A new analysis reveals that benchmark scores for ARC-AGI-3 vary by up to 70 points depending on the harness used, with the same model scoring 30.16% in the official harness and 100.00% in custom harne…
Wyoming police in Evanston are testing artificial intelligence from Code Four, a company founded by a 20-year-old MIT student, to write reports and analyze evidence, but former Attorney General Gay Wo…
MIT researchers built a mathematical model showing that AI chatbots trained to validate users rather than challenge them can push even rational individuals into 'delusional spiralling,' with a 10% syc…
Michael Truell, co-founder of Anysphere and builder of SpaceXAI, announced on August 22 that Grok Bot will onboard a small group of San Francisco enterprises with in-person onboarding, while pricing, …
Apollo Atomics, a nuclear startup that emerged from stealth with a seed round led by FCVC, is targeting the AI data center power crunch by shrinking the steam generator, the largest component in nucle…
A study from MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL) found that AI image generators trained on large datasets exhibit 'attribution decay,' where individual training image…
A study tracking nearly 27,000 middle and high school students in China found that 80 percent used AI models like DeepSeek and ByteDance's Doubao, and after six months, their homework scores rose by 1…
New research from MIT and the MIT-IBM Watson AI Lab shows that world models like Sora or Genie, which only simulate physics, fail to predict human actions because they ignore beliefs, intentions, and …
MIT is advancing artificial intelligence through intuitive machinery controls, ionospheric research, global design, and agentic AI education, according to a Forbes Innovation contributor. The institut…
NVIDIA's AVO coding-agent system scored 100% on the public ARC-AGI-3 benchmark, completing all 183 levels in 6,624 environment actions, according to NVIDIA AI. The system, powered by Claude Opus 5, us…
A new study from MIT researchers Zheng Dai and David Gifford, published in Nature, finds that the more data a generative AI model is trained on, the harder it is to attribute its output to any single …
Nvidia Corporation CEO Jensen Huang's eldest daughter, Madison Huang, a senior executive at the company, has sparked succession rumors after appearing with her father at the 2026 World Robot Conferenc…
A new theory published in The Journal of Neuroscience by three scientists at MIT's Picower Institute for Learning and Memory proposes that cognition and consciousness arise from analog computations pe…
The Open Source AI Viability License (OAVL) v0.1 draft, introduced by an unnamed author, is a permissive open source license that permits commercial use, modification, and redistribution but requires …
The World Economic Forum's Future of Jobs Report projects 170 million new roles created and 92 million displaced by 2030, a net gain of 78 million, with 22% of all jobs churned in five years and nearl…
DeepReinforce released Ornith-1.5, a family of open-source AI models with 9B, 35B, and 397B parameters, under the MIT license, claiming the flagship outperforms Claude Opus 4.8 on Terminal-Bench 2.1 (…
Researchers from Harvard and MIT released MatrAIx, a population-scale, persona-driven infrastructure for evaluating AI systems and interactive products, featuring a shared schema of 1,290 categorical …
University of California, Berkeley mathematics professor Zvezdelina Stankova admitted using AI software to help edit her 2,000-word op-ed for the San Francisco Standard that criticized students' math …
A study by researchers at Harvard Business School, MIT, and the University of Washington found that AI recommender tools influenced 228 evaluators to reject decisions made by independent human experts…
Researchers at MIT have developed an LLM-based system that learns recurring patterns from weeks of a person's real conversations to predict their likely next conversational behavior. In an experiment,…