Agents
July 19, 2026
Why AI agents fail: the seven failure modes
Gartner predicts over 40% of agentic AI projects will be canceled by 2027. The failures follow patterns, seven of them. The taxonomy: what breaks, why, which real incident proved it, and…
July 7, 2026
Agent memory: what should persist, and what should not
A plain filesystem storing markdown files scored 74% on standard memory benchmarks, beating dedicated vector systems. The…
July 6, 2026
How AI agents actually work: the loop behind the hype
Everyone is building "agents" and almost nobody explains what makes one. Not the marketing, the actual machinery: the…
July 1, 2026
What to measure before you deploy an agent
Agents that succeed 60% of the time on a single run succeed 25% of the time across eight. Task completion is the metric…
June 29, 2026
Human in the loop is weaker than it sounds
When the AI was wrong, experienced radiologists went from 82% accurate to 45.5%. A review step is only a control if the…
June 28, 2026
What is MCP? The standard that wired AI into everything
The Model Context Protocol went from a November 2024 announcement to the connective tissue of the entire agent era in…
June 25, 2026
Your agent works in the demo because the demo is the easy case
A demo is not a small production system. It is the survivor of a search process designed to find something that works,…
June 23, 2026
Most agent projects will be cancelled. This is why.
Gartner puts the cancellation rate above 40% by the end of 2027. None of the three stated causes is a model problem, and…
June 19, 2026
How AI memory works, and why the model remembers nothing
Ask whether a chatbot remembers you and the honest answer surprises people: the model itself remembers nothing. Language…
June 19, 2026
Multi-agent AI gets worse as you add agents
Orchestrator steering accuracy falls from around 60% with three agents to about 21% with ten. Coordination is not free,…
May 21, 2026
Will AI take my job? What the evidence shows
The frightening headline numbers and the reassuring ones are both real, because they measure different things. AI acts on…
Other subjects
Foundations
27
Theory, history and the parts that will still be true in twenty years.
Language & meaning
21
How language models handle language, and the places they do not.
Evaluation & evidence
12
Measurement, benchmarks, and how to tell whether a claim is worth anything.
Building with AI
18
Retrieval, cost, infrastructure and the practical business of shipping something.
Safety & governance
11
Alignment, security, regulation and who is accountable when it goes wrong.