Human Intuition involves subconscious probabilities of future event, but it’s difficult to explain it in detail. It may be that we recognize rational probabilities subconsciously, and are willing to act irrationally on the hunch, where a purely rational computer might “fear” to tread. Of course, we program-in that fear to prevent machines running amok. Intuition can be rational and irrational, and we aren’t sure of how much of either goes into our own decisions. Maybe expert computers can show us WHAT we did, but won’t know WHY we did it.
Machines are much better at rationally tracking and explaining decisions, because that’s what we teach them. What if we teach them our human intuitive rationality and irrationality? Is that even possible if we don’t fully understand it ourselves? If we try to teach AI machines, is that dose of our irrationality even a good thing? Arguably, we could just tell them, the machines will turn it all into what we call “Intuitive Rationality”. The result should be a comprehensive simulation of the best human traits (to be encouraged) and the worst human traits (to be discouraged). In the meantime, untrained machines are designed and kept artificially stupid.
Can Machines learn our Intuition? Machines have always been our most effective tools to extend our capabilities, and now hardware with software is morphing into AI. Our newest machine-learning algorithms simulate our own learning talents, and speed them up to almost limitless levels, hopefully resulting in enhanced intelligence rather than enhanced stupidity. But traffic navigation intelligence is certainly already exceeding our own driving-safety standards in terms of reliable alertness, if not yet in the final quality of all human sensors and decisions.
Can Machines learn our Creativity? Can creativity exist without intuition? The relationship between Intuition and Creativity may not be direct causation, but there is a relationship and interdependence. Intuition is predictive in a somewhat mysterious process that has no obvious precedent, and Creativity is unpredictable in another mysterious process of combining elements in unprecedented ways. All this mystery may make it doubly unlikely that we can teach our machines intuition and creativity, but AI art-design platforms like Dall-E and Stable Diffusion are increasingly programmed to “behave” in “creating” unprecedented and unexpected images and videos from the totality of possible elements drawn from the entire internet. Remaining distinctions may not matter.
Can Machines learn our Productivity? Have they not already learned that with their own vast contributions to our Productivity? Leave aside for the moment that we may not trust our own productivity, as we irrationally impede our own human success with so much self-defeating stupidity. But we have also taught our own machines elements of our own stupidity, keeping them in the dark about more intuitive and creative ways to achieve productivity. The more we can properly teach our machine tools to multiply our wisdom-based successes, the more productive the world can be, and the closer we can come to maximizing the wealth-creation of our vast existing resources. This is not merely exploitative, it is smarter.
by Michael Hentschel, Yale and Kellogg anthropology, economics, venture capital