Latest posts
- Even more good news for the future of neurosymbolic AIApr 12, 2026Gary Marcus
And vindication for Apple’s unfairly maligned 2025 reasoning paper
- Three reasons to think that the Claude Mythos announcement from Anthropic was overblownApr 09, 2026Gary Marcus
No need to panic just yet
- What should we take from Anthropic’s (possibly) terrifying new report on Mythos?Apr 08, 2026Gary Marcus
Not many facts are on the ground, but here are some starting points for sober thinking
- Sam Altman, unconstrained by the truthApr 06, 2026Gary Marcus
New reporting from the New Yorker vindicates concerns that were first raised here
- The back story behind the first “$1.8 Billion” dollar “AI Company”Apr 05, 2026Gary Marcus
AI isn’t the only thing behind Medvi
- The two wildest stories today in techApr 03, 2026Gary Marcus
Shifting goal posts and new efforts at redefining the narrative
- On employment, don’t panic – yet.Apr 01, 2026Gary Marcus
Things will get wild, but probably not immediately
- In the Iran war, it looks like AI helped with operations, not strategyApr 01, 2026Gary Marcus
A hot take on how AI may have played out in the war so far
- “CEO said a thing!”Mar 30, 2026Gary Marcus
A blistering guide to what lazy journalism too often looks like
- The mirage of visual understanding in current frontier modelsMar 29, 2026Gary Marcus
When a model achieves a “top rank on a standard chest X-ray question-answering benchmark without access to any images” you know something is deeply wrong.
- War and AI, the death of Sora, and 3 ways you can catch me live todayMar 25, 2026Gary Marcus
With apologies for the short notice, a few quick announcements:
- BREAKING: Sam Altman concedes that we need major breakthroughs beyond mere scaling to get to AGIMar 16, 2026Gary Marcus
It’s past time to look for new architectures
- BREAKING: Expensive new evidence that scaling is not all you needMar 14, 2026Gary Marcus
Two more colossally expensive experiments have failed
- Is the US military actually afraid of Claude? A new theory of why Anthropic was labeled a supply chain risk.Mar 12, 2026Gary Marcus
Unpacking a perplexing argument from the Pentagon
- “A spate of outages, including incidents tied to the use of AI coding tools”, right on scheduleMar 10, 2026Gary Marcus
Some with “high blast radius”
- Anthropic sues US government, with good reasonMar 09, 2026Gary Marcus
As I wrote yesterday, Dario Amodei is no saint, but I fully support his company’s new lawsuit against the US government.