What Happened

Jason Liu (@jxnlco), an AI instructor and developer, tweeted about "Tibo resetting rate limits." He was referring to Thibault Sottiaux, an OpenAI engineer who works on the Codex coding model and handles API rate limit adjustments for heavy users.

Summary

The tweet points to a recurring issue: developers who hit OpenAI's usage caps sometimes need someone at OpenAI to manually raise their limits before they can keep working. For developers relying on Codex for coding assistance, this creates delays.

Analysis

I identified "Tibo" as Thibault Sottiaux based on developer complaints in Twitter threads, Reddit discussions, and OpenAI community forums - his name comes up repeatedly in conversations about Codex API limits.

The underlying tension here is resource allocation. OpenAI uses rate limits to prevent abuse, but heavy legitimate users get caught in the same net. GitHub issues and forum posts show this is a common frustration.

Liu has a decent following in the AI developer community, so his complaint gets more visibility than most. Whether OpenAI's manual reset process can scale as more enterprises adopt their tools is an open question.

Some developers may look elsewhere. Anthropic and Meta's Llama models offer different access terms, and self-hosted options avoid the rate limit problem entirely. How much this actually affects OpenAI's position with developers is hard to say - convenience and model quality still matter more for most use cases.

Impact Assessment

• Significance: Medium
• Categories: Developer Tools, Technical Insight, Industry Trend