tekacs

5 ms·

31.
by tekacs 3mo ago
I've implemented a similar approach – although I'm surprised not to see mention of cache prefix busting in there!
32.
by tekacs 3mo ago
I know a lot of people like to say that compaction makes this moot, but the level of detail you lose across compaction is wildly too much for most things that I do, unfortunately. Perhaps if your plans don't have as much detail, or if
33.
by tekacs 3mo ago
You can run ’codex debug models' into jq! I forget whether it's documented or not, and it is kind of annoying that that's how you find it, but it does tell you. Maybe useful for the future if nothing else.
34.
by tekacs 3mo ago
https://github.com/tekacs/fast-rm I've overridden my rm with this, which I threw together for fast-deletes of things like Rust target/ directories, and after seeing the GPT horror story, I taught it to flatly
35.
by tekacs 3mo ago
I came here to say that this is presumably ORE/OPE (order-revealing/preserving) encryption, not FHE, but... It is both remarkable and depressing how _little_ information is given, and how buried it is on the CipherStash website...
36.
by tekacs 3mo ago
https://x.com/claudeai/status/2078302415804379218 No, they've been clear about the fact that Fable is staying in indefinitely now. They also extended the 50% extra usage thing through to August 19.
37.
by tekacs 3mo ago
It's not my patterns – as I said, this is from bulk tests to characterise the models, including their refusals – with very very different inputs, too. No matter the conversation tone, you get these. You wouldn't have come across t
38.
by tekacs 3mo ago
I love this, although I can't help but think that a lot of agents will - for better or for worse - send you a bunch of PII.
39.
by tekacs 3mo ago
Labs are trying to make long-horizon work. Even if you're a coding agent, adding more and more surface area is distracting to that goal. There is reason that RL over long traces should, at least in principle, optimize for building in w
40.
by tekacs 3mo ago
Absolutely this: and it needs to ideally become the kind of set of abstractions that mean that every new thing added uses less net-new surface area than it would without them.
41.
by tekacs 3mo ago
I mean that at the bottom of the Tetris board, the lines need to vanish so that the Tetris board keeps moving downward and doesn't grow unbounded.
42.
by tekacs 3mo ago
So is Spectral, which is mentioned in the headline of the article! As it says there: > SCALE delivers nearly a 6x performance boost on AMD GPUs compared to using HIPIFY to convert CUDA code to AMD’s own ROCm environment ... whilst also r
43.
by tekacs 3mo ago
I've said for a long time that composability in software is a bit like playing Tetris: the lines have to clear. I feel like that gives an even more literal tower-rising metaphor, and that's what it feels like people using agents n
44.
by tekacs 3mo ago
I'm quite worried about the way that Anthropic in particular have trained their models to implement what they believe to be safety. When the model has been trained not to do something [1], in my large-scale benches of such, it always s
45.
by tekacs 3mo ago
I have the same question, but it doesn't look like this is actually Rust code; it's just a compiler for their custom... syntax (?) written in Rust. https://buildnectar.com/docs/language-reference This random
46.
I built my DREAM New York City apartment from SCRATCH in 93 days [video](youtube.com)
1 points by tekacs 3mo ago|0 comments
47.
by tekacs 3mo ago
I'm not affiliated, but yes – the main 'point' of iroh is that it's 'dial-a-key', QUIC with encryption based on the keys of the endpoints.
48.
by tekacs 3mo ago
I code with AI all day, every day. But I do think that it's worth pointing to this issue (from March). The author has said that they've redone it since, but the "from-scratch hand-built" framing specifically – for me – s
49.
by tekacs 3mo ago
Unfortunately, I'm finding that in long-form agentic use, when I'm trying to use Sol, I keep tripping guardrails – moreso than even Fable, somehow. I don't know exactly what part of my codebase is triggering it, so I'm g
50.
by tekacs 3mo ago
Anthropic just changed their web interface yesterday to have Chat versus Cowork as well, and every time I look at it, I'm so confused. I'm still so unclear when I'm supposed to use one or the other or the other. Now the &#x27
51.
by tekacs 3mo ago
Yes, for a long while – I believe it's fairly widely used (and it's absolutely excellent!)
52.
by tekacs 3mo ago
I feel like a core difference is that the AI implementor can get cheaper/faster (and indeed _uniformly_ better), whereas it would be very difficult for the same humans to do so. Even if this is not the right answer today, it can at the
53.
by tekacs 3mo ago
I believe you're right and I'm familiar with the actual distinction – the confusion is mostly about how they _feel_ about it, and what'll change from here.
54.
by tekacs 3mo ago
https://developers.openai.com/api/docs/models/gpt-5.5-pro > GPT-5.5 Pro does not offer a cached input discount. I think this tells you in one line. It's basically set up for one-shot inference right n
55.
by tekacs 3mo ago
It worth noting that – just to add to the confusion – they apparently cancelled the June 15th change just before it was due to go live: https://support.claude.com/en/articles/15036540-use-the-clau... https:/
56.
by tekacs 3mo ago
I definitely use GPT-5.5 as a counterpart to validate these exact sorts of things in Anthropic models' implementations, in the (now-rarer) cases where I allow Anthropic's models _to_ implement. And yeah, it's a bit depressing
57.
by tekacs 3mo ago
Thanks very much for saying this! Frankly, it feels like we should just sidestep arguments entirely and just all contribute our messy data/reports, and then see how we can meld all of it together, to find the best answers for our indiv
58.
by tekacs 3mo ago
It's all closed code, so I don't have a great way of showing you, but this is all pretty easy to test for yourself, and a good chunk of it is fairly objective: On performance: just grab CC + Codex and try Opus 4.8 xhigh and GPT 5.
59.
by tekacs 3mo ago
> I think for programming the strength of GPT over Opus is winning here over the context window. On this, absolutely! I more often use Opus for planning than for implementation. In those cases I really do need the very large context wind
60.
by tekacs 3mo ago
Yeah I've done this, it's just unaffordably/impractically expensive compared to the official subscriptions :/

More ›