Like most of the internet, I've been diving into Fable 5 over the last 24h. And like most of the internet, I've been pretty blown away with the quality. But as I've been using both Fable and GPT-5.5, I couldn't help but notice there are clear differences in approach which make the two models behave quite differently. And we're seeing two very different training regimes play out. For any frontier model, accomplishing real work is an exercise in context management. The model needs to solve a problem across a very large number of tokens; some are explored via tool calls, others are the model thinking. Then it needs to produce a result. To get models to solve harder and harder tasks that run for increasing amounts of time, you need to figure out how to scale that context management. OpenAI: the oracle Since roughly ChatGPT 5.3-Codex, I've noticed that the model has improved a lot at dealing with long context windows. It stays coherent even across long-running tasks or /goal…
No comments yet. Log in to reply on the Fediverse. Comments will appear here.