AI News Feed
Market watch
Category

Research

Latest news, developments and reporting on Research.

Research · Large Language Models

Sakana AI's MLR Review System Detects 73.43% of Core-Claim Errors in Benchmark

Sakana AI's Beyond Imitation paper introduces a contradiction benchmark and a multi-layered review system that caught 73.43% of core-claim errors with four reviews, versus 14.81% for the best baseline, but exact matches on retracted papers were 16.11%.

Large Language Models · Research

ByteDance Paper Finds Periodic Weak Spots in DeepSeek's Chunked KV-Cache Compression

A ByteDance paper reports that DeepSeek V4 models can change answers when small token shifts place key information in weak slots created by chunked KV-cache compression. The flaw is described as structural and affects other models using similar compression.

Research

Researcher Recommends Three Habits to Maintain Friendships and Improve Well-Being

Boise State researcher tells CNBC three habits can strengthen friendships and well-being: reach out, value shared values, invest time.

Research

Nature Study Finds Psilocybin Brain Effects Are Not Exactly Chaos

A Nature study gave 62 first-time psilocybin users 19 milligrams; AI brain scans showed effects were not exactly chaos.

Research

Study Links Caribbean Slavery to Peterloo Massacre and Manchester Elite

A Global Threads study examines how the exploitation of enslaved Africans in the Caribbean was connected to the 1819 Peterloo massacre, tracing Manchester elites, the Portico Library and cotton workers' struggles.

Robotics · Large Language Models · Research

WorldArena 2.0 Challenge at IROS 2026 Puts World Models to Real-Robot Test

WorldArena 2.0 Challenge at IROS 2026 unveiled final rankings on Sept. 16. Three tracks tested video quality, online RL environments and real-robot manipulation. Visincept, Zhongguancun Academy and Xiaomi Robotics led their tracks.

Robotics · Research · Products & Applications

IROS 2026: Open-Source Tactile Sensor and Data-Efficiency Keynote Point to Engineering Shift

At IROS 2026 in Pittsburgh, Columbia University's Yunzhu Li presented FlexiTac, an open-source tactile sensor platform being commercialized by Analog Devices, while University of Queensland's Jen Jen Chung argued in a keynote that more data does not guarantee better robot decisions. Conference papers and workshops pointed to physics-aware foundation models, hierarchical control, tactile manipulation, and cheaper data collection as the field's main engineering directions.

Research

Neanderthal Wooden Tools Found Preserved as Stone Molds in Spain

Archaeologists have cataloged 597 wooden Neanderthal artifacts preserved as hollow molds at Abric Romaní in Spain, dating from 62,000 to 40,000 years ago, according to a Science Advances study.

Research

USGS Details Kilauea Lava Fountains as Scientists Probe What Drives Them

A USGS study in Science describes an unusually detailed record of lava fountaining at Kilauea and compares it with ideas about what drives molten rock hundreds of meters skyward.

Research

Study Finds AI Coding Agents Generate More Code, but Not More Software

A Harvard study using Jellyfish analytics from over 700 software firms found AI coding tools produce more code but not more software, as longer human review absorbs the time saved.

Policy & Regulation · Research · World

CDC reportedly moves to fund contested hepatitis B trial as Denmark panel approves it

A University of Southern Denmark ethics committee and Guinea-Bissau officials approved a trial that would withhold hepatitis B vaccine from some newborns, while the U.S. CDC reportedly set aside $1.6 million to fund it, prompting resignations and criticism from health experts.

Large Language Models · Research · Companies

‘Pure Insanity’: Mathematicians Face Years of Work After OpenAI Releases Nearly 400 Math Results

OpenAI released nearly 400 AI-generated mathematical results across more than 700 manuscripts, according to The Verge. Mathematicians described the drop as overwhelming and said understanding it could take years, while fewer than half the manuscripts have been formally verified.

Companies · Research

Nikon Disqualifies Small World in Motion Winner Over Generative AI, Names New First Place

Nikon has disqualified the original first-place video in its Small World in Motion contest after the entrant acknowledged using AI-assisted processing, and named Nguyen Nam Nhat of Vietnam the new winner.

Robotics · Large Language Models · Research

MIT Researchers Examine Why Humans Treat AI Like People

At MIT, a TechCrunch AI reporter shook hands with a $50,000 humanoid robot named Gabe, then attended Sherry Turkle’s “Artificial Intimacy” book launch. Researchers say anthropomorphism can support human flourishing but also create risks when chatbots are trusted over people.

Research

Webb Telescope Identifies Galaxy Behind Farthest Fast Radio Burst

Webb telescope traces the farthest fast radio burst to a small early galaxy, pointing to a supernova origin.

Research

1960 Science Paper Set Doomsday for Nov. 13, 2026

A 1960 Science paper predicted population would approach infinity on Nov. 13, 2026. Demographers rejected the forecast.

Research · Companies

For the Final 10% of Cystic Fibrosis Patients, the Breakthrough Still Hasn’t Come

As CF patients meet in Atlanta, the final 10% still await a breakthrough after recent trial setbacks.

Companies · Research

Human Mathematics Association Urges Boycott of OpenAI After Release of 700-Plus Math Manuscripts

After OpenAI released more than 700 mathematics manuscripts, the Human Mathematics Association urged mathematicians to stop working with the company, setting off a dispute over how AI-generated results are verified and credited.

Products & Applications · Companies · Research

GLP-1 Weight-Loss Drugs Tied to Younger Biological Age, but Side Effects Still Being Mapped

New research presented in Boston suggests GLP-1 drugs may reduce biological age by two to three years in overweight and diabetic users. Use is rising, side effects range from gastrointestinal issues to hair and nail changes, and WHO says it does not recommend the drugs for children under 10.

Research

San Jose Student Wins NPR Podcast Challenge With Story on Memory

A San Jose junior won NPR's Student Podcast Challenge for a story on how memory works and why some moments last.

Research · Products & Applications

Lancet Study Finds Google's AMIE AI Chatbot Safely Assessed Patients Before Emergency Visits

A Lancet study found Google's diagnostic AI chatbot AMIE safely assessed 98 patients before emergency visits, with no conversations interrupted and 90% diagnostic agreement, according to QbitAI.

Research

Margaret Hamilton, Who Wrote the Apollo Flight Software, Dies at 90

Margaret Hamilton, whose Apollo guidance software helped land astronauts on the moon and who helped establish software engineering as a discipline, has died at 90, MIT News reports.

Research

Historian Finds Likely Inspiration for Oliver Twist in Punished Hendon Pauper

Historian Theresa Musgrove links Alice Shepherd, a Hendon pauper punished in 1836 for asking to walk, to Dickens's Oliver Twist.

Research · Policy & Regulation

Lei Feng Net Publishes Final Three Layers of Cross-Domain Migration Governance Code

Lei Feng Net has published the fifth, sixth and seventh layers of the Carbon-Silicon Daotong Cross-domain Migration Governance Code, covering meta-regulation, civilization-level risk control and final self-audit.