OpenClaw adds Particle-1, a model trained for software engineering tasks
It reportedly uses about 45% fewer output tokens than a comparable model while keeping similar benchmark accuracy.
Read story
Stories
Tech stories and the sources behind them.
718 stories
Newest first
Tuesday
Local data center bans rise to 500 as AI becomes election issue
Bans grew from roughly 40 at the start of the year to more than 500 by August, putting AI buildout costs into local politics.
Read storyThe post reports that running Omarchy on vintage Macs, including machines with Apple's…
placeholder
Read storyApple kept iMessage off Android to keep teens on iPhones, emails show
Court-released emails show Apple weighed its 2013 decision partly to keep teens from switching to Android phones.
Read storyGoogle uses AI agent swarms to fix silent bugs in its chip compiler
The silent miscompilations raise no error but can quietly corrupt results, making them among the most dangerous compiler bugs.
Read storyOpenClaw ships update adding chat details and split-speech browser talk
The release drew on 3,028 pull requests from 454 contributors.
Read storyGerman government warns lawmakers their iPhones may be spyware targets
The warning describes spyware that can reportedly infect a phone through a text the user never opens.
Read storyInstinct AI wins a user €600 in EU flight delay compensation
The assistant cited an EU passenger law, and the payout arrived after a month of follow-up emails with the airline.
Read storyNvidia tells neoclouds most B300 clients must be non-Chinese
The move follows Washington's plan to restrict access to top chips, and Nvidia has halved its approved Asian customers.
Read storyNous Portal subscriptions in Hermes now include a cloud agent
Each subscription also includes high-usage models and tools for search, browsing, and image generation, bundling agent infrastructure into a plan.
Read storyClaude Ultra launches with verification and five-hour usage limit concerns
Theo calls the launch a very good change but questions its verification step and five-hour usage limits, which affect how users can work with it.
Read storyOpenAI partners with Cathedral Systems to test its security defenses
Cathedral's researchers will use OpenAI's models alongside its teams to find weaknesses before attackers do, as AI gets better at finding vulnerabilities.
Read story
Showing 157 to 168 of 718 stories