All Stories — Page 122

The YC-backed startup says users can re-rank 25 million videos and purchase selected clips for $3 per hour.

Tests show the refusal has expanded from living writers to dead authors, while preserving prompts based on atmosphere, pacing and genre.

The Chicago founders pivoted from plastic waste and are building a containerized pilot at a wastewater plant that is expected to go online in 2027.

Repeat AI-infrastructure founders Tom Hamer and Jesse Clark are betting parallel decoding can move stronger models into voice and agent workflows.

Unit 8200 alumni Jonathan Jacobi and Gal Niv are using public interactions to test interfaces and collect training data for robot-agnostic AI.

On Deck co-founder Weisser is applying his fellowship experience to a three-month San Francisco cohort with mentorship and a $100,000 grant for each selected entrepreneur.

The YC-backed system turns prompts and camera feeds into editable programs, with the founders claiming a working policy can be built in 20 minutes.

A July 27th video posted to X shows Jensen Huang saying Anthropic's Mythos model should be available to everyone after an interviewer raises its purported ability to breach a bank.

Paradigm and Caffeinated co-led $370M in equity, while $100M in debt backs a 2027 power test and targeted 2028 deployments.

Feyn Labs expanded BiRefNet, trained FeyNoBg on 26.1K images, and published the weights to support its specialist-model pitch.

The new model claims a 96% CyberGym score and half‑the‑cost operation, and debuts alongside the Perception agentic security system.

The startup says a retrospective ulcerative colitis analysis predicted treatment response at 0.76 AUROC and could have maintained statistical power with roughly 450 fewer patients.

AMD is evaluating the agents, while Synopsys reports early debug cycle-time reductions of 25% to 40% without disclosing benchmark details.

Brett Adcock announced on X that Figure's autonomous robots now carry visitor badges that let them open any campus door.

The Bitcoin veteran who first joined Coinbase in 2014 when the platform ran on Heroku now leads its technology organization.

The release includes checkpoints and a technical paper that claim 2.5x more intelligence per compute unit and native visual understanding.

The reported 90.9% CyberGym score puts system design ahead of model choice, though Wiz has not published Atlas methodology or access details.

The undisclosed investment and strategic partnership give Safe Superintelligence an order-of-magnitude compute increase after Nvidia reviewed the lab's closely guarded research.

Vercel Labs' compiler uses static lowering by default, embeds QuickJS only through an opt-in mode, and rejects code it cannot safely translate.

In a July 26th X thread, ex-Anthropic staffer Adi Baradwaj claims the company traded safety for revenue and notes that most black‑hat hackers use standard Claude Code subscriptions.

The reported guarantee would make Nvidia a financing pillar for a SoftBank-built data center campus that OpenAI would lease.

The post says Z.ai may skip GLM-5.3 and launch GLM-5.5 in August, but it provides no benchmarks, model details, or official confirmation.

On July 26th 2026, the founder warned that apps, docs, dashboards built with Anthropic's Claude are appearing in public search results, raising privacy concerns.

Soar AI says its Model Context Protocol interface gives AI agents access to market-rate flights, no-fee booking, and x402 payments.