7/ This is why a color game matters. The models still had the knowledge. What they lacked was the ability to hold focus and resist a pull across a long stretch of input. That gap is the finding.
LLMS
-
AI models default to reading over naming colors
By
–
6/ The cause is in how these models are built. They were trained on text, so reading words is their strongest instinct. Naming the color means fighting that instinct. A human brain can suppress the urge. The model defaults to reading.
-
GPT-4o accuracy falls sharply with more words and color mismatch
By
–
4/ Everything fell apart. GPT-4o scored 91% on 5 words. At 10 words it dropped to 57%. At 40 words it hit 15%. When the colors and words were mismatched, accuracy fell to near zero.
-
AI models ace short lists but struggle with longer ones
By
–
3/ Researchers ran this test on the top AI models. GPT-5, Claude Opus 4.1, Gemini 2.5, and others. On short lists they aced it. 90% and up. Then the researchers made the lists longer.
-
The two most absurd weeks in AI history
By
–
The last two weeks in AI have been the most absurd period in the history of tech. I can't even keep up. And that's literally what I do for a living. → Anthropic launched Claude Fable 5, its most powerful model ever made public. At the forefront of
-
Sparse models of that size: 40-60 GB, fast; dense models too slow
By
–
Sparse models of that size use about 40-60 gb and are fast enough for me (but dense models of that size I agree there bare too slow)
-
Anthropic CEO discovers Chinese founder giving away architecture surpassing Claude for free
By
–
Le CEO d’Anthropic qui ouvre X (Twitter)
— Jouhatsu | AI Influence Operator (@Jouhatsu_ai) 13 juin 2026
et découvre un fondateur chinois d'IA à 20 milliards de dollars distribuer gratuitement, en 40 minutes, l'architecture exacte qui surpasse Claude https://t.co/tZDSjCaIKh pic.twitter.com/IXs4QWENGPThe Anthropic CEO opens X (Twitter) and discovers a Chinese AI founder worth $20 billion distributing for free, in 40 minutes, the exact architecture that surpasses Claude.
-
Training a Mythos model requires noticeable amounts of compute
By
–
And by regulatable amounts of compute, I mean training a Mythos class model uses enough power and chips that national governments will obviously notice. No ine is training a model of that size without permission
-
Mythos-class models won’t be open due to regulation and risks
By
–
I don’t think this is going to result in more open weights models, as I wrote before the Anthropic news, if Mythos-level models are considered risky, China will also not want them to be open. And you can’t build a Mythos-class model without a very regulatable compute footprint