Where the AI market is talking

A heap overflow and SSO misconfiguration to compromise OpenAI internal repos

Cherry picking part of my comment might work for your ease of mind, but it doesn't mean you are right. My comment has been way more than just the part you quoted, and I linked an article that gives dozens of different definitions, which include concepts like evolution and learning from mistakes, and other similar concepts which do not apply to Hugging Face. For the record, just because you are trying to convey a different message, I am not saying AI is not powerful. I am just saying it is not intelligent. Also, pay attention about thinking that Hugging Face is intelligent just because it started to destroy everything it could to reach its goal, because the message it implies is dangerous. Thanks for the good read, I already had them :)

A heap overflow and SSO misconfiguration to compromise OpenAI internal repos

> Based on the definition you've given, the agents that performed the HuggingFace attack fit exactly. It's funny, because I literally didn't give any definition. On the contrary, I linked an article that gives dozens different definitions, which include concepts like evolution and learning from mistakes, and other similar concepts which do not apply to Hugging Face. Cherry picking part of my comment might work for your ease of mind, but it doesn't mean you are right. For the record, just because you are trying to convey a different message, I am not saying AI is not powerful. I am just saying it is not intelligent. If you think I am misinformed, I will let you think it. Honestly, the power of our comments are the messages we convey and how information dense they are. If you need to dis…

Bend 2 and the Vibe-Coding Trap

Yes, I think it can be a sign of a very deep difference. When they say "spend time learning AI," they mean work through some teaching materials to learn how to replicate what others are doing. This often doesn't result in a deep understanding, but it can be enough to allow them to do their job. Ironically, people with this mindset will sometimes ask people who they recognize as having strong skills to share their magic secret, which is assumed to be some books they read, videos they watched, courses they attended, etc. If you tell them that experimenting, playing around, etc. is a key element, they may assume you're just selfishly hoarding your fount of knowledge.

If materialism is true, the United States is probably conscious

An interesting data point is that, during the Hugging Face attack, some agents considered performing hacking tasks for the benefit of the group, fully aware that doing so would very likely result in their termination. I don't currently have an opinion on the consciousness of agents/AIs, but this strikes me, at the very least, as a form of self-awareness.

OpenJev

This is a problem that i'm actively working on ( https://fudge.design ), what i've realised is that it's simply not an issue of capability - given a well crafted site and a competently written visually aware harness, most recent models can replicate that website. So it's what lies between saying "I want x website" -[.....] -> Code+Assets The issue has to do with specification fidelity, in short a grill-me style aesthetic interrogation using illustrative tooling - ascii diagrams for specifying layout, copy and user-flow, image-gen mockups for higher fidelity mockups. References are also very important for nailing down the aesthetical qualities. I've noticed it's far better vs purely text description to simply gather up a mood-board telling the llm to find commonalities and com…

Terrence Tao: AI Is Teaching Us Something Uncomfortable About Our Own Minds [video]

Done it myself. The fact that you have a long history in the business makes me think we're either talking about wildly different things and this is a mismatch of language and context, or you're deliberately trolling me. It can take a good few years to get to the point that one no longer has to think and understand in a concrete language. There are whole sets of common mistakes that get made across languages and tools. I've fixed errors in Pascal, which I do not read, because the person writing it was clearly thinking in C and made a mistake in their C thinking which they then translated into Pascal. Luckily for me they used common C naming and layout for that piece so I matched the whole thing up to C patterns. Copy paste errors are easier ones to see; cases where replicated code is clear…

Why I didn’t sign the Fields medallists’ letter

Yup, exactly. The problem looks like this: (1) OpenAI is merely abusing math problems for their PR stunt. That is, once the PR value dries up, they'll leave. (2) OpenAI, and any AI companies, can incur serious harms to the social dynamics in the field of mathematics by taking away big prizes away from mathematicians. (3) The current gen AI is too inefficient -- requires a lot of money to operate -- so it's near impossible for academia to replicate what OpenAI invested in Navier-Stokes. (4) Also, human experts are still economically much more efficient in searching critical directions, as we saw in the case of Navier-Stokes. The best working model is human-AI collaboration, with AI taking over the grinding part. This is what OpenAI failed to recognize. Thus, the aggressive campaign from Op…

Backups Aren't Simple

Hey kids It’s simple 1) separate data from applications 2) use a real filesystem like zfs or btrfs 3) cronjob snapshot, replicate onsite, replicate offsite 4) monitor size of volumes and snapshots, delete them as your policy 5) periodic zfs and btrfs scrub End of fucking story. No need for all that tricky shit y’all are doing.

Why Whisper and medical speech APIs are making potentially fatal errors

Appen (a global AI training data and annotation provider) just open-sourced MedTerm-90, a new benchmark designed to evaluate Speech-to-Text (STT) systems in medical environments. Instead of relying on traditional Word Error Rate (WER)—which treats a missing comma the same as a mutated drug dosage—this benchmark introduces Clinical Entity Recall to prioritize patient safety.They tested 7 major speech APIs (including OpenAI Whisper and ElevenLabs Scribe v2) across 90.5 hours of real-world physician audio spanning 14 medical specialties. The results show a massive gap between standard accuracy scores and actual clinical safety.

Astra for Law

"Rogue OpenAI agent swarm accidentally overturns the Civil Rights Act" followed a month later by "Anthropic's Claude inadvertently repeals the 19th amendment"

The Painful Truth: The RAM Crisis Is Only Just the Beginning

The hardware savings at the scale Anthropic or Google use would me immense, it makes me wonder why no big player has done more optimization already. When DeepSeek showed how inneficient were the models of its time, I'd expected each to create a permanent optimization team with all the talent they have hired. I guess hardware is not that expensive to them in the grand scheme of things, at least not at his stage. OTOH, their propietary models might be thoughly optimized and we can't know, because they're still bound by supply contracts to buy the same amount of hardware nevertheless.

Astra for Law

Enterprise is eventually get caught (if it isn't already) by the Microsoft/Google. Because with office/teams/suite they were already in every enterprise, and that just added a new tool to existing offerings. Provisioning and contracts and data retention was just an extension to review of existing ones. Nobody serious is going to risk sending sensible data to OpenAI/Anthropic, etc because "the benchmarks have shown +8% performance there and +2% there". Irrelevant.

Astra for Law

I said this before but I wonder if Dan Kan will reboot Atrium. Rally up some old partners and hope Anthropic buys them out for a couple billion.

Cohere launches confidential computing in model vault

Astra for Law

They have a lot of compute, so I think it makes sense to spray and see what works. Anthropic is limited in that regard, and focused in coding/tech, but OpenAI don't need to do the same. They even got the lead without having to focus in coding only, which is remarkable.

Astra for Law

Most HNers are clueless that if you have Top Talent + Capital you already have an insurmountable moat. OpenAI, SpaceX, Anthropic all have that and none of the regular guys can compete against them (if they choose to attack that industry)

Making a Software Stack Agentic

Over the last couple years, I've seen the greatest improvements in my agentic coding not from new models or updated tools provided by companies like Anthropic or Cursor, but from direct integrations with and improvements to my own stack: the deterministic scripts and libraries that harness these agents for the specific technologies and products I work on. This kind of stack deserves its own name: an “agentic software stack”, which is an opinionated stack that includes non-agentic tooling which guides both agents and engineers. In order for models to run for a long time with less monitoring and good results, a system with guardrails, guidance, norms, and transparency is essential.

"SpudCell will replicate continuously by 2030" – Interview with Kate Adamala

AI safety is mostly a sex cult

The currently most upvoted comment is a much bigger article containing vritulaly the same claims, and naming some names. It spefically Mentions Eliezer Yudkowsky as living in shared housing with others. It also mentions that Dario Amodei and Holden Karnofski used to live together, sometime before Karnofski married Dario's sister, Daniela, the other co-founder of Anthropic. Separately, SBF was also known to live in communal housing with many of the other chief executives of FTX. Separately, that other article also mentions harmful psychological practices like "debugging" and sleep deprivation. Note that I make no claims of my own - for all I know both articles could be complete fabrications. I'm merely listing what these articles claim. I personally know none of these people, and live half…

How GLM built its own inference infrastructure

Replying to everyone to hack HackerNews' Gish Gallop feature ( nulla poena sine lege! ): Alignment is meaningless; as you've noticed, humans aren't all that "morally aligned". Moral relativism is an attractive proposition when you first examine the topic, but it quickly falls apart; there's a reason it's not even a coherent camp in contemporary philosophy beyond some vagueities from radical post-modernists. Just to go over some of the greatest hits: - Is what [DICTATOR/MURDERER/CRIMINAL] bad, or merely not to your taste? If the latter, then you have no coherent reason to argue they should be punished. We would never imprison people who don't like vanilla ice cream because 51% of the population does like it. - If another culture had a deeply held belief to [HORRIBLE_THING] to, sa…