But this is all the easy part

The hard part is getting an innovation into service without blowing stuff up, and then freezing it into the standard way things get done.

That requires the ability to flip from #innovation to #formalization. You have to document the innovation, test it across a wide number of scenarios, write up how it works, what it does, and who can use it and when. You must rebuild existing workflow to use it well, and you must have metrics to check that it works as designed
/6

@corbden So, this whole complex of ideas (#philosophy)a is connected to (among others) the concepts of #incompleteness (#Gödel #Goedel), #creativity #cognition #psychology #aesthetics #pedagogy #learning #AI #AGI and the limits of #algorithms and #formalization.
There are known instances of physical entities for which it can be shown that they are not computable. I think human beings are like that and AGIs must be like that. The limits of the Turing-computable are not the limits of the possible.