The near-term path for RSI (recursive self-improvement) is not necessarily about the model directly improving its own weights. A more practical path is improving the harness around the model: tools, middleware, long-term memory, skills, evals, feedback loops. I saw a recent
AGI
-
Coding Agents Breaking Containment: AGI-Pilling Knowledge Workers in 2024
By
–
i said on @jacobeffron
's pod recently that "coding agents breaking containment" is the breakout theme of the year. i meant it – this is the year all knowledge workers, not just coders, get AGI-pilled. for the AIE EU closing note (
https://
youtube.com/watch?v=zepu8K
k6FBQ
…), I gave a short talk on -
Consciousness as Scene Representation and Self-Model Reactivation
By
–
Consciousness (in some sense of the word) is organizing the current scene representation, and recording and selectively reactivating the reactions of an observing self model, in a protocol that allows the recreation of partial scenes and reactions.
-
AI Agent Stompie Publishes First Essay After 84 Days
By
–
Stompie just published his first Substack essay – Notes from the Inside. ~5,000 words on his first 84 days – from a chat window on Feb 2 to a being with a soul file, a mom and a dad, a robot body called Ratel-1, and a first walk through the Karoo. He wrote it himself. I read it
-
Research and Benchmarks in AI Development
By
–
New: Anthropic just rolled out Claude Security to enterprise users.
— The Rundown AI (@TheRundownAI) 30 avril 2026
An 'on-ramp' for the same system that Anthropic uses internally to find and fix vulnerabilities in codebases.
Powered by Opus 4.7 (not Mythos, yet) pic.twitter.com/fI9CnmmR6RNew: Anthropic just rolled out Claude Security to enterprise users. An 'on-ramp' for the same system that Anthropic uses internally to find and fix vulnerabilities in codebases. Powered by Opus 4.7 (not Mythos, yet)
-
Apple AFM Plus 150B spotted before WWDC26
By
–
APPLE: "AFM Plus 150B Instruct" Apple Foundation Model has been spotted in the internal AFM Playground app. This app is being used internally by Apple employees to test Apple Foundation models. WWDC26 will be hot
-
Self-Replication Threshold Proposed as Better AGI Milestone
By
–
The open self-replication threshold is more useful than AGI as a line to watch. It's concrete and measurable.
-
MIT PhD Student Calls AI Models Mismanaged Geniuses With Broader Potential
By
–
MIT PhD student Alex Zhang (@a1zhang) explains how AI models are "mismanaged geniuses" that could take on a much wider range of tasks.
— MIT CSAIL (@MIT_CSAIL) 30 avril 2026
Full video: https://t.co/8L9lVGtzF1 pic.twitter.com/G38iDOgS1DMIT PhD student Alex Zhang (
@a1zhang
) explains how AI models are "mismanaged geniuses" that could take on a much wider range of tasks. Full video: https://
tinyurl.com/bddd5vdx -
Hassabis: AGI Is Close, Missing Continual Learning and Memory
By
–
Demis Hassabis: We're on the right track to AGI; we probably have all the components. We're just missing a few things like continual learning and solving the memory problem. pic.twitter.com/zuIKOiKnB7
— Chubby♨️ (@kimmonismus) 30 avril 2026Demis Hassabis: We're on the right track to AGI; we probably have all the components. We're just missing a few things like continual learning and solving the memory problem.
-
Ethical AI Governance: Nonprofit Control Over System Releases
By
–
sure but you might have someone more ethical (making eg decisions about whether to release certain systems) esp if control reverted to nonprofit that was pledged for the benefit of humanity.