With cloud computing becoming popular I think there is a very good opportunity in the future for a big lisp system whose primitives kept ugly clustering issues

With cloud computing becoming popular I think there is a very good opportunity in the future for a big lisp system whose primitives kept ugly clustering issues hidden from a typical developer. With that in mind, I would have them choose a few kinds of applications suitable for EC2 and then give me a proposal on how they would build a platform for other non-guru lisp programmers to easily run lisp programs that would make good use of as many EC2 instances as were available. It's no good if they come up with something only they can use - whatever they propose has to be usable by someone who may have only read SICP or the little schemer. Should the selected application area have reasonable market potential, I'd then try and find some potential customers or partners. Ideally these would be indifferent to how their problem was solved but could understand why lisp+this platform is a good fit for their problem. If no idea produced by the team made sense to any potential customer, then I'd give up on them and go find some other team.