I thought OpenAI said fine tuning and not bigger models were the future?

Side note: Wonder if Twitter will be filled with graphs of the motivation crowd’s interpretation.

Tiny circle = GPT4 Huge circle = GPT5