Journal
Long-form writing on memory, consent, and what private AI looks like in practice. Updated when there is something worth saying.
Privacy & Control
What ChatGPT, Claude and Gemini store by default, how to see and delete it, why deleting a chat does not delete a memory, and what UK GDPR does and does not give you.
23 July 2026 · 5-minute read
Memory
Why ChatGPT, Claude and Gemini forget what you told them, what each one actually remembers, whether you can move your memory between them, and what a portable memory looks like.
23 July 2026 · 5-minute read
Field notes
AI Twin works in more than 100 languages. Here is what that means today, what we are still improving, and an open invitation to help us get the translations right.
22 July 2026 · 4-minute read
Practical
Forward any email to AI Twin and it becomes typed, searchable memory. Here is how Magic Inbox works, and where its limits are.
15 July 2026 · 4-minute read
Foundations
AI Twin is a personal operating system for your memory. Here is what that phrase means, what it does not mean, and why we use it deliberately.
8 July 2026 · 4-minute read
Practical
The line we are holding for privacy-led professionals. What “showing your regulator” actually means, and what AI Twin commits to.
22 May 2026 · 5-minute read
Field notes
Most coverage of the EU AI Act addresses enterprise compliance teams. Individuals using AI products day-to-day have fewer guides. Here is what the Act actually changes for you, the user, when it takes effect.
16 May 2026 · 4-minute read
Field notes
A founder essay on the commercial price of a policy we are not bending on. Slower personalisation. No network effects. No free training data. Here is why it is worth it.
16 May 2026 · 3-minute read
Field notes
The market does not need another chatbot. The problem worth solving is what the chatbot forgets. Here is the founder argument for picking a smaller surface and a longer trust horizon.
16 May 2026 · 3-minute read
Foundations
UK GDPR was written before AI memory layers existed but applies to them anyway. Here is how the six principles and four most relevant rights map to memory products, in plain language.
16 May 2026 · 6-minute read
Foundations
The EU AI Act's enforcement window is bringing the consent conversation forward. What is a consent layer, what does it actually require of a product, and why ticking the terms of service box does not count.
16 May 2026 · 6-minute read
Foundations
Most AI memory today is chat logs in a database somewhere. Typed memory is something different. Here is what it means, why it matters for retrieval, and what the consent implications are.
16 May 2026 · 6-minute read
Foundations
The phrase is used loosely by vendors who are anything but. Here is what model-agnostic actually means at three levels, and why most of the AI tools you use today fail two out of three.
16 May 2026 · 6-minute read
Foundations
The phrase is everywhere now but rarely defined cleanly. Here is what a memory layer is, what it is not, and why most current definitions are written for engineers rather than the people who would actually use one.
16 May 2026 · 5-minute read
Newsletter
Join the newsletter. AI Twin is open to early users now. We will tell you when wider sign-in opens.
