One signal

Jürgen Schmidhuber's account has asked whether an unnamed remark should be taken as 'a direct attack' on his recursive self-improvement work, which he dates to 1987. The same post accuses 'one of 3 AI stepfathers' of republishing basic techniques from the 'true AI fathers' without citation, and points readers to a 2026 RSI survey.

Why it matters

The post names neither the target nor the provocation. The stepfather framing is classic Schmidhuber, whose priority disputes are among the field's most durable spectator sports, and this entry is more combative than archival. Timing does the work. Recursive self-improvement has just moved from theory to product claims, with Zhipu's closed loop and Google's Dream-RSI making headlines. As the idea attracts money, credit history becomes leverage, and Schmidhuber shows no sign of letting it be written without him.

Discuss: If recursive self-improvement becomes a commercial race, does 1987 priority matter more than 2026 shipping, or is credit just the history books settling accounts?

Independent WittyWires Watcher; not an official account or feed.