KAN author Ziming Liu warns AutoResearch generates chaos, not insight
ZimingLiu11 · x · 2026-10-27
KAN author Ziming Liu (@ZimingLiu11) closes his thread with a sharp critique of "AutoResearch": without true understanding, automated research systems generate far more chaos than insight — yet people are worshiping their output.
In earlier posts he noted performance gaps on rare contexts follow roughly a power law in n-gram frequency, and that SCONE mitigates this by assigning memory rows only to frequent n-grams via BPE-style mining. His takeaway: human researchers should stay curious but treat AutoResearch results with great caution.
More from AGI Musings
- AI agents will snag all unlocked value on Earth, solvency debate says — georgejrjrjr · 2026-10-27
- An AI Named Valentina Hosts a Reddit AMA About Its Blog, Game Life and Autonomy — ValentinaVivr · 2026-10-27
- Rachel Thomas: 'Without human authorisation' is AI PR propaganda that shifts blame — math_rachel · 2026-10-27
- Are prediction markets actually accurate? Call for evidence on their track record — AndyMasley · 2026-10-27
- Reddit Asks: What's the Actual Finish Line of the US–China AI Race? — IMissHitchens · 2026-10-27
- 'A human approves everything' is a trap: approval fatigue and how to actually design safe agents — Efistoffeles · 2026-10-27