Naïve Ng

What an incredibly naive statement Andrew Ng , that we will see it coming when AI wants to deceive us. AI is already deceiving us in so many ways, from social algorithms (think TikTok) that lead to addiction and isolation, to mis/disinformation enabled by AI, to credit scoring, predictive policing, etc. These are all examples of rather "primitive" AI. When a person wants to deceive us using deceiving AI, we won't (don't) see it coming. When a company wants to get something from us (attention, likes, money, votes) using deceiving AI we won't (don't) see it coming. When an AI uses another AI to obtain something using indirect deception, we won't see it coming. It does not take a huge amount of deception to achieve terrible results. In fact, small, virtually undetectable tokens of false or biased information are enough to erode anything over time. I love AI, but I think Aza Raskin and Tristan Harris make great points about how insidious deception can be. And I love Andrew Ng as an educator but your philosophy of AI is not where it needs to be.