My father David's first posting out of Harvard Business School and the Centre d'Études Industrielles was to Alcan's sales office in New York. This was the 1950's when the three-martini lunch was a required skill. One day his boss told him he would be entertaining a big customer who liked his martinis double and in sequence. He would be expected to keep up.
His boss then offered a piece of advice that HBS had somehow failed to impart: When the customer orders his first double martini, you order gin on the rocks. That will impress the other fellow. Then, as you nurse it, the ice melts but the glass still looks full, and you keep up with the big boozer, drinking mostly water.
David spent a lot of time in Madison Avenue eateries those days, and if the conversation fell short it didn't matter, because the eavesdropping was great. Think Mad Men.
So there came a day when two guys in the next booth were having a heated discussion. They were obviously partners in an ad agency, and they were arguing about whether to hire someone. The point in favor was how intelligent the prospect was. One man remained unconvinced. Finally, he said in exasperation:
“OK, OK. I admit, the kid is smart. But how does he stack up, wisdom-wise?”
AI is hot. It's the next big thing. It's machine learning that's so fast that humans can't keep up. It's tireless workers minting profits at the speed of silicon.
I met Marvin Minsky at the MIT Experimental Music Studio, a precursor of the Media Lab. He was a great guy: interesting, charismatic, challenging, musical, and, you got the feeling, delightfully wrong. Minsky was one of the great early leaders of artificial intelligence research, and he saw it not only as a technological pursuit, but rightfully, as a philosophical one.
One day in the summer of 1982 Minsky came down to talk to a group of computer music composers. He happened to mention that his team had written code to compose Bach. Before leaving, he invited us upstairs to visit his lab.
Up I went. I was really curious to hear some AI Bach. I mean, that's gutsy, exposing your theories to a test that mapped what your machines could do against the vast and deep capacities of a human being listening to music.
When I got there, I looked around for the speakers. There were none. How, I asked, do you know if the machine's Bach is any good if you don't listen to it? Ah, they said, we have another program that analyzes Bach, and it tells us.
That's the day I learned that when it comes to Artificial Intelligence, the first word is more important than the second.
I picture two Mad Men arguing today about AI. I picture one saying how smart the machine is. I hear the other saying: “OK, OK. I admit the robot is smart. But how does it stack up wisdom-wise?”
I want a term we can all use to highlight the wisdom — that is ever-present in this world — that machines will never possess.
I want this term to highlight all the intelligence around us that is not machine learned. I want it featured as much as AI, so we don't lose hope and perspective. I am thinking: Organic Intelligence.
Organic Intelligence is not just about humans. It's also the resilience of ecosystems, and the synergies of countless living systems in proximity. It's climate, and intuition, and music.
Organic Intelligence is something that is fundamentally worthwhile. We don't wonder why the weed grows, even if we try to discourage it. We will use AI to do worthwhile things and we will use it to do worthless things. Any entity possessing organic intelligence will never be worthless.
Organic Intelligence is something that is based in real time. For ecosystems, it's years and millennia. For humans, seconds, days, and years. Machine intelligence that exists out of real time is categorically unreal.
Organic intelligence emerges through two synchronizing processes: Being and creating. These don't need to be engineered. They just are.