Anthropic explains that the next big step for humans and AI is an AI that builds itself through recursive self-improvement

Anthropic explains that the next big step for humans and AI is an AI that builds itself through recursive self-improvement

By Lance Eliot
Publication Date: 2026-06-07 07:15:00

In today’s column, I explore the belief that AI will ultimately result in AI being built for us, rather than for software developers and engineers to do. This point was recently announced in a blog post by Anthropic. Anthropic’s AI researchers claim that the method underlying this effort will most likely be based on recursive self-improvement (RSI).

Let’s talk about it.

This analysis of AI breakthroughs is part of my ongoing Forbes column coverage of the latest developments in AI, including identifying and explaining various impactful AI complexities (see link here).

The goal is top AI

Avid readers know that I have repeatedly grappled with the vexing question of how humanity can further develop AI into some kind of cutting-edge AI. However, there is already an ongoing debate about what the cutting-edge AI itself will be; see my analysis at the link here.

Some insist that the top AI would be AGI (artificial general intelligence), so AI would be on…