[bainsfather]

I read the Quanta article because I am interested in optimisations for deep neural nets.

Am I right in concluding that your work doesn't apply to that area, as it is looking at a much 'easier' case?