[02:21] <floridagram-bot> <AdamOutler> https://www.youtube.com/watch?v=2Pi7Smh3b2s
[03:53] <floridagram-bot> <govatent> I have a regular ir harmony remote
[14:08] <floridagram-bot> <KMyers> @RazPi - https://www.engadget.com/2018/11/14/intel-neural-compute-stick-2/
[14:38] <floridagram-bot3> <AdamOutler> Seriously?  AI is way overrated.  The example shown is a drone.  The damn thing just needs a linear feedback algorithm.
[14:39] <floridagram-bot3> <AdamOutler> What is this really for?
[17:00] <maxolasersquad> Here's a cool system monitoring tool for Linux named Hegemon. https://github.com/p-e-w/hegemon
[17:00] <maxolasersquad> It's a work in progress but is already really cool.
[20:00] <floridagram-bot3> <RazPi> @KMyers OMGGGGGG
[20:07] <floridagram-bot3> <RazPi> @AdamOutler the value of neural networks becomes more apparent  when you consider that you can have many more than three layers, convolutions to discard the necessity of pattern location in an input space, and that you can connect multiple specialized networks together in any directional flow you like. Moreover you can feed them back into themselves to cover patterns over time. The whole picture result of this
[20:07] <floridagram-bot3> is simply that NN's actually depict a generalized programming language.