@Sophty_@sama Nobody still cares about 4o except a community of boomers stuck in 2024. Deploy an OpenClaw, fine-tune your SOÛL.md and you’ll have your perfect confidant with whatever personality you want. The model is not the product anymore — your setup is.
@kgnvsk@steipete Instructions in SOUL.md.
e.g. "**Tool-first, always.** When a task involves data you don't have in context, call the tool IMMEDIATELY. Do not announce what you will do — do it."
@steipete@steipete Great work on the GPT personality tuning! For those of us with heavily customized SOUL.md to match Opus : should we trim it down now or does the new layer stack on top ?
The MiniMax Token Plan was designed from the beginning to be used across third-party harnesses.
There will be more good ideas of how to use AI coming from outside the AI labs than in them.
Limiting AI subs to first-party products kills these ideas before they are ever born.
@bcherny@jaredctate Whatever the excuse, this was a catastrophic signal about Anthropic’s real stance toward open tools and power users. You had the dev community on your side.
Instead of consolidating that lead, you just revived OpenAI.
@malvei100x20@p_duval Pas forcément, sur un horizon d'investissement très long la volatilité n'est plus un problème.
D'autre part on peut aussi très bien tirer du rendement de la forte volatilité d'un actif (straddle long).