doorss

Anthropic published a policy paper outlining two scenarios for US-China AI competition, arguing it is essential that the US and its allies maintain a lead over authoritarian governments like the CCP in AI development.

The Anthropic Institute shares its research agenda, focused on investigating AI's real-world impact from within a frontier lab and sharing findings publicly.

A survey of 81,000 Claude users finds that people in roles more exposed to AI have greater concerns about AI-driven job displacement, with concerns also higher among younger workers and those in lower-income roles.

The FT argues that whether AI can perform a task is only a small part of the employment picture, and the narrative around AI replacing jobs overlooks many practical factors.

An essay argues that access to frontier AI models will increasingly be gated by economic costs and national security considerations, limiting who can use the most powerful systems.

Simon Willison reflects on Mitchell Hashimoto's observation that programming languages are becoming less of a lock-in factor, as AI makes cross-language porting increasingly feasible.

The Rust project is formalizing a policy on LLM-generated contributions to the Rust compiler, addressing quality and attribution concerns around AI-assisted code.

arXiv has introduced a policy imposing a one-year submission ban on authors who include AI-hallucinated references in their papers.

An Ontario audit found that AI-powered medical note-taking tools used by doctors routinely introduce factual errors into patient records.

Research demonstrates that Mullvad VPN exit IP addresses can serve as a fingerprinting vector, potentially undermining the privacy protections users expect from the service.

A detailed writeup on physically removing the cellular modem and GPS module from a Toyota RAV4 to prevent the vehicle from transmitting location and telemetry data.

Metan tekoälylaseilla on kuvattu naisia salaa, ja huolet lasien väärinkäytöstä kasvavat. Myös Metan työntekijöille päätyy arkaluontoista kuvamateriaalia.

Bluesky published technical details on how its "For You" algorithmic feed is served, highlighting the platform's unique approach where anyone can run their own custom feed implementation.

Bluesky experienced a sustained DDoS attack lasting nearly a full day, causing service disruptions for users of the decentralized social media platform.

Uudellamaalla annettiin vaaratiedote drooniuhasta. Sisäministeriön tietojen mukaan Suomeen harhautui ainakin yksi drooni, mutta vaara todettiin ohitetuksi.

Norjalainen Blastr Green Steel peruu väliaikaisesti massiivisen terästehtaan suunnitelmat Inkoossa.

Etla arvioi Iranin sodan taloudellisen vaikutuksen Suomeen jäävän pieneksi. Teollisuuden lisäksi kasvua odotetaan myös rakentamiseen ja palvelualoille.

TV advertising is gaining renewed value as streaming platforms like Netflix and Disney+ introduce ad-supported tiers, looking for new revenue sources beyond subscriptions.

A leaked deck reveals that StackAdapt, an OpenAI advertising partner, is selling ChatGPT ad placements targeted based on the relevance of user prompts.