AIDE² presents first documented evidence of recursive self‑improvement

Researchers behind AIDE² claim to have observed recursive self‑improvement. The finding marks what they describe as the first documented evidence. Recursive self‑improvement refers to a system enhancing its own capabilities iteratively. The experiment involved the AIDE² architecture operating under controlled conditions. Performance metrics showed measurable gains after each improvement cycle. The result is being discussed as a milestone in artificial intelligence research. Critics caution that the evidence remains limited to the specific setup. Future work will aim to replicate the effect across broader AI models.