AI BRIEFING · 2026-01-15

AI Briefing — January 15, 2026

The 10 most important AI stories of January 15, 2026, hand-curated from the day's coverage. Every story links to its primary source.

02

XAI Grok 4.20 and OpenAI GPT 5.2 Are Solving Significant Previously Unsolved Math Proofs

Next Big Future AI ↗

A Mathematician with early access to XAI Grok 4.20, found a new Bellman function for one of the problems he had been working on with my student N. Alpay. Not an Erdős problem, but original research. Demonstrates AI generating novel math objects (Bellman functions for optimal control) quickly, linking to isoperimetric profiles and Takagi function ...Read more

05

Diagnosing Generalization Failures in Fine-Tuned LLMs: A Cross-Architectural Study on Phishing Detection

arXiv ↗

The practice of fine-tuning Large Language Models (LLMs) has achieved state-of-the-art performance on specialized tasks, yet diagnosing why these models become brittle and fail to generalize remains a critical open problem. To address this, we introduce and apply a multi-layered diagnostic framework to a cross-architectural study. We fine-tune Llama 3.1 8B, Gemma 2 9B, and Mistral models on a high-stakes phishing detection task and use SHAP analysis and mechanistic interpretability to uncover th...

← 2026-01-14 All briefings 2026-01-18 →

Track the tools behind the news:

→ AI Tool Leaderboard → Browse the directory