OpenAI Codex Engineering Lead publicly asks users: Peak Pricing, Trough Giving — yay or nay?

According to 1M AI News, OpenAI Codex Engineering Lead Thibault Sottiaux posted on X that there is a significant load gap between Codex's peak and off-peak periods, and the team is looking to smooth out the traffic distribution to better utilize compute resources efficiently. He directly asked users for feedback: Is it reasonable to offer more usage during off-peak times and introduce multiplier consumption during peak hours? He mentioned that the team has some ideas but wants to hear from users first.

Addressing a similar compute scarcity issue, OpenAI chooses to ask before acting. In contrast, Anthropic's response to the Claude Code quota crisis is a usage guide: avoid using Opus, refrain from opening contexts of 1 million tokens, don't resume idle sessions, no compensation mentioned, and no user input requested.