dcre

20 ms·

361.
by dcre 1y ago
I suspect that says more about Grok than anything else.
362.
by dcre 1y ago
I agree we should not take CEOs at their word, we have to think about whether what they're saying is more likely to be true than false given other things we know. But to trust Zitron on anything is ridiculous. He is not a source at all
363.
by dcre 1y ago
I think at that point there is strong financial pressure to figure out how to continuously evolve models instead of changing new ones, for example by building models out of smaller modules that can be trained individually and swapped out. J
364.
by dcre 1y ago
Why two months? It was almost a year between Claude 3.5 and 4. (Not sure how much it costs to go from 3.5 to 3.7.)
365.
by dcre 1y ago
The Amodei quote in my other reply explains why this is wrong. The point is not to compare the training of the current model to inference on the current model. The thing that makes them lose so much money is that they are training the next
366.
by dcre 1y ago
ICYMI, Amodei said the same in much greater detail: "If you consider each model to be a company, the model that was trained in 2023 was profitable. You paid $100 million, and then it made $200 million of revenue. There's some cost
367.
by dcre 1y ago
What are we meant to take away from the 8000 word Zitron post? In any case, here is what Anthropic CEO Dario Amodei said about DeepSeek: "DeepSeek produced a model close to the performance of US models 7-10 months older, for a good dea
368.
by dcre 1y ago
Interesting, thanks for linking that. ACP seems simpler. Nice comparison here by gpt-5: https://gist.github.com/david-crespo/e9a9ab88df73d4c2702ef04...
369.
by dcre 1y ago
Obviously the protocol is designed to encourage other agents to build their own integration. It seems like the kind of thing LLMs would be good at, so I expect to see it soon, especially for the open source ones.
370.
by dcre 1y ago
To me this (coupled with their DeltaDB announcement[0]) feels like Zed trying to get out of the business that Cursor is in, and maybe even kicking them on the way out, if the protocol helps bring-your-own-agent take off. These agent tools a
371.
by dcre 1y ago
It doesn't seem to matter: people have posted tons of examples on social media of non-AI base images that it was equally able to hold steady while making edits.
372.
by dcre 1y ago
Alarming hands on the third one: it can't decide which way they're facing. But Gemini didn't introduce that, it's there in the base image.
373.
by dcre 1y ago
The key is “stream” in the error message. If you turn off streaming in llm (looks like --no-stream will do it) then I think it should work. I don’t understand why they gate streaming specifically but that’s what they seem to be doing.
374.
by dcre 1y ago
I don't think anyone really knows, but I also don't think it's quite an either/or. To me a more interesting way to put the question is to ask what it would mean to say that GPT-5 is just applying patterns from its traini
375.
by dcre 1y ago
I understand that is what a lot of people are saying. It doesn’t match my experience.
376.
by dcre 1y ago
> To access their best models via the API, OpenAI now requires you to complete a Know-You-Customer process similar to opening a bank account. This is not entirely true. It was entirely true for o3, but for GPT-5 it is only true for strea
377.
by dcre 1y ago
The first premise of the argument is that LLMs are plateauing in capability and this is obvious from using them. It is not obvious to me.
378.
by dcre 1y ago
The other commenter is more articulate, but you simply cannot draw the conclusion from this paper that reasoning models don't work well. They trained tiny little models and showed they don't work. Big surprise! Meanwhile every oth
379.
by dcre 1y ago
Trusting things to work based on practical experience and without formal verification is the norm rather than the exception. In formal contexts like software development people have the means to evaluate and use good judgment. I am much mor
380.
by dcre 1y ago
This is not a good analogy because reasoning models are not choosing the best from a set of attempts based on knowledge of the correct answer. It really is more like what it sounds like: “did you think about it longer until you ruled out va
381.
by dcre 1y ago
Have you heard of Zed before just now?
382.
by dcre 1y ago
We don't know why that is sufficient to enable the models to develop the capability, and we don't know what they are actually doing under the hood when they employ the capability.
383.
by dcre 1y ago
Totally unwarranted level of confidence here about both claims.
384.
by dcre 1y ago
It is sort of amazing that it works, and no one knows why, but empirically speaking it is undeniable that it does work. The IMO result was achieved without any tool calls to a formal proof system. I agree that is a much more plausible-sound
385.
by dcre 1y ago
The problems are not in textbooks. I’m curious what would count as an out of distribution problem for you. Only problems no one knows how to solve?
386.
by dcre 1y ago
How about the International Math Olympiad? https://arstechnica.com/ai/2025/07/google-deepmind-earns-gol...
387.
by dcre 1y ago
"The bet was that by the following year, the application inference would cost 90% less, creating a $160 gross profit (+80% gross margins). But this didn't happen, instead of declining the application inference costs actually grew!
388.
by dcre 1y ago
gpt-5 with medium reasoning, no search: "Short answer: use the system zstd C library that ships with iOS; Apple’s Compression framework still doesn’t expose zstd. Link libzstd and call its streaming API from Swift." (and then like
389.
by dcre 1y ago
I think the concrete issue this points to is the thing that dynamically decides when to use reasoning failed to choose it in this instance. Sam Altman said it was broken on release day.
390.
by dcre 1y ago
Most likely the key here is web search. Later I will try the post’s example with gpt-5 with search. I would be surprised if it didn’t say the same thing. From a product point of view, it seems clear that just as they have work to get the mo

More ›