chis

3 ms·

1.
by chis 9d ago
Personally I didn't find that guy very convincing. I think current frontier AI would already probably enable a non-expert to do gain-of-function work on a virus which could result in something like Covid. And AI looks to keep getting
2.
by chis 11d ago
I gave up on fable 5 after I asked it to critique my PR and it spit out a page of complete nonsense technical jargon. Like, to the point that I had to review the feedback with other models and try to parse what it was saying and ultimately
3.
by chis 13d ago
> I don’t think AI is going to usher in an extinction event. In fact, even if nobody were to slow down, I really don’t think humanity would have much to worry about. I mean sure. If you feel that way then your p doom is zero, and it make
4.
by chis 14d ago
Chinese AI companies are a lot smaller in terms of capital, GPUs, and headcount as compared to American. The fact they’ve kept close so far is super impressive
5.
by chis 14d ago
GDP growth and inflation are different things. If inflation-adjusted GDP goes up, that means the average person is able to buy more things they want. Compare living in the US vs India today for the median citizen
6.
by chis 14d ago
Even at the scale of 100 people there is a coordination problem. If the 50 most conscientious researchers quit, then the 50 left behind would be the more aggressive group and be less interested in AI safety. Personally I agree though. I wou
7.
by chis 14d ago
Nvidia is limited by the number of engineers they have. It might turn out that the AI market is so lucrative that it’s best to reallocate their gaming-focused engineers to AI. Apple TV is at least a growth market for them, whereas gaming is
8.
by chis 15d ago
HN comments are just on a different planet lately lol
9.
by chis 16d ago
Fantastic answer thank you
10.
by chis 16d ago
> Opus
11.
by chis 16d ago
Are there cybersecurity concerns in the frontend? I would have thought you have to assume the client is untrusted and only do security work on the backend
12.
by chis 17d ago
The OG pros honestly sucked and had horrible reliability. The other models have treated me well
13.
by chis 17d ago
The connection type has nothing to do with it though. Modern bluetooth is basically transparent with 256kbps AAC, like you'd have to be an omega-expert listening in a silent room to identify it vs lossless. Certainly the difference y
14.
by chis 18d ago
I think you really have to squint to call this a disproof lol
15.
by chis 24d ago
Certainly it's not possible, yeah. A better test might be to have the AI report either confirmed good, confirmed toxic, or unknown. Even an experienced mushroom hunter needs to humbly put many mushrooms into the "unknown" c
16.
by chis 26d ago
what are you smoking
17.
by chis 26d ago
Sometimes I’m amazed anyone is willing to live without a git ui. The graph structure git uses is just such a natural fit for visual processing, plus the default git ux is borderline user-hostile until you master it. What’s also funny is tha
18.
by chis 1mo ago
You’re still stuck paying for a 3 year contract for deprecating hardware. And the 256gb model which is still not enough is $230 a month. And it runs models 8 months behind frontier
19.
by chis 1mo ago
This is just super unlikely to occur in the near term compared to some of these other risks. It's not like an instance of fable could just introspect into itself and pull out the weights. Model weights are stored encrypted and are hi
20.
by chis 1mo ago
So reading more I guess the judge narrowly ruled that destructive scanning is legal because of the one to one replacement, but didn’t rule on whether or not non-destructive scanning would be legal. It’s still not tested if scanning in that
21.
by chis 1mo ago
Whatever judge ruled that it was legal to scan and use books if and only if you DESTROY a copy of the book was truly a moron. The law gets bent by judges all the time to be practical and fair, they should have used this opportunity to give
22.
by chis 1mo ago
I think Fable is the beginning of Anthropic switching to training models as agent-first, tool second. It’s certainly the best model if you want something to work autonomously without supervision and don’t care to read the code. The code and
23.
by chis 1mo ago
Have you tried Sol? Just curious. I don’t wanna sound like a shill, just feels like every generation it’s important to reevaluate models and pick the best again. I do find myself returning to 4.6 for casual conversation - asking it to help
24.
by chis 1mo ago
Anthropic is lucky that they've built a lot of loyalty over the last year that they can burn through right now. I see people talking about switching back to Opus 4.8 rather that using 5.6 Sol, which is wild. My current approach is to
25.
by chis 1mo ago
It’s very difficult to intentionally find time to be bored. It’s sort of the opposite of an activity, and the brain hates it. One passable option is to take a 20 minute walk with no stimulus beyond the scenery. Journaling regularly. Medita
26.
by chis 1mo ago
You really have to be able to hold both opinions at once about Elon. Yes, Tesla and Spacex are unbelievable companies and perhaps nobody else on earth could have made them happen - so he’s a genius. But also he repeatedly promised autopilot
27.
by chis 2mo ago
It’s funny how different everyone’s experience is with this stuff. To me the diminishing returns are more around not going crazy with prototyping or running with xmax thinking all the time. I haven’t found it hard to stay under the usage li
28.
by chis 2mo ago
https://www.pangram.com/blog/third-party-pangram-evals You can also just try it yourself I guess really what convinced me was how it perfectly agrees with my own judgement.
29.
by chis 2mo ago
This has been pretty well studied. Pangram is about as good at an expert human with a 98% detection rate and <2% false positive, for flagging AI generated text. https://arxiv.org/pdf/2501.15654
30.
by chis 2mo ago
92% of this text is detected as AI. It may be time for Hackernews to integrate a Pangram detector into the UI, similar to what substack is doing :)

More ›