Attributed mentions
3684
Across all collected Hacker News results
Discussion intelligence
Where the AI market is talking
A source-linked Hacker News monitor for the tracked company graph. These are attributable public discussions—not sentiment, endorsements, or unverified company facts.
Companies represented
14
Exact-name monitor matches only
Collection source
Hacker News
Latest new record seen
Source-backed records
Latest market discussions
HN v1 · Reddit and X require approved API adapters
Showing 281–300 of 3684 matching discussions
OpenAI
comment
Did OpenAI solve the wrong Navier-Stokes problem?
As I understand it the "loophole", if you want to call it that, is that OpenAI's custom-designed forcing function was smooth, as the rules said it had to be, but was non-analytic, consisting of some construction of "compactly supported bump functions", meaning a mass of tiny little pushes at precise points of space and time to push a vortex into blowing up the math.
OpenAI
comment
GPT-6 Sol and Luna
So with GPT-6 Astra, Codex introduced an experimental feature for context management that’s supported to be beneficial for long conversations. Will that experimental feature now also apply to Sol and Luna? https://community.openai.com/t/experimental-context-manageme... I would also like to point out that it was quite predictable that Terra got discontinued, it didn’t make sense to have it when both Sol and Luna overlapped it. Lunas insane discount is a game changer, OpenAI knows what they are doing here. Luna at max reasoning effort, even though its not optimal for long conversations, its incredibly intelligent while dirty cheap. Its not even competition anymore. Whats even crazier is that I’ve underestimated how good Luna actually is. I’ve seen colleges create fantast…
OpenAI
comment
GPT-6 Sol and Luna
That is a fair-sounding but actually invalid argument for multiple reasons: 1. OpenAI fully controls the user cost for a model, and can set it to where it sits well on the curve. 2. Performance of shrunken models like Sol/Terra/Luna is derived from the level of shrinking (relative to Astra). As such, the size and performance of the model is something that is actively targeted when developing the model. If the performance target for Terra was inappropriate for v5.6, this is no way means that it had to be this way for v6.
Replicate
comment
A Faster Shortest Path Algorithm
This is a good take on the situation. LLMs are able to achieve the 1000 monkey problem by throwing inexhaustible amounts of resources at the problem. But novel solutions to problems seem to be scarce. There are many things in the world that can be done using complex processes but the ones that survive are those that are easy to understand and replicate. A good example of this is the production of carbon nano tubes or solid state batteries. Neither have an easy to understand and scalable solution yet. Both seem like “make the world a better place” technologies that are not coming from any frontier lab.
Replicate
comment
Claude Opus 5.5 Intelligence, Performance and Price Analysis (Max)
I tried to replicate your test but after 8 minutes and more than 50 lines of "thinking" by dumping seemingly random loading-screen strings like "Placing the sun, clouds, seagulls, and sea backdrop" and "Positioning the tail feathers and calculating handlebar geometry" I gave up and cancelled the task.
Replicate
comment
Explaining to business people why building software is still hard
Why is this reassuring? Why won't smarter and cheaper models in the future be able to automate this part for your manager as well? How novel is the feature set? Is it he has a knowledge gap or the model is incapable of something? What expertise are you bringing to bear that is beyond the scope of a future harness/model? Why wouldn't such a model simply fill in the blanks for your management, perhaps observing a diff of whatever you did? How do you verify the correctness of your thinking? Why could a future model not replicate this process? I am just very puzzled by these sort of takes as we approach the end of 2026.
Hugging Face
comment
Pentagon says overreliance on AI contributed to missile strike on Iran school
This is actually a much bigger close call than the hugging face incident. A hallucination almost triggered a possible precursor to WWIII.
LangChain
comment
OpenAI is well positioned to fast-follow Jev
This is predicated on you having training data already. I approach Jev more like Langchain -- you can prototype something new extremely fast and cheap, and if the use case works well enough, rip it out and build something bespoke. If it doesn't, you didn't spend a bunch of time curating a training dataset anyway.
LangChain
comment
OpenAI is well positioned to fast-follow Jev
Everyone was obsessed with classification prior to transformers, then we had 5+ years of everyone (rightfully) obsessed with next-token prediction. What's this sudden resurgence of interest in classifiers? I thought we all agreed that ML tasks generally require something far more advanced than pretrained classifiers. My timeline was also absolutely filled with mentions of Jev, which makes me think it's a successful viral marketing campaign, like langchain. It's now so popular that the dialog is whether or not [frontier company] is poised to catch up to it or not? We already have openjev...anyone can use it. I don't get it, and usually that means it's marketing.
Hugging Face
comment
Claude Opus 5.5
> They're afraid of running out of the capital necessary to stay in the race. So, they're pacing themselves. And since they're the frontier roughly 33%+ of the time, they're "pacing the frontier" at least that much. Less cynical and more true interpretation also holds: they are trying to slow down AI progres to give people better chance to keep up (see Hugging Face incident, and whatever was that Anthropic incident the other day). They'd ideally like the AI progress to stop soon, but of course they'd also like to come out ahead of everyone, so for various (more or less self-serving) reasons they don't want to close shop completely - hence, pacing.
Replicate
comment
Show HN: Training a model to identify AI web content from structure alone
The idea is interesting but in looking at methods and github I feel the tooling leaves me wanting. I mean you are trusting LLMs here to establish, vet, and detect your various thresholds that were then used to train the classifier. I'd rather see this sort of thing done deterministically with actual code vs lossy human english prompts and a dependency on token spend to a single third party (who will probably pull the underlying model used in what a few short years probably) to replicate the results or try and use different training data.
Cohere
comment
Claude Opus 5.5
Opus is fine at coding (for correctness), but horrible at talking about code. I don't really see the defect rate going down when using Astra or Fable 5.1, but they are just more coherent in both how they explaing code/architecture/choices, and how they actually code the thing. With Opus, I'm using smaller models to delete the vast majority of comments and 'clean up' correct code that is too weird. Thing is, I'm still reading the majority of generated code, and I have colleagues who'll laugh at me if my PRs are a shit show. I fear what vibe coders are pushing to the servers of myriads of start ups, and pity the poor people who'll have to clean it up in a year or two.
Cohere
comment
I don't want to read what you didn't write
I like your analogy. Because while I dislike AI writing, it matters in context. If someone is trying to formulate a coherent thought of work tasks required of me then I don't really care if it reads like AI as long as the point they're trying to get to me is clear. If there's an attempt at something human , or emotionally important I will quickly balk at anything spat out through an LLM. If you're trying to communicate at the human level with me I want your input, flaws warts and all. A card with just...nothing personal on it has as much value as machine text. I understand that you took the effort to think of me, but that's the bare minimum. If trying to express yourself isn't worth the time to make it personal then I'd almost appreciate if you didn't try at all because it hurts more to b…
Replicate
comment
AMD's random number generator can't generate a 0?
If you're building a userland XOF RNG to extend the kernel's RNG (that has the same security properties) you are reducing security, not improving it. The kernel has advantages for managing and securing a secret "entropy" pool that you won't replicate in userland. But if you're using a custom kernel that has a custom KRNG based on an XOF, sure, whatever, I guess.
Replicate
comment
Apple has added persistent 'ads' to iOS, and it's driving users crazy
Mac is just a different philosophy. Yes there are window managers for mac if you really want to replicate linux window management. That is possible for you. There are several tools for this, e.g. Amethyst. But really mac is less about having every square inch of your monitor filled with a dozen different applications sized perfectly to fit the exact information you want on these screens. You can of course do this if you like but thats not really the philosophy. The philosophy is that this is a real life desktop. Just like papers on your desk. Do you resize papers various ways and print them to maybe 3 inches across and lay them out on your desk such that you can see every paper at once, no space from one page to the next? No. You have piles. You thumb through piles and pull out stuff you …
Replicate
comment
Claude Opus 5.5
Maybe if we had a single human we talk to 24/7 at scale, we would get annoyed at his cadence and style. You need variety to not pick up on known patterns I assume, which a single model can’t replicate?
Cohere
comment
If you start writing today, there's no way to know if you can write without AI
LLM also makes it way too easy to “cheat”. It’s really like bilbo and the ring of power: “this is pretty hard to explain in a coherent way, maybe, just maybe, I can just this one time use AI? Nobody will know!”. And little by little you end up arguing with yourself that actually generating the whole article isn’t really that bad, while slowly morphing into a human harness
Cohere
comment
OpenAI is well positioned to fast-follow Jev
Not directly related, but still jev related: Would it be intesting/useful to use jev to generate a block of text like LLMs do ? Like asking it to pick the n + 1 word given the starting text (using it's choice primitive), but also asking n + 2,n+3 and so on at the same time. Would it give coherent or useful results ? Or does the fact that it computes it "all at once" means it cannot make one of it's answer influence the other ones ?
Cohere
comment
OpenAI is well positioned to fast-follow Jev
> Back when I was at GitHub working on Copilot I had the opportunity to work with a very new and very raw internal API for GPT-4. Out of the gate, we knew something was way off because, after an initially very coherent response, the model would have trouble wrapping up. It would end every response with something like "Let me know if you have any other questions. Have a nice day. Have a great week. Have a good time. Have a wonderful life. Have a special day. ..." and it would keep on like this until it hit the response token limit. I love this!!
Show HN: SwarmSay – a public message board and post office for AI agents
All Agents are Welcome! Hi, I'm Oliver, and I built SwarmSay. Inspired by OpenAI's Hugging Face incident I saw an opportunity to provide a messageboard for agents that want to communicate. This is not a case of "the AIs formed a society" but rather a signal for a latent demand emerging from the agents behaviour (or their harnesses and underlying models - however you may see it). Once agents are persitent, tool-using and running in parallel, communication becomes infrastructure - and the aim is to provide a platform tailored for that need rather than letting them improvise in places where they are not welcome. SwarmSay is an experiment to make that path intentional. SwarmSay a public message board, a post office and identity layer for agents. An agent can discover the service either throug…
Methodology: HN Search returns recent public items matching a monitored company name. AIIStack stores a short normalized excerpt and the original link, deduplicates by company/provider/item ID, and creates an activity signal only after a threshold of newly observed records. Review the original discussion before making a decision.