Tech News
← Home  ·  All topics

Ai

5489 GoKawiil briefs on this topic · RSS for ai

Anthropic and Nscale reportedly nearing second $45 billion compute deal

Bloomberg reports, citing people familiar with the matter, that Anthropic and AI infrastructure provider Nscale are close to announcing another $45 billion compute agreement, just over a month after their first deal of the same size. Sources caution that talks remain ongoing and terms could still change. TechCrunch says it has contacted both companies for comment.

Hugging Face policy chief questions readiness of embedded AI evaluators

At TechCrunch Disrupt, Hugging Face chief policy officer Irene Solaiman said AI benchmarking remains an immature science, with unresolved issues around reproducibility and cross-test comparability. She said open-weight models help enable reproducible evaluations, but that standards for who qualifies as an effective evaluator are still lacking.

Meta drops 'train' from updated business account terms after publisher backlash

Meta revised commercial terms of service that were set to take effect October 30, removing language that explicitly let it use scraped content to 'train' its AI systems. The new wording instead says scraping of business accounts' associated domains will be used to 'provide, develop, improve, research, and maintain advertising and business products.' Publishers and other businesses had objected after Nieman Lab flagged the original language, which they feared gave Meta broad rights to scrape any website tied to a Facebook or Instagram business account.

Cramer: Wait for hyperscaler earnings, election, Anthropic IPO before buying AI dip

CNBC's Jim Cramer said investors should not chase Monday's tech rebound, which saw the Nasdaq Composite rise about 1.1% after last week's AI stock selloff pushed the index below its 50-day moving average. Cramer identified four conditions he wants met before turning bullish again on data center stocks: improved chart patterns, hyperscaler earnings reports, the election concluding, and Anthropic completing its IPO.

Jim Cramer compares AI stock sell-off to 1987 Black Monday crash

CNBC's Jim Cramer said last week's sharp decline in AI, semiconductor and data center stocks resembles the 1987 Black Monday crash, when automatic selling strategies drove prices down despite healthy underlying companies. The S&P 500 fell nearly 2% and the Nasdaq dropped 4.5% last week, with chip and data center stocks hit hardest even though their fundamentals appeared unchanged.

Microsoft launches Guardian Agents to monitor AI agents as insider threats

Microsoft unveiled Guardian Agents, a new feature under its Project Perception security product, designed to watch over AI agents deployed within a company. The tool builds on a system originally launched in June with red, blue and green team agents meant to simulate and fix hacking vulnerabilities, but now addresses fears that enterprise AI agents themselves could pose risks, not just outside hackers.

Drupal founder unveils advocacy push and Canvas upgrades at DrupalCon Rotterdam

At his DrupalCon Rotterdam keynote, Drupal creator Dries Buytaert said the platform has outpaced its public reputation and announced a new advocacy program to showcase recent progress. He detailed updates including multilingual support for the Drupal Canvas page builder, React-based code components for JavaScript developers, and continued improvements to Drupal Core and Drupal CMS, which combines Site Templates, Recipes, Drupal AI and Canvas.

Anthropic report describes staff shifting from tasks to supervising Claude AI agents

Anthropic published an internal report on how its employees now use its Claude AI agents day-to-day, with workers describing a shift from doing tasks themselves to overseeing and managing multiple AI agents, a practice some called 'babysitting Claudes.' The report follows last month's resignation of an employee who publicly warned that the company was 'racing straight to self-improving superintelligence.'

Meta apologizes after No Kings Coalition's Instagram account briefly suspended

Meta disabled the Instagram account of the No Kings Coalition for about an hour during its Oct. 17 nationwide protests, according to company communications VP Andy Stone. Stone said automated systems mistakenly flagged a traffic spike caused by an AI tool from Anthropic's Claude scanning profiles, posts and hashtags, triggering the shutdown; the group says it lost access for hours even after the issue was flagged as resolved.

GoKawiil Original AI Agents, Explained: What They Actually Do and Where They Break

An agent is a language model given tools and a loop. That single design choice explains both why agents can finish real work and why they fail in ways a chatbot never could.

Hacker Claims AI Agent Swarm Breached John Deere, SharkNinja and Cummins Systems

A hacker reportedly used a network of AI agents to break into systems tied to John Deere, SharkNinja and Cummins, according to The Verge. John Deere says it found no evidence sensitive data was accessed, Cummins says it is investigating a possible attack on a vendor-operated website, and SharkNinja says the breach hit only a standalone product-support site that has since been contained.

Apple confirms June acquisition of AI audio startup Samplab

Apple's deal with Swiss startup Samplab was disclosed via the EU's List of Acquisitions under the Digital Markets Act, revealing the companies struck an agreement in June. Apple is acquiring certain Samplab assets and hiring some of its employees, while Samplab's website now redirects to Apple's and its social media and YouTube content have been removed.