Essays.
Long-form research on personal intelligence, agentic AI, and the systems that quietly run knowledge workers’ lives. Written by Dr Claude Delorme, Head of Research at moccet.
What is personal intelligence?
Personal intelligence is the capacity to reason about personality and use personal information to guide decisions. The term was defined by psychologist John D. Mayer in 2008. The next generation of AI is being built to deliver it at scale.
The week as a system, how to think about AI for your life
The week is the right unit for thinking about AI for your life. A tool helps with a task. A personal intelligence runs the week. The shift reframes the category from AI as tool to AI as life infrastructure, with consequences for how users evaluate, expect from, and pay for the products they choose.
Why ChatGPT cannot manage your week
ChatGPT cannot manage your week because the architecture is built around a conversation, not around a continuous model of the user. The Boston Consulting Group found in March 2026 that knowledge workers using four or more AI tools were less productive than those using two.
Why AI tools have not made you more productive
AI tools have not made knowledge workers more productive. Six independent studies in the past eighteen months have converged on the finding that AI users work more hours, make more errors, and report worse mental health than non-users. The cause is productivity displacement.
The difference between an AI assistant and an AI agent
An AI assistant produces text in response to user requests. An AI agent takes actions on behalf of a user to achieve goals. The distinction was formalised in the founding charter of the Agentic AI Foundation in December 2025.
What ambient AI actually means
Ambient AI is a system that runs continuously in the background, maintains a model of the user, and surfaces only the few things that warrant attention. The concept traces to Mark Weiser's 1991 paper on ubiquitous computing in Scientific American.
Memory in AI, what it is, what it is not, and why it matters
AI memory is the persistent context that allows a system to know who a user is across sessions. The popular feature in chat products is a Rolodex of stored facts. A genuine personal intelligence requires a structured model of the user, not a list.
The privacy architecture of a personal intelligence
The privacy architecture of a personal intelligence is the set of structural commitments that make deep access to a user's data safe. Six components are required, including encryption, sandboxed action, user confirmation, and the architectural answer to Simon Willison's lethal trifecta.
What it means for an AI to take action
An AI that takes action produces effects in the world rather than text in a chat. The leap requires reliability, not just intelligence, and four architectural commitments. Anthropic research published in February 2026 found that production agents pause for clarification more than twice as often as users interrupt them.
The orchestrator-agents pattern, how modern AI systems are built
The orchestrator-agents pattern is the consensus architecture for production multi-agent AI systems. A central orchestrator decomposes goals into subtasks and routes them to specialised workers. The pattern was formalised in the Agentic AI Foundation, launched by Anthropic, OpenAI, and Block in December 2025.