@scaling01

Who Is This Person?

The Twitter account @scaling01, with display name 'Lisan al Gaib', appears to be run by an individual deeply involved in AI research, particularly in benchmarking large language models (LLMs). The account became active around December 1, 2024, with the user posting over 14,000 tweets in the first year. Content focuses on AI model evaluations, comparisons (e.g., Gemini, DeepSeek, GPT variants), benchmarks like LMSYS-inspired arenas, and critiques of industry practices. Recent activities include discussions on model parameters, training costs, and new releases like Gemini 3 Pro, with posts garnering significant views (up to 662k) as of November 2025. No formal organization is tied to the account, suggesting an independent researcher or enthusiast.

High credibility in niche AI benchmarking due to detailed, evidence-based posts and community resonance. Lacks formal affiliations or verification, which slightly tempers broader authority, but no red flags for inaccuracies or disputes. Suitable as an insightful voice in AI discussions, though independent status warrants cross-verification of claims.

Assessment by Grok AI

What's Their Track Record?

No documented fact-checks, corrections, or controversies found. Posts demonstrate technical depth in AI topics, using methods like Glicko-2 ratings for benchmarks, with apparent accuracy in model comparisons based on public data. Historical posts show consistent, data-driven analysis without evident misinformation.