Skip to content
Tech News
clear
Topics: Today This Week This Month This Year

OpenAI's ChatGPT Work splits into cloud and local versions with distinct capabilities

OpenAI launched ChatGPT Work on July 9th as a paid feature for $20+/month subscribers, and it actually comes in two forms: a cloud-based version accessible via chatgpt.com and mobile apps, and a local version built into the desktop app (formerly Codex) that can access files and run programs on a user's computer. The cloud version, described as the more interesting of the two, adds capabilities not found in standard ChatGPT Chat, including selection among multiple models and reasoning-effort levels.

Multiple Laptop Thefts Hit Exhibitors in Gamescom's Business Area

Several exhibitors at Gamescom in Cologne reported stolen equipment from booths in the show's business zone, including developer Ryan Laley's new Alienware laptop, two laptops from publisher iam8bit, and two laptops plus a Steam Deck from Tessera Studios. The thefts occurred despite the area being restricted to press and industry badge holders, with organizers claiming 24/7 security was in place.

Global timekeepers may scrap the leap second before October vote

Standards body CGPM will vote in mid-October on ending the leap second, the periodic one-second adjustment used since 1972 to keep atomic clock time aligned with Earth's rotation. The move is being accelerated because the next adjustment could, for the first time, require subtracting rather than adding a second—a scenario computer systems have never had to handle. In its place, officials are proposing a leap hour that would only realign timescales once they drift a full 3,600 seconds apart.

Space scientists propose four lab 'archetypes' to guide AI adoption policies

A team of astronomers, led by Sarah Burke-Spolaor at West Virginia University and Michelle Ntampaka at the Space Telescope Science Institute, posted an unreviewed white paper on arXiv outlining four archetypes for research groups to help them decide how to use AI tools like large language models. Rather than proposing a single universal policy, the framework maps each lab type against its research priorities using a radar-style diagram to suggest which AI applications, from code generation to writing assistance, best fit its values.

1990s Super Metroid FAQ writer hand-justified 17,000 words without software

A retro gaming FAQ author named rs1n manually wrote a lengthy Super Metroid guide in the late 1990s so that every line's right margin ended flush against the text width, without using any justification software. He achieved this simply by carefully choosing words and phrasing as he wrote, avoiding the awkward double-spacing or hyphenation typically seen in monospace text justification.

Pope Leo XIV issues encyclical Magnifica Humanitas warning on AI's effects on human dignity

Pope Leo XIV has published an encyclical titled Magnifica Humanitas, accompanied by a video presentation, addressing how artificial intelligence affects the human person. The document, discussed alongside remarks from Cardinal Michael Czerny, frames AI's trajectory as a matter of human choice and institutional responsibility rather than an inevitable outcome of the technology itself. It focuses on three central risks: loss of inner freedom despite tool use, erosion of critical judgment amid abundant data, and diminished real human connection in digital life.

Linux Foundation reports AI scrapers now dominate its git server traffic

The Linux Foundation's IT infrastructure director revealed that AI crawlers hunting for clean, LLM-free training data are hammering the organization's kernel.org git servers, consuming more CPU power than all legitimate human and tool access combined. Because cgit can generate near-infinite valid URLs for commits, diffs, and patches across hundreds of forks, scrapers are redundantly fetching duplicate data at massive scale.

Continuous diffusion language models see renewed research momentum

A wave of new research papers is reviving interest in continuous diffusion techniques for language generation, an approach that had largely been overtaken by discrete diffusion and autoregressive methods in recent years. The author reviews this resurgence, tracing why continuous diffusion for text fell out of favor and why several research groups are now revisiting it as an alternative to token-by-token autoregressive generation.

Chess.com expands beyond chess with poker site Gambit, eyes go and mahjong next

Chess.com quietly launched a free educational poker platform called Gambit in May, letting players learn poker and build an Elo-style rating without real-money stakes. The company says it's now developing similar rating systems for other classic games like go, backgammon, and mahjong, using AI-assisted coding to build these platforms faster with smaller teams.

Today's top topics: facebook cambridge analytica new mexico lawsuit meta openai
Browse all topics →  ·  Today's trending topics →