At Friday's group meeting, Raffi mentioned a researcher in Japan
who works on evolutionary algorithms with human input. He
explained that he hoped to incorporate some of the same ideas
into his work: a human user would be presented with designs for
devices, and could choose the ones that looked most reasonable
and which ones looked completely unreasonable. I think the
user should give the program feedback by clicking
I also think there's far too little King James English in mathematical writing. For that matter, there's far too little skater slang. I could write sentences like this:
And lo! as a result of theorem 3 and some totally snarly algebra, there was a continuous Schur decomposition.
Hmm. Perhaps I'd better stick to a more conventional style.
- Currently drinking: Black coffee