New Model Generation Starts to Feel Like AGI
Crazyscientist1024 · reddit · 2026-07-16
Based on his trial experiences with RL and image generation, the author believes the latest batch of models has clearly become more adept at "research" and more creative.
He noted that past models often only made incremental 5% or 10% optimizations within existing codebases, tending to generate "slop" code. However, after using Fable and Sol, this issue was significantly mitigated; the models even proactively suggested interesting experimental directions and cleaned up legacy codebases better.
He concludes that this generation of models will significantly accelerate AI/ML research, leading him to believe that Recursive Self-Improvement (RSI) is approaching, and AGI is imminent.
Related event: New Model Generations Bring AGI Closer to Reality(3 posts)→
More from AGI Musings
- Claude Code skill uses 10 Markdown rules to make outputs ADHD-friendly — alex_verem · 2026-07-22
- AI Power Demand Exposes US Energy Gap, Urging Shift from Scarcity to Abundance — bradneuberg · 2026-07-22
- ControlAI CEO says an international ban on superintelligence is needed to avert extinction risk — zetalyrae · 2026-07-22
- Gary Marcus says LLMs still cannot really do math on their own — GaryMarcus · 2026-07-22
- Gary Marcus says LLM math skills are like knowing only a car’s engine size — GaryMarcus · 2026-07-22
- AI may make digital work infinitely leveraged while offline life gets more human — illscience · 2026-07-22