ML is a subset of AI.
AI also includes speech and language processing (like those voice activated assistant thingies that people generally like), machine vision (like ANPR and self-driving cars that generally stop cars crashing) and deep learning (artificial neural networks).
Funny cynic up there in the usenix presentation makes valid points particularly around algorithm opacity and trainer bias, but don’t doubt that AI-linked technology is very relevant and addressing real world problems. Definitely a toddler in some applications and undoubtedly tech companies are falling over themselves to label their products and services AI.
The Tay chatbot thing is mildly alarming. For more odd stuff see http://aiweirdness.com
His summary statement that “technologists are detached from the ramifications of things they build” is the issue here but that’s not specific to AI. Potato hand is mint!
