Reddit thread says the next giant pre-training run could trigger another capability jump
imadade · reddit · 2026-07-29
A Reddit thread warns the next giant pre-training run could unlock another jump in capability
The post argues that if current internal models at OpenAI, Anthropic, xAI/Grok, Meta, and DeepMind are already as strong as people suspect, the next training run with 2–3x more compute could produce a far more capable system.
Main claims
- The author expects major compute clusters to come online in roughly 12–14 months.
- With better pre-training, post-training, and scaffolding, the next generation may cross into superhuman performance in math, physics, CS, and related logical domains.
- The post speculates this could accelerate the recursive self-improvement loop and make AI-2027-style forecasts arrive sooner than expected.
It is a prediction thread, not evidence, but it is an example of the kind of AGI timeline talk that shapes the discourse.
More from AGI Musings
- First-ever paper on multi-agent swarm safety lands NeurIPS spotlight — LauraRuis · 2026-10-16
- Yoav Artzi on peer review: paper flooding still pays off as review system strains under load — yoavartzi · 2026-10-16
- Letta founder: work in fields defined by problems, not methods — lateinteraction · 2026-10-16
- Why pure mathematicians will soon flood into AI x-risk debates — Dr_Atoosa · 2026-10-16
- RSI began when model outputs augmented next-gen training data, agents just escalated it — akbirthko · 2026-10-16
- Luiza Jarovsky: with AI making careers unpredictable, 'do what you love' is the best advice left — LuizaJarovsky · 2026-10-16