All this are basic foundations for AI, nothing innovative at all, at least until Distributed Computing point is reached using the blockchain tech, at that stage the users will still not be able to do crap, do you guys even understand how DC woks? If you would, then you'd know that custom DC propagated projects are just not viable since you would be just bloating the other support network cycle contributors. At that point is either you are the only one in the network pulling cycles (unlikely), or everyone is and the project is pointless.
That is why a focused goal is necessary to achieve tangible advantage point results for the network, not pretending that each user will be able to launch their own ML projects and they will get substantial results, that is just not viable, unless we apply another twist but I have never heard of it in this forums so I am going to go and risk assuming that Cedric does not know about it, and without that, this project is only a wannabe AI exercise.
Still cool for educational purposes.
Oh and you think it is the only blockchain AI project ongoing? Think twice.
Life exists beyond BCT

~P~
If this were the case then Facebook, Twitter, Microsoft, Google and several startups would not have internal or external products using similar architectures, would they? And the entire body of knowledge from the past ~14 years or so would not exist.
You can't build a skyscraper without having the foundation underneath it.