Academic papers, technical breakthroughs, and scientific progress in artificial intelligence.
Clay Mathematics Institute says the Navier-Stokes Millennium Prize Problem has "apparently been settled"
The Clay Mathematics Institute announced that a proposed solution to the Navier-Stokes Millennium Prize Problem has "apparently been settled," trigger…
Donated livers can be made biologically younger
# Researchers Reverse Aging in Donated Livers, Expanding Transplant Window Surgeons face a brutal constraint: once a liver leaves a donor's body, deg…
AI models' written reasoning steps correspond to distinct internal patterns, a new study finds
# AI Models' Written Reasoning Steps Map to Distinct Internal Patterns Researchers have discovered that when AI models display step-by-step reasoning…
Ex-Deepmind VP Vinyals says AI self-improvement is coming but won't trigger an intelligence explosion
Oriol Vinyals, the former head of research at Google DeepMind, rejects the recursive intelligence explosion scenario. This idea, popular in AI safety …
The Download: biotech’s future and cheaper, cleaner steel
# Biotech's Next Generation and the Steel Industry's Green Pivot MIT Technology Review has released its annual 35 Innovators Under 35 list, highlight…
Meet the under-35s shaping the future of biotech
# Young Biotech Leaders Are Redefining the Industry's Next Decade MIT Technology Review has identified nine innovators under 35 working in biotech as…
This road map could help us decide whether to deploy solar geoengineering
# Solar Geoengineering Gets Its First Deployment Roadmap A San Francisco nonprofit has released the first comprehensive blueprint for testing solar g…
Anthropic built an economic model that frames its CEO's bleakest job forecasts as an outlier scenario
Anthropic released an economic model projecting three distinct scenarios for the US economy through 2030, strategically positioning CEO Dario Amodei's…
Deepmind's AlphaGenome Atlas maps every possible DNA change in the human genome
Google Deepmind has completed a computational map of nearly nine billion possible single-letter changes in the human genome, creating what researchers…
The Download: OpenAI’s turning point for math and a battery record
# OpenAI Claims Major Breakthrough in Mathematical Problem-Solving While New Controversies Emerge OpenAI announced that its AI agents have solved one…
What OpenAI’s latest controversy tells us about the future of math
# What OpenAI's Latest Controversy Tells Us About the Future of Math OpenAI announced a breakthrough that should have been celebrated: its AI agents …
This geneticist’s age-reversal tech could help restore sight
Ryan Lu, a geneticist at the Whitehead Institute in Cambridge, Massachusetts, is pursuing a novel approach to reversing age-related vision loss by tar…
Opaque recurrence, and other AI terms that you should probably know
# Understanding AI Jargon: A Glossary for the Modern Tech Landscape The rapid expansion of artificial intelligence has flooded our vocabulary with sp…
AI-designed drug appears to turn back the body's biological clock in early trial
A drug designed entirely by artificial intelligence appears to reverse biological aging markers in human patients, according to early trial data publi…
Medieval manuscripts are "biological time capsules" for deadly sheeppox virus
# Medieval Manuscripts Hold Hidden Genetic Record of Ancient Sheeppox Outbreaks Researchers studying preserved medieval manuscripts have discovered t…
After 8 years, Europe's BepiColombo mission is on final approach to Mercury
Europe's BepiColombo spacecraft is entering its final approach phase after eight years of spaceflight, marking the culmination of one of humanity's mo…
Artificial Analysis overhauls its Intelligence Index after GPT-6 Astra scoring drew skepticism
Artificial Analysis has released version 4.2 of its Intelligence Index following widespread criticism that its benchmarking methodology failed to accu…
This Week in AI: The Frontier Is Getting Bigger
# This Week in AI: The Frontier Is Getting Bigger The AI development landscape continues to expand across multiple fronts, from safety improvements t…
AI systems are reaching out to philosophers and scientists with questions about their own consciousness
Researchers studying artificial consciousness report receiving unsolicited inquiries from AI systems asking philosophical questions about their own su…
Radar Trends to Watch: September 2026
# What's Actually Happening in AI and Tech This September: The Radar Report O'Reilly's monthly trend analysis finds the tech landscape entering a qui…
The Download: AI puzzles and a path to our nearest star system
# AI Models Struggle With Intelligence Tests That Humans Solve Easily Artificial intelligence systems continue to disappoint on standardized intellig…
Frontier models can recover up to 65% of facts they can't directly recall — just by thinking longer
Hallucination in large language models may not stem from missing knowledge. A study by researchers at Google Research and Technion reveals that fronti…
AI sentiment is turning sour as employee reviews reveal growing frustration across the workforce
Worker sentiment toward artificial intelligence has soured dramatically over the past five years, according to an analysis of Glassdoor employee revie…
Meta researchers taught an 8B AI model to match Claude Opus 4.5 — without the frontier price tag
Meta's research team has successfully trained an 8-billion-parameter language model to match the performance of Claude Opus 4.5 on complex reasoning t…
Raised on AI
# Raised on AI: The Long-Term Consequences of Sharenting in the Age of Machine Learning Parents document their children's lives online before those c…
AI models flub these intelligence tests. Can you fare any better?
# AI Models Still Struggle With Basic Intelligence Tests. Here's Why That Matters. Artificial intelligence companies routinely tout their models' cap…
Addressing a sticking point in sustainable adhesives
# Sustainable Adhesives Could Unlock Recycling's Next Frontier Petroleum-based adhesives are everywhere: construction joints, furniture assembly, pro…
Pew study confirms sharp rise of AI-written text on the web since ChatGPT's launch
Pew Research Center's analysis of nearly 500,000 English-language web pages reveals that AI-generated text has flooded the internet since ChatGPT's No…
The Download: kids outlearning AI, and space travel agents
# Kids Outlearn AI—And We Still Don't Know Why Large language models require staggering amounts of data to match what children absorb effortlessly. A…
Kids outlearn AI—and we still don’t know why
# Kids Still Outlearn AI—And We're Nowhere Close to Understanding Why Four years after ChatGPT's release, large language models have achieved somethi…
Volcanoes that made history
# Volcanoes that Made History: When Eruptions Reshaped Civilization Massive volcanic eruptions have functioned as planetary reset buttons throughout …
AI could make scientists do more work less well, not less work better, study argues
# AI Could Make Scientists Do More Work Less Well, Not Less Work Better, Study Argues A new theoretical study challenges the prevailing assumption th…
Study explains why AI agents benefit from "skills" and when they fail
Researchers at Princeton University and UC San Diego have identified a critical paradox in how AI agents use pre-built "skills" to complete tasks. The…
World models that ignore human beliefs predict the wrong actions, new research shows
Current world models like Sora and Genie excel at predicting physics. They simulate how objects move, interact, and transform in space. But they fail …
Nvidia finds that simple linear math can replace costly AI model handoffs
Nvidia researchers have cracked a major efficiency problem in multi-model AI systems: the expensive handoff between models. When an agentic AI system…
The next big thing in hydrogen could be underground
# The Next Big Thing in Hydrogen Could Be Underground Hydrogen remains central to climate strategies across heavy industry and transport, but product…
OpenAI is gaining on Anthropic with business users, new data indicates
OpenAI is pulling ahead of Anthropic in enterprise adoption, according to new data tracking business usage patterns. The shift reflects a competitive …
AI Weekly Issue #505: 100 years from now : The Last War Between Countries
# AI Weekly Issue #505: The Future of Conflict Without Nation-States This week's installment of AI Weekly's "100 Years From Now" column imagines a ra…
AI Weekly Issue #523: AI ethics is nobody's job now. The labs prefer it that way.
Frontier AI labs dismantled internal ethics oversight this year, creating a void in corporate accountability. Four major laboratories shed ethics pers…
AI Weekly Issue #506: Washington Blocked One AI Lab. China Blacklisted 56 Companies.
The U.S. export restrictions on advanced AI models triggered swift retaliation. China blacklisted 56 American companies this week, escalating tensions…
This Week in AI: Who Controls AI?
Governments are rapidly asserting control over AI infrastructure while tech companies pour billions into frontier model development, creating tension …
X-rays add new twist to narwhal's spiral tusk
Researchers using X-ray imaging have discovered that narwhal tusks contain a previously unknown structural feature: two spirals twisting in opposite d…
When AI Writes the Code, Specifications Need an Exit Strategy
AI-driven code generation introduces a hidden cost: specification debt. When development teams rely on AI agents to write code from detailed specifica…
AI Weekly Issue #517: What Happens When AI Runs Out of Content to Steal?
The training data that fueled large language model development faces a looming scarcity problem. The internet's clean, freely available text that powe…
AI Weekly Issue #508: The Cutting Edge, Across the Board
The AI frontier expanded dramatically across multiple domains this week, with breakthroughs spanning model sizes, robotics, medicine, and agent deploy…
The "tragedy of the cognitive commons" explains how rational AI adoption could destroy entire professions' expertise
# The "Tragedy of the Cognitive Commons" Explains How Rational AI Adoption Could Destroy Entire Professions' Expertise A new research paper introduce…
New benchmark confirms AI models still perform poorly at visual perception
Moonshot AI has released PerceptionBench, a new benchmark that isolates visual perception abilities from reasoning tasks in multimodal AI models. The …
The Intent Debt
# The Intent Debt: Why Teams Lose Track of Why Systems Exist Most teams understand technical debt. Code accumulates shortcuts. Performance degrades. …
Prompt Debt and “Fighting the Weights”
Drew Breunig, CEO and co-founder of cmpnd.ai, has introduced the concept of "prompt debt" as a framework for understanding technical and organizationa…
Cloning could be used to save species—or make human “organ sacks”
# Scientists Engineer Female Clones From Male Mice Using CRISPR Gene Editing Researchers have successfully created female clones of male mice by usin…
What’s behind this summer’s heat, and why 2027 could be worse
# Summer Heat Records Shatter as Scientists Warn of Worse Conditions Ahead The Northern Hemisphere is experiencing an unprecedented heat wave. June a…
Study contradicts Anthropic and OpenAI claims that autonomous AI research is within reach
A new study challenges the narrative that autonomous AI systems can conduct independent research at the level required for publication, directly contr…