This AI is really human-like and allegedly doesn't give pre-programmed responses. It's makers Robots without Borders say the project is open source but I couldn't find the code anywhere.
You can't find the source code because it doesn't exist. The whole thing was a 2015 kickstarter scam, and now it's a patreon scam. All the tech "demos" are obviously pre-scripted videos.
There was actually a question on Quora about it, in which one of the answerers managed to find, among other things, the question and answer that his fake AI gave in one of the videos.
$\begingroup$ After reading the Quora post I do not feel that it is a scam. $\endgroup$ Jan 22, 2018 at 3:37
1$\begingroup$ @harsh99 The fact that his (solo developed) super-AI, that is "completely unscripted," answered 2 complicated social questions with answers that were both found on Quora is pretty damning evidence. Add the "open source" (but not anywhere) aspect of the project, the failed kickstarter, the patreon, the only tech demos being youtube videos, and you have what should be an obvious scam. It's sad that anyone would fall for it and donate money to that scam artist. He claims to have an AI more advanced than anything at Google/Facebook/Microsoft and can't prove it with a simple web demo. $\endgroup$– sageJan 22, 2018 at 12:16
1$\begingroup$ @sage The fact that the open source code isn't available anywhere troubled me too. And yes, I also noticed the Kickstarter didn't even get $150. So yeah, I guess it's a scam. Thanks! $\endgroup$– harsh99Jan 22, 2018 at 14:25
1$\begingroup$ However, I'm tempted to add it might be programmed to fetch answers from the web. $\endgroup$– harsh99Jan 22, 2018 at 14:34
1$\begingroup$ While I also believe it's a scam, there is no evidence to back up the claims you make in your answer, either. Do I believe the youtube videos are scripted/pre-programmed responses? Yep. Has anyone, such as yourself, proved it? Nope. Has the author admitted they're fake? Again, nope. So, ironically, the information contained in your answer is no more reliable than the information you're criticizing. $\endgroup$– arkonMay 24, 2018 at 10:30