got it, makes sense.
how do you think about building a moat around Spirit of Satoshi? how easy/hard do you think it will be for someone to replicate/compete with your model?
The moat is in the quality of the model.
Honestly - it's been over 6 months of tinkering and fucking around with data structures, data styles, data mixes, training formats.
This whole process is more art than science. And unless you know Bitcoin very intimately, you're not going to compete.
And if you do, you need to really have a good grasp of training models. Which honestly - not many people do. We have 4 Data scientists who happen to be Bitcoiners too, working with us - and we've all been pulling our hair out.
So I'd say perhaps the outward facing "moat" will be the quality of the model, and the inward moat is the experience we've had (trial, error, experimentation).
That's hard to replicate.
reply