These days there are so many claims of AI breakthroughs that a real one would risk going unnoticed.
RESEARCH
-

GPT-5 Visual Reasoning Capabilities Show Weakness in Testing
By
–
gpt5 visual reasoning is the weakest so far in my testing
-
Five AI Tribes: Core Tricks and Methodologies Explained
By
–
Each AI tribe has a trick it does everything with:
Symbolists: heuristic trick
Analogizers: kernel trick
Bayesians: conditional independence trick
Evolutionaries: crossover trick
Connectionists: embedding trick -
Google’s 270M-parameter Gemma targets Apple market
By
–
The 270M-parameter Gemma is Google's pitch to Apple.
-
ML4A Invisible Cities: Creative AI Project Exploration
By
–
that was a fun project https://
opendot.github.io/ml4a-invisible
-cities/
… -

The Next Big Thing Will Start Out Looking Like a Toy
By
–
The Next Big Thing Will Start Out Looking Like a Toy (1978, colorized)
-
AI Development Approach Matters More Than Platform Choice
By
–
The important thing is the approach not the platform, so that's what we wanted to focus the start on
-
Rapid Technical Evolution Through Industrial Revolution and World Wars
By
–
To be fair most of fast technical evolvement he saw in form of industry towns or worldwars
-

Human-AI Teams in Medical Diagnostics: Interface Design Challenge
By
–
Increasingly, the findings of controlled experiments go much further: doctors with off-the-shelf AI outperform those without in diagnostics… but AI alone outperforms doctors. Harder to know what to do with that. What systems or interfaces will result in better human-AI teams?
-

AI Services Specializing Between Persuasiveness and Accuracy
By
–
The future is hard to predict. Still, one read on "AI persuasiveness decreases AI accuracy" is that AI services could speciate on prioritizing "warmth" (sycophancy, consumer capture) or accuracy. Personality-first: ChatGPT, Meta, Grok.
Accuracy-first: Gemini? Claude API?