A new method for certifiable randomness

I came across this paper. What’s your thoughts @ScottWilber ?

It doesn’t provide anything new or useful for MMI. The “certification” method is way too slow to be of any use as an entropy source, and the bits generated by the processing are not any better than – or even as good as – from other sources.

I will provide a chapter analyzing entropy sources for MMI.

Thanks for taking the time to forward the reference.