Updated daily401 verified stories
Material AI papers, studies, datasets, evaluations, and scientific advances with direct links to the underlying research.
Every story links to the strongest available evidence: original sources when available, otherwise clearly attributed reporting.
What happened, why it matters, and what to watch — no jargon tax.
When the signal is thin, we publish nothing rather than padding the feed.
A growing stream of verified perspectives for people who need to understand AI without chasing hype.
An arXiv preprint introduces Consilience, a framework that selects and certifies how language-model agents communicate when each holds only part of the evidence. The authors report higher decision accuracy and greater communication efficiency than fixed or unstructured discussion protocols on hidden-profile tasks.
A new arXiv preprint argues that enterprise finance systems using large language models should be judged not only by answer accuracy but also by whether every claim can be traced to authoritative evidence. Its proposed framework improved citation traceability in a 145-question evaluation, while producing…
A new survey proposes a common framework for studying AI agents that act through terminals, arguing that model quality alone does not determine their behavior.
An arXiv preprint introduces VortexChat, a framework that uses a large language model to coordinate photonic-device generation, refinement and electromagnetic simulation. The authors report that it autonomously designed and helped fabricate a terahertz vortex-beam multiplexer whose measured performance matched…
A new preprint proposes a hierarchical memory system that decides which parts of an AI agent’s execution history should remain active. On GAIA-Text, the authors report higher accuracy, lower prompt-token use and less persistence of poisoned information than linear memory.
An arXiv preprint presents SAGE, a framework that organizes AI-powered SQL operations into three relational types and reports a 358-fold cost reduction for one factorable join workload.
An arXiv study of AI systems that choose whether to act or stay silent reports a trade-off between stronger decisions and inspectable reasoning, and warns that common faithfulness tests may overstate what reasoning traces reveal.
A University of Nottingham study is testing an AI app that analyzes speech, facial expressions and movement to flag possible depression during or after pregnancy. The BBC reports that the trial is in its final stages, but the system is intended to support clinicians, not diagnose patients.
South Korea’s Ministry of Science and ICT has begun a six-year, 20 billion won program to develop AI systems that use scientific laws alongside data. DongA Science reports that four university-led projects will focus on causal modeling, physical-law interpretation, long-context architectures and more efficient AI…
NPR reports that researchers used AI trained on DNA sequences to design new viruses that infect bacteria. About 16 of roughly 300 tested designs produced functioning viruses, while experts warned that more accessible biological design tools could create new safety and governance challenges.
NVIDIA says a preview feature in its Dynamo inference platform can restore an LLM worker in seconds by keeping a preinitialized standby engine on the same GPUs and sharing model weights in memory. In the company’s benchmark, recovery took 7.3 seconds instead of 283 seconds after a worker failure.
Glitchwire reports that Figure AI has brought Index out of stealth after collecting 16 million user-uploaded videos across 108 countries. Figure says the platform has 264,000 downloads, 43,000 weekly active contributors and has paid contributors $15 million, although the supplied report does not independently…
One useful briefing each week
Get the week’s verified AI news, original data, useful tools, learning picks, and fresh AI jobs.
Hiring an AI professional or launching a useful AI product? Put it in front of people who came here to learn and act.
Post an AI job Submit an AI tool