Intelligence is not Artificial: Why the Singularity is Not Coming Any Time Soon

We are bombarded daily with news of stunning achievements by Artificial Intelligence (AI) programs and promises of more to come. We are told that AI systems are getting as intelligent as us, and that soon they will be even more intelligent than us. It is worth pausing. Let’s take a deep breath, sit down and let’s do a reality check on the hype. Has there really been so much progress in the last few years? Is this progress truly accelerating? Are we heading towards an exponential singularity? And kind of singularity? If it is so scary, why do we do it at all? What is truly unique about today’s AI? What is truly reasonable to expect (and to desire) from AI?

After a tour of recurrent neural networks, reinforcement learning, generative adversial networks, walking robots, automatic machine learning, etc , we can get a pulse of the discipline. I am not afraid of AI, I am afraid that it is not coming soon enough. Expect a contrarian’s view that: 1) AI is just computational math applied to automation, i.e., good old Computer Science (call it what it is and nobody freaks out); 2) The theory of the singularity belongs to the realm of religion, not science; and 3) Automation will cause hyper-employment. There are many aspects of our technocratic society to be afraid of, but perhaps not the one you are thinking of.

Presenter(s)