What models miss, and what we know about it.
Our analyses on AI, its limits, and evaluating the skills models cannot measure.
Scientific validityJune 8, 2026 · Aziz Goumiri
Emotional intelligence: Surprise! AI can't read emotions
Reading an emotion in context is hard, yet fundamental at work. Understanding what the other person feels in real time helps you communicate better.
Scientific validityJune 8, 2026 · Aziz Goumiri
Finding rules from a handful of examples: the skill that still resists AI
AI is famous for being extremely good at logic, yet it fails certain logic tests. Why?
Scientific validityMay 19, 2026 · Aziz Goumiri
Negotiation: AI knows Fisher & Ury, and loses anyway
LLMs recite negotiation theory, yet lose money on every deal. What that means for your salespeople, and how to test for it.
Scientific validityMay 19, 2026 · Aziz Goumiri
Adaptation: why AI stays frozen when the market moves
LLMs handle a stable world just fine. When the market shifts, they cling to what they already know. Why that happens, and how to assess this skill in your candidates.