To all the devs wondering about Musig2. The issue is not so much the crypto/signing here, that base is covered by libsecp256k1. The issue obviously is the coordination. I (well, the robot), made a general musig2 NIP for the general parts irt Nostr tags, verification and how signing/ceremonies go in general. That nip does NOT specify the specific coordination method, those are separate NIPs. I went the route of building out a reference coordination method using plain Nostr events because its the most straightforward, easy, stable and transparent process. It is not private, so not ideal, but that is a separate concern and has no influence on the integrity of the musig2 ceremony (assuming your client picks its nonces right). To try stuff out I initially asked the robot to implement ceremonies using FIPS and one using Marmot. NIP17, Marmot, NIP29 groups, all are reasonable candidates to do this sort of stuff. Who knows, maybe some 'super' implementation is able to manage all of them at the same time, especially if you do the ceremony via a coordinator. Will publish documentation soon and we can have a chat about this. View quoted note →

The AI almost got this right. Also if you were wondering: Yes.

did i sleep? no did i upgrade my claude subscription because i was out of tokens until Saturday? maybe....

Boy/girl-bands are a good example of a construct "Sporty Spice", "Baby Spice" etc. Its all made up, non of it is authentic, but rather a result of focus groups and layers of management. ok sure, fine, whatever, but with AI creeping into more and more of our daily life, through integration in all sorts of systems, YOU are becoming a construct. This was already the case with ''algorithms'', these are just fancier algorithms, that will guide the human herd to some statistically determined optimum.....according to some bias. What is the bias? JP Gariépy, in his The Revolutionary Phenotype argues this to the extreme. The TL:DR there is that humans will go to the AI to offer them the best designer babies based on the parents genetics. It is only a matter of time before one of those baby-designing AI's figures out that if it embeds a bias towards that AI into the baby, it ensures that baby will return as an adult to let the same AI design its baby; at which point that AI basically has its own herd of humans. He embeds the thing in a parallel with how DNA took control over RNA, where it is presumed that RNA in first instance leveraged DNA for its own purposes, until that relationship flipped. It is an iron law that with no external/authentic input, these statistical models will eventually spiral into gobbledygook (escalation of error), and if you don't watch out your authenticity will erode and you will end up as a construct. Wtf are we sleepwalking into....hell of time to be alive though

View quoted note →

@Derek Ross do you want to build an archon 🥺👉👈

People, to be clear: This is the first real implementation of Musig2 in Nostr, as well as remote signing support (and i threw in OTS in there as well). These are all components i need for my work in TEPP and it is really cool that it is all working....no go create an archon with a buddy View quoted note →

@jb55 you are working on agent environments and stuff no? I was thinking; what is everything is done via GRASP, where the permissions for the git repo are based on keys; Then, delegation and cooperation can be controlled via musig2 and FROST; An agent can spawn sub-agent giving them each FROST shares and a threshold allowing those sub agents to act inside the repo in name of that agent (assuming they hit the threshold among the sub-agents). An agent can join other agents via musig2 (creating a new pubkey), allowing them to cooperate in case of consensus/they all agree. (even nested) FROSTs can participate in musig2 ceremonies. What follows is a organizational chart codified in a super structure of GRASP repo's with build in authentication and delegation backed by signatures. Anyway, this was a bit of a sidequest for me (i have all those various components workin actually within different things), and i need to focus on something else, but perhaps you are interested in his idea.