AI is not new – in fact the term was coined in 1956. Why then, if it’s been around for decades, and we are now on our third major wave of AI interest and hype, do we keep making the same mistakes that have brought us to two previous AI Winters? Are we still overpromising what AI can deliver?