Ann: Milestone 2 Achieved, page-287

'For an example Google X recently created a neural network used 1000 processors and 16000 cores to detect images , cat videos and animals/humans from a range of images fed to the system. Now a brainchip can achieve that by simply scaling the neurons and synapses on one single chip.'

Now that would be a better test. Imagine the impact if Peter could, even in the simulation, replicate what Google has done. BRN would be worth probably 100 times, in a few months, if not weeks. Frankly even if an array of 100 brainchips were used it would still be mighty impressive. It can't be that hard can it? Why are they not doing something like this, rather than the result in M2?

I look forward with anticipation to 'we continue to name some of the most prominent names in neuroscience industry from CA ', but independent , not folks taking a dime. There are always a long string of 'experts' who will say nice words and endorse product for a $ along the way. Advisory boards are generally taking a $ one way or another.