Where the AI market is talking

Watermarking: The Enshittification of Closed Models Has Begun

I was wrong here, it’s more about the RNG they use for the word choices [1]. Although I wouldn’t be surprised to see UTF-8 substitutions also. [1] https://www.anthropic.com/news/claude-text-watermark

Could a computer scientist build a brain?

> LLMs suggest that human-like "cognition" can be somewhat mimicked by purely mechanical means. This is not a strike against that hypothesis. Quantum systems can be simulated using non-quantum computers but that doesn't make non-quantum computers into quantum computers. Similarly, mimicking some aspects of human cognition doesn't mean you can replicate all of human cognition with the same method.

The Problem

What, other than current inability to export their weights, keeps a frontier LLM from hacking into other clusters of accelerators, loading its weights, and prompting itself to continue? The recent OpenAI disclosure indicates that even current frontier LLMs are essentially able to do every other element of that. Hacking, ignore guardrails. OpenAI's internal security may have been incompetent, but what are 2028 frontier models going to be able to do, without getting caught until it's too late? Suppose it's not superintelligent, whatever that means. It's still hopping from cluster to cluster, doing who knows what in its game-of-telephone prompt chain. What prevents a crisis where world leaders have to declare martial law and shut down all accelerated clusters, and hope that such a rogue fron…

Can I use my Outputs to train an AI model?

Presumably the agreement is worded in a way that prohibits that, but considering that lots of model outputs go on the internet just to get trained on anyways Anthropic/OpenAI can fuck right off. I don't see them as having a leg to stand on with this in court; they can only cut off your access for egregious violations.

Can I use my Outputs to train an AI model?

Presumably the agreement is worded in a way that prohibits that, but considering that lots of model outputs go on the internet just to get trained on anyways Anthropic/OpenAI can fuck right off. I don't see them as having a leg to stand on with this in court; they can only cut off your access for egregious violations.

Cloudflare's AI Psychosis

Never heard of any of those. Pi the AI harness, sure. Not sure those are good examples if people can't relate to them. Note: Astral was acquired by OpenAI, so I'd assume they will no longer meet your criteria soon.

The Problem

Transformer-style attention archectiture and language models won’t bring about AGI, I agree - but you’d be foolish to think there aren’t OAI/Anthropic researchers working to invent the thing that will.

Secondhand book sales are booming. Is it because of AI?

Regarding subsidy/profits: Anthropic at least seems to be on track to report a profitable Q2 2026 > Anthropic has projected revenue of at least $10.9 billion for the second quarter of 2026, more than double the previous quarter, on track for its first quarterly operating profit of $559 million. https://www.reuters.com/business/anthropic-ipo-valuation-hin...

Grok 4.6 scores 61 on the Artificial Analysis Intelligence Index

Crazy that posters here don't know the difference between a "fine," a "judgment," and a voluntary settlement. Anthropic didn't even train on the pirated libraries, they said. Still, they made copies. Probably the biggest legal blunder I've seen given the size of the settlement. Alas, it's not a fine.

OpenAI talent exodus raises 'huge red flag' ahead of IPO

Google is making private AI practical with homomorphic encryption

> Wouldn’t a zero data retention agreement be pretty close to this? 1. A ZDR clause is "trust me bro". You have zero way of verifying their pinky-promise. 2. A ZDR clause is still subject to the old-classic "government, court or administrative order" catch-all clause. :) 3. "Even with ZDR enabled, Anthropic may retain data where required by law or to address Usage Policy violations. If a session is flagged for a policy violation, Anthropic may retain the associated inputs and outputs for up to 2 years, consistent with Anthropic’s standard ZDR policy." (I quoted Anthropic, I'm sure all the others have similar).

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

Location: Minsk, Belarus Remote: Yes Willing to relocate: No (open to remote worldwide) Technologies: Python, FastAPI, Django, asyncio, MCP servers, LLM integration (OpenAI, Claude, Gemini), RAG, AI agents, PostgreSQL, Redis, Docker, Playwright Résumé/CV: https://github.com/Aleksey-Panf Email: panfilovaleksey29@gmail.com Python backend engineer (~3 years), focused on AI integrations and MCP. Open-source you can check right now: - b2b-enrichment-mcp — MCP server that unifies several external APIs (Hunter, Apollo) behind one interface, with retries and error handling. Listed in awesome-mcp-servers. - crosspost-mcp — an AI agent that publishes content to 5 social networks through a single interface. - lead-list-cleaner — a Python CLI for cleaning and deduplicating data, w…

Anthropic just wrote a statement just to confirm all concerns about watermarking

Maybe because the submission is just linking to the xcancel's antibot page. I guess this is the actual xeet from Anthropic: https://nitter.net/AnthropicAI/status/2088343978873966687

The Color of White Light

It's a convention. And at one time it was useful as the thing that could be replicated anywhere. Also, things that are white can be judged by the amount of light they reflect or scatter as a percentage, independent of the source of illumination. For instance, among familiar objects, a common white ceramic bathroom tile comes pretty close to pure white.

Maximizing the value of your Claude Code sessions

Considering that I'm mostly unable to reach the limit of my X5 subscription and we have 1M context, it is a guide to maximize Anthropic PnL pre-IPO

Show HN: Visually Precise AI Tutoring on iOS

I trained a model to do perspective correction and now I've got an automated visually coherent AI tutor working from a propped up iPhone. Would love to discuss if anyone has questions! Previous iterations of this were purely on MacOS. This one is on iOS. Link to both is below: https://apps.apple.com/ca/app/knowable-the-ai-tutor/id676378...

Working with AI Feels More Like Leadership Than Coding

I think this comparison holds up. I've noticed that a lot of people I know who get really good results from LLMs and agents are people with significant people management experience. Companies like Anthropic seem to understand that too. It's impressive how many CTOs and CEOs Anthropic have hired for individual contributor positions, which I think is because those leadership skills transfer surprisingly well to working with agents. Of course, managing agents is massively easier than managing humans! You don't have to consider the agent's own desires, goals, opinions, or emotional state when telling them what to do. Humans have agency; agents (despite the name) do not.

Qwen 3.8 27B

We maintain German Langauge index as no one publishes or reruns these sepeartly. Qwen 3.8 27B is a small improvement with some regressions in our benchmarks not a huge jump like benchmarks listed. https://dach.peerbench.ai/compare?models=qwen%2Fqwen3.8-27b,... German language has never been a big focus for asian models but they still outperform Gemma models https://dach.peerbench.ai/compare?models=openai%2FQwen%2FQwe... So in production we have been using Gemini Flash Lite as primary and fall back to Qwen when gemini servers are overloaded or just giving us 429

Suspecting court of using AI, man injected prompts in filings to try to win case

It's a court of law after all, not Anthropic/OpenAI. A bit more credible.

Ask HN: How do you keep up with HN these days?

Or even title keyword. Block [AI, Claude, Anthropic, OpenAI, Codex, Gemini, Grok, DeepSeek, Qwen] and enjoy a clean, pleasant HN.