I came across this on X (AGI at home) and checked out EXO Labs which, other than the site’s colour scheme, reminded me of Grin and Grin’s mission. I too, fear a world where power is concentrated into the hands of a few and it seems that technology could allow for super concentration. Decentralised tech seems to be the way forward. I’m not sure at this stage how Grin and AI connect but perhaps the idea is that both AI and money need to be forces of good.
One thing I did notice was that LLMs and the Cuckoo Cycle algo are both memory-bound. I expect that there’ll be a point soon, if not already, where a GPU could outperform current G1 ASICs. Perhaps AI hobbyists would mine Grin when not training models.
Most Ai framework runs off a token based system, it would be possible to run an Ai model on a server and offer the Ai service for public use through an api, using grin coin to pay for token usage directly. Problem with this idea is Ai is rapidly becoming easier to run locally and cheaper or even free.
Although I suppose something like that could be built into a grin wallet to allow interaction with Ai within the wallet to automatically pay for inquiries. I really liked how yeastplum was thinking of ideas about grin + ai, really sucks the community didn’t eeem to be very positive on the idea.
I was negative about it because of the core priorities. People want AI to be able to transact with each other but we don’t even have a GUI wallet that allows humans to transact with each other. I mean come on lol.
For example “using grin coin to pay for token usage directly” is not possible until we have a payment proof flow where servers can verify that payments have been made by the user to authorize the usage.
It’s a cool idea for a community project though and I hope Yeastplume still explores it.
I wish there would have been a little more discussion rather than him leaving at the tiniest bit of pushback, but honestly why was the OC not involved in helping set priorities? Why is the OC not meeting to discuss priorities internally at all?
Community helps to set priorities when asking what they need, good example is above. Github, this Forum, Telegram chats are public places, your voice can be heard also at Keybase. Payment proofs are top priority and this is not only GUI part.
Does anyone have a link to the most recent conversations on Payment Proofs? Perhaps we could regenerate some excitement around it. It sounds like grin could easily become part of the conversation on the best technology for AI agents to use, especially with its scalability advantage. The built-in amounts blinding could also be super useful for it in terms of business strategy. Privacy maximalists give grin a hard time regarding its linkability but they may be overlooking the steps in between.
One other thing that I think could be super useful is Atomic swaps, in particular with a currency like ETH or SOL or any other established currency that has a DEX. If there’s anything said about that too please let me know.
If my memory serves me right, Yeastplume did start work on tromp’s proofs quite some time ago. It might be in the grin-wallet contracts branch, but I can’t remember if it was finished or not.
The contracts error is familiar but I don’t know why the sender wouldn’t be specified in this case. It would need a bit of debugging, but unfortunately, I don’t have the time (or motivation) to look into it right now:(
People can still discuss things freely and share ideas.
