Researcher's optimistic hunch on neural networks draws skepticism
A blog post claiming AI models may be easier to interpret has sparked debate among experts.
On a report by MIT Tech Review AI
Beat
Research
MIT researchers report AI agents playing in virtual world
An experiment suggests AI agents can learn through unstructured play, though critics question if it's genuine.
On a report by MIT Tech Review AI
New benchmark suggests AI models struggle with intuitive visual reasoning
A new test of visual reasoning shows frontier models lag far behind human performance, raising questions about world models.
On a report by Import AI
AIpressr
From Draft to Wire in 24 Hours
Schema-marked, editorially reviewed, distributed to the people who cover AI — and the LLMs that cite them. Cite-ready by default.
Submit a release →
Anthropic reportedly finds new CRISPR-like enzyme system with AI
The AI company says its Claude model helped identify a new biological mechanism in a matter of hours.
On a report by TechCrunch AI
OpenAI's new math advisory group appears to offer limited oversight
The AI lab forms an external panel after its models reportedly solved over 100 open problems, drawing mathematician criticism.
On a report by TechCrunch AI
World model startups appear to be delaying competition through secrecy
Leading AI labs are keeping product plans under wraps, which may reflect a strategy to defer market rivalry.
On a report by TechCrunch AI
Anthropic researcher resigns, alleges AI labs gamble with existential risk
A departing researcher warns that leading AI labs are racing toward self-improving systems they believe could kill humanity.
On a report by TechCrunch AI
Google DeepMind alum reportedly pursues world-model AI for robots
A former Google researcher is applying his model-based reinforcement learning techniques to physical robotics.
On a report by MIT Tech Review AI
Hugging Face Blog says benchmark audit reveals mixed signals in LLM evaluations
A new method analyzes what individual benchmark questions actually measure, finding safety and reasoning scores often conflated.
On a report by Hugging Face Blog
Anthropic research suggests automated systems may improve AI alignment
A new paper explores using AI to train other AI models, raising questions about the future of human researchers.
On a report by TechCrunch AI
DeepMind Blog: Google pilots cryptographic double-blind AI model tests
Google partners with external institutes to test a Gemini model in a secure environment where neither side sees the other's data.
On a report by DeepMind Blog
Inherent's small AI agent reportedly beats larger rivals at paper replication
A London startup founded by DeepMind alumni claims its compact AI teammate outperformed frontier models on a scientific benchmark.
On a report by TechCrunch AI
Nvidia research suggests AI harnesses may matter more than models
A new study indicates the scaffolding around an AI model, not the model itself, is key for complex tasks.
On a report by TechCrunch AI
Hugging Face Blog says ASR models may be gaming benchmarks, not transcribing audio
Research suggests top speech recognition models can reproduce benchmark errors, overstating their real-world accuracy.
On a report by Hugging Face Blog
Study suggests AI authors one-third of web pages since ChatGPT launch
Pew Research analysis indicates a significant portion of new web content shows signs of AI authorship, with commercial sites leading the trend.
On a report by TechCrunch AI
Microsoft Research expands Skala DFT tool to major chemistry codes
The deep-learning functional's wider software integration aims to accelerate adoption in scientific and industrial workflows.
On a report by Microsoft Research