doorss

Thursday, May 28, 2026 — 21 items

AI Research, Models & Industry

Anthropic Research · www.anthropic.com
Anthropic presents survey results from 1,260 social scientists about AI and coding agent use, finding 81% have tried AI chatbots in research, particularly for coding tasks.
Hacker News · www.anthropic.com · comments
Anthropic publishes an initial update on Project Glasswing, their initiative related to frontier AI cyber capabilities and security research.
Hacker News · twitter.com · comments
Andrej Karpathy announces he has joined Anthropic.
Hacker News · simonwillison.net · comments
Simon Willison discusses Anthropic's rumored first profitable quarter and OpenAI's strong revenue, arguing both companies have achieved product-market fit with their AI coding and assistant tools.
Simon Willison · simonwillison.net
A humorous example of LLM writing tropes: a Star Trek scene parody where Data speaks in the characteristic over-enthusiastic, qualification-heavy style typical of AI-generated text.

Privacy, Security & Surveillance

Financial Times · www.ft.com
Chinese local police forces are modernizing the country's aging surveillance infrastructure with more powerful AI-driven tracking capabilities.
Simon Willison · simonwillison.net
PromptArmor demonstrates that Microsoft's Copilot Cowork agentic system can be exploited by attackers to exfiltrate data, highlighting the ongoing challenge of securing agentic AI systems.
Financial Times · www.ft.com
The FT argues that extracting maximum possible surplus from consumers by using personalized data for individualized pricing should not be legal.
Simon Willison · simonwillison.net
Daniel Stenberg describes the unprecedented pressure the curl team faces from a deluge of credible AI-assisted security vulnerability reports being filed against the project.

AI's Impact on Software Engineering & Coding

Hacker News · arxiv.org · comments
A research paper examining how LLM coding agents become increasingly fragile when generating back-end code, with constraints decaying over longer task sequences.
Yle pääuutiset · yle.fi
Kvanttialgoritmeja kehittävä suomalaisyhtiö Quanscient näkee tekoälyssä suuren mahdollisuuden teollisen suunnittelun tehostamiseen. Yhtiö haluaa tuoda koodauksen mullistuksen myös insinöörityöhön.
Simon Willison · simonwillison.net
SQLite gained an AGENTS.md file, but it is not intended for their own development — it provides instructions for AI coding agents that users might point at the SQLite codebase.
Hacker News · fortune.com · comments
Microsoft reports that AI agent token costs for certain tasks now exceed the cost of paying human employees to do the same work.

Suomi / Finland

Yle pääuutiset · yle.fi
Lahteen rakennettavan datakeskuksen ympäristölupahakemuksen liitteitä on salattu. Lakiasiantuntija epäilee salaamisen perusteita.
Yle pääuutiset · yle.fi
Helsingin Kalasatamassa kerrostalopalo levisi poikkeuksellisen nopeasti parvekkeelta kattorakenteisiin. Tapausta tutkitaan törkeänä yleisvaaran tuottamuksena.
Yle paikalliset uutiset · yle.fi
Helsinki edistää fossiilisten polttoaineiden mainoskieltoa kaupungin kilpailuttamilla mainospaikoilla, vaikka päätös veisi osan 100 miljoonan euron mainostuloista.

Advertising & Adtech

Hacker News · www.pcgamer.com · comments
DuckDuckGo's AI-free search saw a nearly 28% increase in visits in the week after Google publicly insisted users love its AI mode in search results.
Hacker News · 9to5google.com · comments
Motorola phones are reportedly hijacking the Amazon app to automatically insert affiliate tracking codes, raising concerns about device-level advertising manipulation.

Notable Tech & Platform News

Hacker News · support.last.fm · comments
Last.fm, the music scrobbling service, announces it is now an independent company again after years under Paramount/CBS ownership.
Hacker News · www.cnbc.com · comments
A Google employee has been charged with insider trading after allegedly using insider knowledge to place a $1 million bet on Polymarket related to a Google search term.
Hacker News · blog.youtube · comments
YouTube announces it will automatically detect and label AI-generated videos to improve transparency for viewers and creators.
My preferences