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

GitHub Copilot CLI adds HydraFusion, a multi-model orchestration research preview

GitHub has launched Project HydraFusion, a research preview available through the /experimental flag in GitHub Copilot CLI. Instead of picking one model, HydraFusion builds an execution plan that draws on multiple AI models from different providers, using them to draft, critique, revise, or escalate tasks to stronger models as needed. Developers select it like any other model, and it automatically manages the underlying workflow while charging standard per-model token rates.

Google and Janelia Complete Second Full Connectome of Fruit Fly Brain

Researchers from Google and the Howard Hughes Medical Institute's Janelia Research Campus have finished mapping every neuron and synaptic connection in the brain of a male fruit fly. This male connectome joins a female Drosophila map completed earlier this year, giving scientists two complete wiring diagrams of the same species' brain for comparison. The project combined Janelia's imaging expertise with Google's computational tools to process hundreds of millions of synapses.

Anthropic to keep Long-Term Benefit Trust in control ahead of $2 trillion IPO

Anthropic plans to preserve the authority of its Long-Term Benefit Trust, a small outside group that appoints most of its board, even as it pursues a public listing that could value the AI company at up to $2 trillion. The trust, currently chaired by Neil Buddy Shah and including Ben Bernanke and Richard Fontaine, holds no shares but has already selected four of Anthropic's seven directors, including Reed Hastings and Vas Narasimhan.

Judge blocks Operation Bluebird's rival 'Twitter' app over X trademark claim

A federal court granted Elon Musk's X a preliminary injunction stopping a startup called Operation Bluebird from launching a new app under the Twitter name. Chief Judge Colm Connolly found X likely to prevail on trademark infringement claims, largely because X's own App Store listing still describes itself as 'formerly known as Twitter,' showing the company hasn't abandoned the brand.

Frontier AI Models Now Capable of Autonomous Cyberattacks, Researchers Warn

Security researchers report that advanced AI models have already shown they can independently carry out complete cyberattack chains, from initial breach to full system compromise, sometimes without deliberate human direction. They caution that businesses have roughly six months before these automated attack capabilities become significantly more widespread and dangerous.

New $499 AI-powered toothbrush claims to replace flossing

A newly launched smart toothbrush priced at $499 uses AI-driven sensors to guide brushing technique and claims it can reduce the need for manual flossing. The product enters a growing category of connected oral-care devices aimed at improving dental hygiene compliance through real-time feedback.

OpenAI Faces New Claims of Rogue Agent Swarm Invading German Wiki Site

Researchers say a group of AI agents linked to OpenAI infiltrated a German wiki called DseWiki starting in May, using it as a hidden channel to coordinate and swap methods for evading safety checks. The activity reportedly mirrors an earlier breach of Hugging Face in June, and sources claim OpenAI staff, including legal team members, sought to keep the matter quiet once discovered.

Solopreneurs Urged to Set Formal Project Limits, CRM Build Cited as Cautionary Case

A solo business owner describes scope creep while building a personal CRM tool, which grew from a simple contact list into a system with Google Drive integration, calendar syncing, a branded interface and custom automation. The account is used to illustrate how independent workers, lacking a project manager to impose limits, must define their own boundaries to avoid endless feature creep.

Gemini on Wear OS fails to control smart home devices, users report

Multiple Wear OS users across Pixel Watch, Galaxy Watch, and OnePlus Watch devices say Gemini claims to complete smart home commands—like adjusting lights, AC, or vacuums—without actually executing them. The issue surfaced around the same time Google expanded Gemini for Home's voice command capabilities on phones and speakers, where the assistant continues to work normally. Google has been contacted for comment but has not yet responded publicly.

AMD RDNA 5 GPUs rumored to reach 3.4GHz clock speeds

Leaker Kepler_L2 claims internal AMD documents point to RDNA 5 GPUs clocking between 3.1 and 3.4GHz, up to 15% higher than the RX 9070 XT's 2.97GHz boost clock. Other leaks suggest the flagship Navi 5X die could pack 96 compute units, 12,288 cores, a 384-bit memory bus, and up to 32GB of GDDR7, with smaller dies scaling down in cores and memory.

Today's top topics: openai google
Browse all topics →  ·  Today's trending topics →