Doesn't seem to show anything the can't already be done with deep learning. Is it just deep learning packaged to be easier to use for end users or something actually new?
The major differences are that it can move from one domain of expertise to another without changing any code, it's a unique method (not a neural network -- but we'll elaborate more on this in the next release), and perhaps most importantly, it can run on a Raspberry Pi (~12MB memory usage).
i don't have sound, so when i watched the video i don't know if there was someone explaining the output of the command line prompts, but from the video it looks like a grad student project, not a system to sell
Sound is definitely required for this video. And you're right, this is not something that's for sale. This video is effectively our "Hello, world!" We're optimistic about the technology, though, and look forward to showing off some really cool things in the near future.