nicce

5 ms·

61.
by nicce 26d ago
You also want prompt/prefix cache. Otherwise there is a lot of duplicate prefill processing if you fork conversation, have a different chat window or anything like that. It therefore also makes subagents much faster.
62.
by nicce 26d ago
>> Fable was the step change for programming > AFAIK: Mistral does not even try to compete in this field. They have released models speficially for programming that ”vibe coding” would be safer. https://mistral.ai/n
63.
by nicce 26d ago
> Just the other day I saw some meme about how if we saw as many ads taking over sites like we do now, that was a sure sign of having your machine getting infected with a virus. That used to be the reality around decade ago for those who
64.
by nicce 26d ago
Better to buy GPUs and RAM with current rate
65.
by nicce 26d ago
It is not just the money but social media is controlled by right wing ideologist. Musk has openly supported right wing parties in Europe.
66.
by nicce 26d ago
Oops
67.
by nicce 26d ago
5h limit is gone with one prompt with Sol or Astra with high or higher thinking.
68.
by nicce 26d ago
Well, that makes them just even slower then
69.
by nicce 26d ago
That MXFP4 is an excellent project. But I have difficulties on reading that README. Is it intentionally generated like that with LLMs?
70.
by nicce 26d ago
2x RTX 3090 is not enough for proper use. ideal is 64GB+ so that you get proper cache and 256k context with good enough models. (e.g. Qwen 3.8 27B with MXFP4). And that is tight already. You would need 3x RTX 3090 - but then the PCIe bandwi
71.
by nicce 27d ago
I bought Apple TV many years ago and haven’t plugged tv into internet since.
72.
by nicce 27d ago
Duolingo has joined to the addiction club
73.
by nicce 27d ago
Both can be true. I was saying marketing trick about the response and selected words ”how great it was”.
74.
by nicce 27d ago
Sounds like marketing trick once again, to be honest.
75.
by nicce 27d ago
Not far away if it could be completely removed with all the other competitive technologies: https://www.uschamber.com/technology/data-privacy/impacts-of...
76.
by nicce 28d ago
People already look me like crazy when I still review all the code changes AI does, and not even using the agent directly. And make it do things differently if it is unmaintainable slop. Their argument is that are you really going to modify
77.
by nicce 28d ago
Meanwhile it still can’t write a sentence without adding three semicolons.
78.
by nicce 1mo ago
You can run it with 2x r9700 with 150-200 tokens per second. It is intelligent enough if you just point the docs / whatever for it.
79.
by nicce 1mo ago
They seem to have good enough general intelligence that missing knowledge is not that big thing. If you are able to have a proper [free search engine], they can do almost anything. Having own local search index about relevant stuff can help
80.
by nicce 1mo ago
Yeah. Unless human can verify it, not sure if it is certain or useful.
81.
by nicce 1mo ago
Many. Too lucrative for certain companies and even governments to allow that to happen
82.
by nicce 1mo ago
> Version bumps should really be about removing deprecated cruft rather than shiny new features. Can there be deprecated cruft without new features? :-D
83.
by nicce 1mo ago
Issue is that it is extremely hard to notice if there is backdoor or some training-related hallucinations that are completely random. E.g. I had random, completely unrelated and irrelevant fetch by Qwen3.8 27B to " https://r
84.
by nicce 1mo ago
I used Fable once. Used through API and asked it to review one 2k word plan. It costed me 15 dollars and haven't used it since.
85.
by nicce 1mo ago
> Apple doesn't design GPUs on-par with Nvidia's efficiency yet How much it matters in inference? Most GPUs have enough computing for that and the bottleneck is the RAM speed and size. And M5 Ultra is becoming to challenge this
86.
by nicce 1mo ago
First thing that came to my mind... why people are so self-centric... and yet, somehow, you feel bad because the same people flex with these CVEs while you can't tell a single huge one you had found under NDAs, while they keep downplay
87.
by nicce 1mo ago
Depends on what you do. For prototyping GUIs without well-defined framerwork, Rust is terrible. TypeScript wins completely. For back-ends, not so clear, since it is easier to write the intent there and even prototype must be somewhat correc
88.
by nicce 1mo ago
> I’ve got some news to share about your new Macbook. As long as you pay for Apple Care, not a problem. But maybe not the point…
89.
by nicce 1mo ago
Thank you! Have to look into it.
90.
by nicce 1mo ago
Money always wins, sadly.

More ›