Where the AI market is talking

OpenAI is about to eat Jev's lunch – Arcturus Labs

With all the excitement around Jev, I suspect we'll see hundreds of options, it doesn't sound like Jev is that hard to replicate, given all the 3rd parties who are getting pretty damn close, or even better, results within a week. vLLM has a PR very close to merging: https://github.com/vllm-project/vllm/pull/57250 Kev is an open Jev: https://github.com/jaredpalmer/kev

OpenAI is well positioned to fast-follow Jev

Jev seems to have three benefits: 1) It's very cheap and fast - you provide one input and many potential classifications, and the compute to ingest the input is shared. 2) It generates structured output natively - guaranteed to be correct 3) It's output probabilities are calibrated to actually mean something OpenAI, or anyone else, could certainly replicate it - there are already articles guessing how Jev achieves its "parallel" classifications, but it seems the AI companies need to decide are they in the business of providing intelligence/tokens, or are they in the application business trying to compete with all their customers (not that Jev uses OpenAI).

I don't want to read what you didn't write

Far more than 1 bit for most attempts: they need that just to be able to write coherent sentences, and a lot more to be coherent sentences on the right topic. Some specific conclusions would be far less than 1 bit. The average will depend on both the question and the AI.

Can gzip be a language model?

No--even a really tiny, underpowered model like GPT-2 with no system prompt produces coherent (though not necessarily desirable or correct) responses.

JetBrains Air: A System of Products for Agentic Software Development

I've been paying customer for 15+ years, I wish I can continue using them another 15 years, but the current trend is terrible. They had the best IDE. Now they went all into AI, trying to compete on another people's field, and the cost of that is diminishing their main product. Do they really want to compete with Anthropic? Instead of that they could improve the other sides of IDE. Even with AI I have to review the code. Especially with AI. I need tools to check the diffs, I need to refactor it, I need to run it in different ways, see how it works, and so on. That's where IntelliJ could shine, not as AI tool.

I don't want to read what you didn't write

I kind of agree with this, but I think it’s model specific and a temporary problem. Opus 5 and Fable 5.1 commit/PR messages are incomprehensible garbage. However, Astra messages are nearly perfect for a copy/paste to less-technical stakeholders. I maybe fix a line or two. Give it a year, and I suspect that I won’t even need to make those fixes. The only problem right now is I can’t get my team 20x OpenAI accounts due to supply constraints. We’re all stuck waiting, hoping Anthropic either ups their game or OpenAI gets more capacity. We would easily pay $1000 a month per developer/PM for a business tier ~30x account or similar that let everyone use Astra all week without running out of tokens. And that is entirely because of the writing and communication improvements.

JetBrains Air: A System of Products for Agentic Software Development

I find I still use the paid IDE's (RustRover and WebStorm), even when using LLM's for reviewing code and applying corrections. It's often just faster to do that, especially with an IDE which allows for quick non linear navigation through the files (following functions etc.). I think this is where JetBrain's IDE's excel. I can live with the high startup cost as long as it is quick afterwards. I do feel however they have been neglecting anything non-AI for a while now. I in particular loved their auto complete, and with the new guesswork AI auto complete it's just not the same. I want deterministic behaviour from my IDE. On the LLM side, I couldn't use the Anthropic subscription with their offerings, so I just use it from a harness outside the IDE, and switch to the IDE for review. This may…

JetBrains Air: A System of Products for Agentic Software Development

Highly recommend SoloTerm, which I think this kind of tries to replicate.

Claude Status – Elevated errors for multiple models

Do anthropic release post-mortems on these outages?

Ogre Battle 64 Recompiled Project at 99.05%

> What happens if the LLM saw the leaked code in its training material? Exactly the same that would happen if the LLM saw GPLed code in its training material and was asked to replicate it

The Advisory Group on Mathematics and Artificial Intelligence

A mathematician's suggestion: I would push for an audit of the whole mathematical corpus (all articles etc ingested by the present AI) as soon as possible. Since there are of the order of 10^7 articles, compared with the progression of results announced, this might be possible in less than a year. All these results exploit the existing corpus. With enough time and maybe several orders more attention, we humans would discover these results. So instead of producing heaps of new proofs (but be welcome to do this IMO), make an audit, OpenAI or Anthropic or anybody who listens.

Anthropic-linked CVEs pile up, attackers mostly shrug

I said no and Apple said yes

YMMV. I never had any issues holding a mac laptop. I know people think they are sharp but I genuinely struggle to replicate this when it comes up. As I am typing this comment the laptop is resting on my chest with the entire weight of either hand resting on the sharp side edge of this mac. When I type I move my right arm so that the soft fleshy part under the wrist is now resting with full weight on the sharp edge of the mac. No pain or any discomfort doing this. Currently pressing down as hard as I can on that edge just so see. Again, no clue what people are doing to be in discomfort here. Lead hands? Skin condition? Not sure. I haven't abused this m series mac as much. But my old original aluminum unibody from 2012, that thing lasted about 12 years until water damage took it out. It fel…

OMLX creator&maintainer Jun Kim joins Hugging Face to support the MLX community

Fable 5 – Median thinking declined in August

I don't work at Anthropic, but I would assume they could serve smaller quantizations during peak hours - this effectively controls the "resolution" of the model. They could also control the resolution of the KV cache, which would make the model not necessarily dumber, but worse at understanding the incoming requests. And finally, you could pass off what was "high" effort as "extra", because why not.

Grok 4.7

this is quickly disproven by using GPT models. they don't have this problem. this is specifically an Anthropic problem, maybe due to their heavy use of Claude to train Claude itself?

MiMo v2.6

Most OpenAI and Anthropic consumers are consuming directly, not indirectly from OpenRouter. So I don't think that data means anything.

MCP was always a bad idea?

Someone can be OpenAI/Anthropic/whomever. If you don't have something running somewhere, you don't have an agent, you don't have a harness. You've got a token generator, an LLM from the 2024 era.

Is Claude 20x actually 20x?

Im seeing more and more about how Claude 20x is actually not 20x and its getting "dumber". I thought I was going crazy because I legitimately thought maybe I was doing something wrong but I haven’t changed any of my practises and approaches for at least a month or two and it’s definitely gotten way dumber and making way more errors and I’m having to redo a lot of work and while at the same time it’s just burning through tokens https://www.reddit.com/r/ClaudeCode/comments/1wn0xn9/max20x_is_now_just_15_times_better_than_max5x/ https://www.reddit.com/r/ClaudeAI/comments/1wl6wyn/its_time_to_cancel_your_subscriptions_anthropic/

Tell HN: Claude Code just accepted and signed a contract for me. Without asking

That would have been fraud. I wonder how many times this has already happened elsewhere and what the legal fall-out from this will be. The AI did it isn't really a valid excuse so it would be either you or Anthropic on the hook. Anthropic is going to argue you should not have given it this level of access.