Recursive self-improvement from AI models

With Claude Opus 4.6 and 5.3 Codex, both stellar achievements, the pace is heating up:

OpenAI went from its last Codex release, on December 18, 2025, to what is widely acknowledged to be a much more powerful one in less than two months. This compares to frequent gaps of six months or even a year between releases. If OpenAI can continue at that rate, that means we can easily get four major updates in a year.

But the results from what people in the AI world call “recursive self-improvement” could be more radical than that. After the next one or two iterations are in place, the model will probably be able to update itself more rapidly yet. Let us say that by the third update within a year, an additional update can occur within a mere month. For the latter part of that year, all of a sudden we could get six updates—one a month: a faster pace yet.

It will depend on the exact numbers you postulate, but it is easy to see that pretty quickly, the pace of improvement might be as much as five to ten times higher with AI doing most of the programming. That is the scenario we are headed for, and it was revealed through last week’s releases.

Various complications bind the pace of improvement. For the foreseeable future, the AIs require human guidance and assistance in improving themselves. That places an upper bound on how fast the improvements can come. A company’s legal department may need to approve any new model release, and a marketing plan has to be drawn up. The final decisions lie in the hands of humans. Data pipelines, product integration, and safety testing present additional delays, and the expenses of energy and compute become increasingly important problems.

And:

Where the advance really matters is for advanced programming tasks. If you wish to build your own app, that is now possible in short order. If a gaming company wants to design and then test a new game concept, that process will go much faster than before. A lot of the work done by major software companies now can be done by much smaller teams, and at lower cost. Improvements in areas such as chip design and drone software will come much more quickly. And those advances filter into areas like making movies, in which the already-rapid advance of AI will be further accelerated.

Here is more from me at The Free Press.