Musk teases Grok 4.8: no benchmark optimization, but 'incredibly useful'
minchoi · x · 2026-10-30
Elon Musk confirmed Grok 4.8 is in the works, admitting it has taken longer than expected but promising it "will not be a disappointment." He stressed that xAI isn't optimizing for benchmarks — saying he doesn't care how Grok scores on them — and that the focus is on making the model "incredibly useful," continuing his critique of leaderboard-driven development.
Related event: Musk Confirms Grok 4.8 Is Coming, Prioritizing Usefulness Over Benchmarks(3 posts)→
More from Models
- Deep Research with Gemini 4 Argon generates a full Amazon returns infographic — gaganghotra_ · 2026-10-30
- Mistral Open-Sources 52B-Active Flagship at Half Price; AI Safety Incidents Roundup — Latent Space · 2026-10-30
- MIT's Soft-MTP Fixes Noisy Gradients in Multi-Token Prediction, Free Perplexity and Reasoning Gains — Massachusetts-Institute-of-Technology · 2026-10-30
- Insider says Google's unreleased model isn't named Xenon, denies rumors — imjustnewatai · 2026-10-30
- ChatGPT's blunt answer on what happens to your data if OpenAI goes bankrupt — alex_verem · 2026-10-30
- Mistral open-weights Le Chonk tops FrontierCode among open models, targets cyberdefense — MistralAI · 2026-10-30