Mysticism and Cybernetics

Posted on Sep 25, 2022

I am writing a new paper on economic planning, and it’s interesting to observe the kind of mystical fervour certain cybernetic themes arouse. I am not sure why, but my suspicion is that cybernetics was the last effort to do anything “AI-ish” not serving moneyed interests. We have had like 70 years since these topics were introduced, and they were all more or less incorporated in RL. Their limitations are more than obvious now. In fact, even if you add every single modern development, they still lag behind the grand visions their inventors had in mind.

This much is clear now, the greats of the field backtrack,

“I think AI systems need to be able to reason,” says Yann LeCun, Meta’s chief AI scientist. Today’s popular AI approaches such as Transformers, many of which build upon his own pioneering work in the field, will not be sufficient. “You have to take a step back and say, Okay, we built this ladder, but we want to go to the moon, and there’s no way this ladder is going to get us there,” says LeCun.

The self-driving car crowd, startups such as Wayve, have been “a little too optimistic,” he says, by thinking they could “throw data at large neural networks and you can learn pretty much anything.

Yet, the legacy of the third wave of AI will be that some of the systems deployed already, in very humble roles, might form the base for a next generation of machines. We might need a 4th AI revolution 20-30 years from now to close the loop.