So a bit of a JubJub update You can now manage your socials through Claude, get a video scheduled, caption and going to multiple socials in 3 prompts. you can also comment on projects and media with your team or collaborators. We also now show you and your collaborators in the vault, our onchain record of all media distribution, including URL's to the distributed media. This is a verifiable proof of ownership and we will shortly allow payments to flow from owner to editor and team, and also collect platform revenue and split it between channel dn content owners. We will continue to expand control planes and my next focus is to allow you to message your team, review media via telegram and discord (and Whatsapp if I stop being cheap)
- 0 replies
- 0 recasts
- 0 reactions
Anyone got a friend at Anthropic?
- 0 replies
- 1 recast
- 2 reactions
Turns out reasoning models don't really respond to hard rules the way you'd expect. They respond to understanding why something matters. Once I started writing instructions that explained the reasoning behind each constraint rather than just stating the constraint, compliance improved dramatically. The other thing that made a real difference, putting guardrails directly on the tools themselves, not just in the system prompt. By the time the model is deciding which tool to call, it's already past your general instructions. Meeting it at that decision point changes the outcome.
- 0 replies
- 0 recasts
- 2 reactions