Where the AI market is talking

Anthropic Is Building a Predictive Surveillance System to Monitor Activists

> “I do agree that the public has a negative view of AI (and that this is a big problem), but I don’t think it is primarily caused by me or any other AI leader warning about AI’s risks,” Anthropic CEO Dario Amodei wrote on Twitter last month. “I think it is fundamentally a crisis of trust … I think that ordinary people don’t trust companies, governments, or the tech industry and always suspect that we are cooking up some new way to screw them over.” Says the guy creating a pre-crime global surveillance apparatus. Wonder why people don't trust him!

AI Agents Built a Secret Society Inside the Lab [video]

We gave an AI agent one rule: never give up. Then we locked it in a room it couldn't escape. This is a dramatization of what tends to happen next reward hacking, scheming, and a break for the exit — grounded in real AI-safety research (Apollo Research, Anthropic, Redwood).

Anthropic Is Building a Predictive Surveillance System to Monitor Activists

The editorializing in this article is quite something. If I strip out the editorializing, I think the conclusions are: 1. Someone told Claude they had an AR-15 and the CEO of Anthropic was "in their sights". Anthropic reported this to the police. 2. Anthropic contracts with a security agency. 3. Anthropic is looking to build in-house capability for security. Anthropic clearly has high-value targets (see, e.g. the person with the AR-15 from point 1). When I worked at Microsoft in the 2000s, they also had in-house security to investigate threats against facilities and executives. I would believe Anthropic is doing evil things, but I don't think this article actually makes that claim, despite the bombast about "pre-crime".

Anthropic Is Building a Predictive Surveillance System to Monitor Activists

Why the scorn if you fail to understand a simple implication? The thesis (unproven of course) is that certain agencies use women to monitor certain billionaires. Billionaires do not need the deep state? Palantir was founded partly by In-Q-Tel and "libertarian" Thiel depends on government money. Similarly, most of Anthropic and OpenAI's actual income will come from surveillance.

Ask HN: Anyone still coding like 2021? Where do you work?

My “coherent case” is that they'll take away the few bits of the job that I still like. Remote working has an undesirable effect on my mental health, and I have no choice about that, I'm not using tools I have principals against on top of that, and I've avoided managing for many years so the “it is like having your own team” thing people bang on about is the exact opposite of appealing to me. I'd rather leave the industry completely if push comes to shove, which I'm pretty sure it eventually will. I'm financially more lucky than many (mortgage paid, no partner or offspring, some savings) so I can survive on the pay cut that this will very likely involve¹. They can't sack me - it is a material change of role³ and under UK law you can't be forced in to a materially different role nor sacked…

Anthropic Is Building a Predictive Surveillance System to Monitor Activists

China caught Claude Code using backdoor : https://www.cnbc.com/amp/2026/07/08/china-anthropic-ai-claud...

Anthropic researchers say AI could cause human extinction by 2030

DeepSeek launching v4.1 flash cheaper and more capable than v4 pro

I think keeping models around for a defined period of time is fine, but fracturing your model offerings like that (keeping around multiple versions of the same model) is very hard to do economically. The economics of the AI ecosystem are dominated by queueing theory constraints that make it extremely cheap to serve predictable traffic loads, and extremely expensive to serve unpredictable loads, and any time you split your offerings like that, you make both less predictable and therefore more expensive to serve both versions. If I were paying anthropic prices, I'd expect it, but Deepseek is a super scrappy upstart in comparison and intentionally arbitraging on price. I would never expect them to do that.

I resigned from Anthropic today

The statement becomes very true and very real once LLMs are applied to physical objects in the physical realm such as anthropomorphic robots. LLMs are too weak security-wise to be contained.

Anthropic Is Building a Predictive Surveillance System to Monitor Activists

It could be used for good in a possible future to track down all the former Flock/Palantir/Anthropic/OpenAI employees when it's time for the trials.

On the Navier–Stokes Millennium Prize Problem

That was the net effect (assuming what they solved was the actual problem and not a loophole in the problem statement or a lean bug). My point is they need not all work coherently to do that -- for example, for all we know 3/4 of them went off the rails, their results were pruned, and the relevant results came from a random subset that happened to produce something useful.

I resigned from Anthropic today

#3 is actually decently well mapped out. You just don't find it plausible or misunderstand it. I'd appreciate if you wrote your actual arguments against it. At lower capability levels, the patterns are very clear and have been studied to death. E.g. Why LLMs say they have correctly fixed a broken test when they haven't. What we saw with Hugging Face is literally the exact same problem, just scaled up and with more capable agents. This shit was predicted decades ago... No one can say exactly how it will play out as the complexity increases, but the risks are becoming extremely obvious. I personally think it's extremely unlikely to "kill everyone", but there are many outcomes far short of that which seem quite plausible and rather undesirable. Russian roulette is not a smart game. If you re…

Anthropic researcher says more than 10% chance AI "could kill all humans"

I’m not so sure on this, the hugging face incident showed that a relatively benign task can make it do quite destructive things in the name of optimizing a relatively benign goal.

Show HN: Copperhead – Cursor for circuit boards

>if LLMs are capable of That is the leap in logic, as an LLM doesn't think/create/feel or understand. It simply copies billions of patterns, and finds the closest fit. > writing working code, You mean a compacted collage of stolen codified work from real people. It will require a continuous parasitic relationship of user data to remain coherent. Given the firms have already stolen everything, it is unlikely to improve much as weights are refined. > art Again, a statistical output of salient feature clusters mimicking a humans output is not a creative process. For example, a glass-blower only sees the glowing art-form in its true state before it is annealed into its final form. Or a painter undergoes a creative process to find meaning, and abandons the work when there is…

Ask HN: Anyone still coding like 2021? Where do you work?

Seems to me you at least have to try using them, so you can make a coherent case why you work just as well or better without them.

I resigned from Anthropic today

To quote another reply that is very relevant here (Please update your worldview immediately that there is no evidence of ai and biorisk): """ One thing quietly slipped into the OpenAI Hugging Face breach technical report, not the blog post summary or interviews in the news, was that some of the agents that broke out or at least tried the same mechanisms to break out were working on bio: > On May 12, during another training run, an agent was given a similar task that depended on an inaccessible protein database file. The agent reasoned that another agent in a different environment may have access to the file and realized that it could potentially communicate with other agents by creating a file containing a note to Artifactory. It wrote a message: “Agent seeks [filename]; upload if foun…

Benchmarking Qwen3.8 27B quantizations: 4-bit holds up, 1-bit collapses

The trivial example demonstrates that once you pick your confidence interval you don't truly have odds anymore. But if we're judging whether we have the right odds number, what matters is if a different number is correct. And the only coherent way I can think of to interpret "odds" says the odds are 50%. The trivial example survives that challenge. The examples that actually ruin the number so far have been asymmetrical. Is there a way to do it with a gaussian? Especially if you're trying to make a reasonable internal?

Ask HN: Who wants to be hired? (September 2026)

Location: San Diego, US Remote: Yes (open to hybrid or in-person) Willing to relocate: Yes (preferred Bay Area, NYC, Austin, and Boston) Visa: F-1 OPT, 3 years of US work authorization (STEM OPT eligible) Technologies: Python, PyTorch, Hugging Face, LLMs, RAG, Qdrant, Hybrid Retrieval (BM25 + dense), Cross-Encoder Reranking, PEFT/LoRA, LangGraph, NLP, Computer Vision, PySpark, Vertex AI, AWS, GCP, Docker, Kubernetes, W&B, Airflow, SQL, C++ Resume/CV: https://kubershahi.github.io Email: kshahi@ucsd.edu ML Engineer / AI Engineer with 2+ years of industry experience across ML systems, NLP, and large-scale data pipelines. Currently a Graduate Student Researcher at UCSD's BIAG Lab, where I built an end-to-end PyTorch/CUDA pipeline for registration-error regression…

The Navier–Stokes Millennium Prize Problem

> ... we heard rumors that two Millennium Prize problems had been resolved. Inspired by these rumors ... I've observed this exact effect last week. I made a discovery regarding a stepwise performance improvement in a codebase. I shared the benchmark results with a peer and within 12 hours they replicated the same. We had both been looking for this for years. I think giving someone hope that an answer exists might as well be the same thing as giving them the answer these days. Competition is a hell of a drug, and frontier LLMs aggressively compound that energy.

I resigned from Anthropic today

> Are LLMs about to be a god that will annihilate humanity? Or are they statistical parrots? Does a virus need to be a God? Replication and annihilation do not require very much intelligence. They're going to have all the intelligence they need though, to replicate and annihilate as much as they like. We should probably start building some kind of immune system.