AISep 2026
How to Read an AI Benchmark Table: The Opus 5.5 Deep Dive
Five checks, nine rows, seventeen tests. A close read of the Opus 5.5 launch table, its footnotes and error…
Read the analysis ↗AN INDEPENDENT PERSPECTIVE ON AI
RESEARCH / LABS / IDEASAI, examined
closely.✳
Papers worth reading.
Ideas worth questioning.
What it means in practice.
Reading notes by Luis Villalobos ↗
Explore the collection ↓FEATURED READING ↗AI
A CLOSER LOOK ↗
September 22, 2026
How to Read an AI Benchmark Table: The Opus 5.5 Deep Dive
Five checks, nine rows, seventeen tests. A close read of the Opus 5.5 launch table, its footnotes and error bars, plus a field guide to the benchmarks every lab reports.
BY Luis Villalobos
FREE TO READ↗
FREE TO READ↗
FOLLOW THE QUESTIONS
The collection.
AISep 2026
Agentic Coding and Persistent Returns to Expertise
You decide what. The agent decides how. A deep dive into 400,000 Claude Code sessions, task expertise, success rates,…
Read the analysis ↗
AISep 2026
Grounding Agent Memory: Your Agent’s Memory Is Guessing
Memory buys the accuracy. Verification buys the margin. A close read of environment-probing curation, its benchmarks, costs, and implications…
Read the analysis ↗
AIJun 2025
Practical Read of the EU AI Act – Part IV (Final)
The final part of the EU AI Act series examines general-purpose AI models, systemic risk, provider duties, incident reporting…
Read the analysis ↗
AIJun 2025
Practical Read of the EU AI Act – Part III
How EU AI Act provider and deployer duties translate into engineering work: quality management, documentation, conformity assessment and safe…
Read the analysis ↗
AIMay 2025
Practical Read of the EU AI Act – Part II
In the upcoming segment, we’ll delve deeper into Part II of the analysis. For those who haven’t had the…
Read the analysis ↗
AIMay 2025
Practical Read of the EU AI Act – Part I
Various regulations, policies, and frameworks aim to govern the rapidly evolving world of AI. Traditional software regulations are becoming…
Read the analysis ↗
AIMay 2025
AI System Cards – Anthropic’s example
As part of the previous blog and with the recent announcement of the new Claude 4 family of models,…
Read the analysis ↗
General TechMay 2025
Probability/Severity Harms Matrix
This isn’t meant to be a comprehensive article about risk management (that’s a topic unto itself) nor is it…
Read the analysis ↗THE PERSON BEHIND THE NOTES
Curiosity is the starting point.
Understanding takes work.
I write about AI research, the labs moving it forward, and the ideas that deserve a closer look. These are my reading notes, questions, and perspectives on what it all means in practice.
KEEP THE CONVERSATION GOING
Stay curious.
Go deeper.↗
Future AI research notes and updates,
from Luis to your inbox.