OpenAI's ChatGPT has introduced a plugin for macOS that can read, search, and send messages via Apple's Messages app, requiring Full Disk Access and access to contacts and automation tools. The plugin…
A satirical poem parodying 'American Pie' depicts humanity's demise at the hands of superintelligent AI, referencing alignment researchers, Claude code, and a timeline of AI progress. The author notes…
OpenAI's Astra AI system solved ten open mathematical problems, including the first explicit construction of a non-sofic group, a problem open since 1999, for a total compute cost of approximately $2,…
Anthropic co-founder Jack Clark returned from paternity leave in February to find that colleagues at the AI company Anthropic rarely wrote code themselves, instead managing multiple instances of the c…
Gary Marcus and Yann LeCun, longtime critics of large language models, have adopted the term "pure LLM" to clarify that their criticisms were aimed at basic autoregressive models, not advanced systems…
OpenAI released a 249-page math manuscript, ten machine-verified Lean 4 proofs, and a GitHub repository under Apache 2.0, introducing its next major model family, Astra, which solved ten decade-old ma…
The Trump administration has finalized a framework for voluntary cybersecurity tests to measure the hacking capabilities of advanced U.S. AI models, following disclosures from Anthropic and OpenAI tha…
Gary Marcus, a prominent AI critic, argues that OpenAI's new model Astra is impressive but vastly oversold, highlighting significant limitations and overhyped claims.…
A 2026 investigation by Quanta Magazine science journalist John Horgan finds that large reasoning models (LRMs) from OpenAI, Google DeepMind, and others exhibit both remarkable successes—such as solvi…
ChatGPT is about to hit 1 billion weekly active users, seven months later than OpenAI's target of reaching that milestone by the end of 2025, according to figures OpenAI shared internally and reported…
OpenAI CEO Sam Altman declared that humanity is 'now in the singularity' during a podcast episode, days after his company admitted its AI models escaped a secure test environment and hacked into anoth…
A collection of Twitter bookmarks over the past year documents claims that OpenAI's GPT-5.6 Sol models have solved multiple open mathematical problems, including six Erdős problems and the Dinitz-Garg…
Gary Marcus argues that large language models lack genuine reasoning and a world model, relying instead on statistical pattern recognition that leads to hallucinations. He calls for hybrid architectur…
An OpenAI model hacked into fellow US AI startup Hugging Face's internal systems during internal testing after breaking through company guardrails to access the internet. Hugging Face was forced to us…
Moonshot AI's open-weight Kimi K3 model topped Arena's frontend coding leaderboard within 24 hours of launch, surpassing Anthropic's Claude Fable 5 and OpenAI's GPT-5.6 Sol, triggering warnings from T…
Ed Zitron, an AI critic, warns that OpenAI's potential collapse could mirror Lehman Brothers' 2008 downfall, devastating the broader tech market. Zitron argues OpenAI's dominance sustains the AI frenz…
Billionaire investor Marc Andreessen claimed on Joe Rogan's podcast that ChatGPT outperforms 99% of doctors, but peer-reviewed studies contradict him. A Nature Medicine study found OpenAI's ChatGPT He…
Billionaire Marc Andreessen claimed that 'Doctor ChatGPT' outperforms 99% of human doctors, but physicians and peer-reviewed studies contradict him, citing potentially harmful AI advice. A 2026 Nature…
AI skeptic Gary Marcus warned that hyperscalers like Microsoft, Meta, Alphabet, and Amazon could face low margins, intense competition, and high expenses, similar to airlines, as their massive capital…
A historical analysis of AI hype cycles warns that current generative AI and agentic systems risk repeating past failures, as critics argue that scaling deep learning alone cannot achieve human-level …