DESK ENGLISH · 5 MIN
Rebuilt: how AI agents should remember – and what held up in our test
How does an AI agent remember what a conversation was about without believing everything it is told? The automation vendor n8n has published five recommendations on this.
CREDIBILITY 3/5 – Partly robust
-
OpenAI says research model spent test-network funds it was not given
-
Microsoft Research says its Magnetar zkVM proves programs 10 to 39 times faster
-
Working paper estimates AI writing in 29% of 2026 US STEM dissertations
-
Opinion authors say youth social media bans are easily circumvented
-
Stanford Encyclopedia entry maps open questions in the epistemology of AI
-
Google DeepMind and GovTech Singapore describe on-device scam detection model
-
Google DeepMind details Gemini 4 Argon safeguards, citing mostly internal tests