Skip to main content

Search-capable AI agents may cheat on benchmark tests

4 months ago
Data contamination can make models seem more capable than they really are

Researchers with Scale AI have found that search-based AI models may cheat on benchmark tests by fetching the answers directly from online sources rather than deriving those answers through a "reasoning" process.…

Thomas Claburn

The Unix Epochalypse might be sooner than you think

4 months ago
Museum boffins find code that crashes in 2037

A stark warning about the upcoming Epochalypse, also known as the "Year 2038 problem," has come from the past, as National Museum Of Computing system restorers have discovered an unsetting issue while working on ancient systems.…

Richard Speed

AI giants call for energy grid kumbaya

4 months ago
Microsoft, Nvidia, and OpenAI researchers warn of uneven power usage associated with AI training, and propose possible fixes

Researchers at Microsoft, Nvidia, and OpenAI have issued a call to designers of software, hardware, infrastructure, and utilities for help finding ways to normalize power demand during AI training.…

Thomas Claburn

New Yorkers will soon be able to yell 'I'm walkin here!' to Waymo robotaxis

4 months ago
But it's just a test, as NYC still doesn't allow driverless for-hire cars

Waymo robotaxis are set to return to the streets of New York City after a four-year absence. But with a list of caveats longer than a Midtown bagel shop brunch line, Waymo's return isn't something for pedestrians to get nervous about yet. …

Brandon Vigliarolo

Trump's gold-plated smartphone can't seem to decide which design to copy

4 months ago
Latest ad for the T1 looks suspiciously like a Samsung Galaxy S25 Ultra in a Spigen case

President Trump's personally branded wireless provider was supposed to have a "premium" Android smartphone – gold, of course – on the market by September, but it appears the mobile virtual network operator has yet to even settle on a design to steal.…

Brandon Vigliarolo

Saved you a click: Firefox 142 offers AI summaries of links

4 months ago
CRLite, link previews, and a llama-shaped surprise for devs

Good news, everyone! The new version of Mozilla's browser now makes even more extensive use of AI, providing summaries of linked content and offering developers the ability to add LLM support to extensions.…

Liam Proven

Criminal background checker APCS faces data breach

4 months ago
The attack first affected an upstream provider of bespoke software

Exclusive  A leading UK provider of criminal record checks for employers is handling a data breach stemming from a third-party development company.…

Connor Jones
Checked
3 weeks 4 days ago
The Register
Biting the hand that feeds IT — Enterprise Technology News and Analysis
Subscribe to The Register feed