“AI is a lie” – O’Reilly
On this interview from O’Reilly Foo Camp 2019, Eric Jonas, assistant professor on the College of Chicago, pierces the hype round synthetic intelligence.
Highlights from the interview embody:
Jonas argues that “AI is a lie”—which means that our expectations far outsize the truth of what’s presently doable. One of many points arising from that disconnect is a stage of company funding within the analysis course of that hasn’t been seen earlier than. He argues this has led to “so much questions on whether or not robots are going to take our jobs, and these kinds of issues, that are all fairly untimely.” (00:41)
Questions of ethics and what position it ought to play are more and more arising in machine studying and AI analysis, particularly within the space of science purposes. Whereas Jonas applauds such inquiry and pondering deeply concerning the social ramifications of AI analysis, he’s involved the questions is perhaps reinventing the wheel: “The info assortment itself typically has severe ramifications that we’ve all been wrestling with for 15 years. It’s exhausting to determine, essentially, when a linear mannequin is perhaps biased, or what precisely which means—there’s been an incredible quantity of pleasure round these questions. However I’m attempting to additionally remind everybody that the challenges there have existed for some time, and there’s numerous good analysis and scholarship on addressing them.” (2:12)