Zhipu's Tang Jie: Parameter scaling hits threshold; future relies on long-horizon reasoning
rohanpaul_ai · x · 2026-08-21
Zhipu AI founder Tang Jie argues that parameter count alone is becoming a weak descriptor of model capability. He identifies four scaling dials: parameters, data, compute per forward pass, and post-training. Tang suggests parameters need only reach a threshold to 'hold the world,' with further gains coming from increased depth per forward pass (long-horizon reasoning) and post-training. GLM-5.3 is presented as an implementation of this 'train longer, think deeper' philosophy.
More from AGI Musings
- Tech Erodes Individualism; Uploading Consciousness Might Just Save Media Logs — yeastsplainer · 2026-08-21
- Agentic search may be the first rehearsal for a web with invisible economic graphs — schwentker · 2026-08-21
- 130 years, 10^22x more compute per dollar: Kurzweil's graph sparks debate — Singularitarian · 2026-08-21
- Rick Rubin: AI is the new punk rock ushering in a golden era of creativity — TheChuckTone · 2026-08-21
- Automation endgame may not be billions of operators — VraserX · 2026-08-21
- Prediction: Cognitive AGI, Autonomous Driving, and Robots Arriving by End of Year — davidpattersonx · 2026-08-21