himata4113

5 ms·

31.
by himata4113 1mo ago
That's because pepsi already had coca-cola's secret recipe. I'm sure if they didn't have it already they would have been more than happy to at least have some knowledge before reporting it, but not like they didn't
32.
by himata4113 1mo ago
The entire point is so you don't have to have a google account. Aurora actually works fine if you do sign in. This is just blocking anon downloads.
33.
by himata4113 1mo ago
We have starlink I am sure we can figure it out. Like... launch a better GPS system with one rocket?
34.
by himata4113 1mo ago
We have starlink I am sure they can figure something out.
35.
by himata4113 1mo ago
USD is falling vs euro so EURJPY is even more bleak. USD in general during those times was way stronger than it is today so those numbers are meaningless.
36.
by himata4113 1mo ago
I don't really care about 'future & children', but even I find this absurd.
37.
by himata4113 1mo ago
urgh, there goes my automatic redirect... any alternatives? Although, I don't see how this will hold up in court as they're pretty much an archive-like site hosting UGC.
38.
by himata4113 1mo ago
You can omit the first part, being neutral towards a controversial figure is okay!
39.
by himata4113 2mo ago
"argument is very weak" regardless as I said.
40.
by himata4113 2mo ago
Well the big problem with china is that they do not respect international law when it comes to technology theft. But that argument is very weak when it appears that a lot of what they do is out in the open for anyone to replicate.
41.
by himata4113 2mo ago
Have you considered that china has been working on these things for the past decade and they're finally getting results out of that? And yes, it is known that they do steal tech as samsung discovered in their tech preview phone bust in
42.
by himata4113 2mo ago
For me I just kind of ignore the fact that I am neurodivergent, it really doesn't matter. I always refused in school to get any help for it since it really bothered me that I should be treated different. I haven't even gotten diag
43.
by himata4113 2mo ago
Don't discount the capability of representing human-like intelligence with simple constructs. Give enough parameters and advanced enough training you can without a doubt create real intelligence. We've seen some of this already wi
44.
by himata4113 2mo ago
You can get away with murder if it can't be proven that it was intentional homocide, that's why detectives will spend an unreasonable amount of time getting a confession and hard evidence ALONGSIDE intent and motivations.
45.
by himata4113 2mo ago
I feel like this is the typical MacOS user stereotype.. Cosmetics over efficiency. No, for the love of god never stop making TUI's.
46.
by himata4113 2mo ago
Probably even higher because openai and anthropic undoubtably have the lowest cost per token generated, especially with cerebras being able to serve a million tokens every 16 minutes.
47.
by himata4113 2mo ago
This is a reminder that getting an air purifier with high airflow filters (NOT HEPA!) is lifechanging in cities with high pollution levels. I recommend getting one of these https://www.youtube.com/watch?v=sQGfO4hNb60 and ju
48.
by himata4113 2mo ago
It directly and indirectly increases your quality of life and it is something you use every single day.
49.
by himata4113 2mo ago
From what I understand when you re-tokenize the output you can simply look at how often certain tokens show up and the position of them, enough of these matches would result it watermarked text. Let's say we are at token 431 and there
50.
by himata4113 2mo ago
rustdesk just works, moonlight and sunshine is janky as hell outside the local network due to the fact it uses like 5 different streams.
51.
by himata4113 2mo ago
There goes the last argument that anthropic had. I think beyond this point we're entering the 'dark scary world' that dario predicted which in fact result in things going on as usual. Really, the amount of fear mongering is a
52.
by himata4113 2mo ago
To put it more accurately: people who don't do software use loveable to build apps for people who do not use software.
53.
OpenAI and Anthropic hidden CoT leaks when given deep_think tool.(twitter.com)
56 points by himata4113 2mo ago|6 comments
54.
by himata4113 2mo ago
From my experience their capabilities are extremely narrow and generally perform terribly when faced with issues outside comparatively narrow training data.
55.
by himata4113 2mo ago
It's not that their strategy is to train smaller models, it's the only choice they have. Training SOTA takes anywhere from 1.5b to 150b. We don't know the real cost of training for the chinese models, but mistral neither has
56.
by himata4113 2mo ago
I doubt it, a lot of what I use llms for is on the very edge of what these models are trained for.
57.
by himata4113 2mo ago
feels like a harness issue, bug fixes work out fine with omp.sh as long as it can be tested properly. Untestable things tend to lead to hallucinated results though.
58.
by himata4113 2mo ago
Compare the performance of a 980 and a 5050 and I am sure that will answer your question. Also models baked into the silicon are able to achieve efficiency that is simply impossible to achieve with programmable circuits, there is a general
59.
by himata4113 2mo ago
ehh, they're pretty good at automated performance research and bug fixes, especially when spanned across hundreds of them.
60.
by himata4113 2mo ago
Neither of these points feel true anymore. Models are very much predictable these days (except anthropic models). The real issue stems from letting them work on their own for far too long. Also we are not controlled by 2 companies anymore a

More ›