscope and thinking hmmm that does sound like me and they aren't like philip k dick us
917.
by ftxbro 3y ago
As everyone is saying in replies, it's not so far out there compared to what some more mainstream-seeming AI people think. As one example consider https://tinygrad.org/ it's been featured recently at the top of ha
918.
by ftxbro 3y ago
they are going to start prompting for tips at the self service gas station pumps, along with the ads that you have to click through before you can pay for gas
919.
A Personal View of Average-Case Complexity (1995) [pdf](gwern.net)
63 points by ftxbro 3y ago|3 comments
920.
by ftxbro 3y ago
> "People think that P≠NP, because they tried it hard, and did not find any NP-hard problem in P." That was the informal explanation given to the computer scientists. But that's not so convincing to non computer scientists
921.
by ftxbro 3y ago
> "10-15 years ago, the basic idea was "P≠NP because otherwise computers could do crazy shit." Everyone is downvoting you because that's not the mathematical explanation, but it's absolutely the informal explanat
922.
by ftxbro 3y ago
those responsible have been sacked
923.
by ftxbro 3y ago
it sounds identical to http://suckless.org/philosophy/
924.
by ftxbro 3y ago
can i use this on myself to achieve a deeper level of laziness
925.
by ftxbro 3y ago
literally leaving money on the table if you don't have a tipping prompt for your service
926.
by ftxbro 3y ago
I guess what I'm supposed to do in 2023 in USA if I have a prediction, like a prediction that increasing params+flops+dataset LLM training will keep unlocking new cognitive capabilities, is to make money from it by doing some kind of i
927.
by ftxbro 3y ago
Thanks I didn't know you could do it. I made one here https://news.ycombinator.com/item?id=36078637
928.
Poll: How many of you take as an axiom that a computer cannot be intelligent?
3 points by ftxbro 3y ago|8 comments
929.
by ftxbro 3y ago
A computer can never be intelligent by definition.
930.
by ftxbro 3y ago
It is possible that a future computer could be intelligent.

More ›