This blog is about computers, but it is also about AI. And AI is about intelligence, and intelligence is about humans, and what we do. I have always thought that studying humans, while not essential to AI, can make a lot of things much easier. And humans are interesting to study in their own sense.

One thing I have always been fascinated by is this: Why do some succeed and others fail? Why do exceptional people have such a difficult time training their children to be as exceptional as themselves?

I have been trying to make myself exceptional for years, and so I have been relentlessly attempting to answer this question: what makes someone exceptional, i.e. far beyond the norm in output.

I go to an elite liberal arts college at the time of writing, Reed College, and there are quite a few exceptional people here. And, of course, I keep tabs about what parts of me and my output that others think are exceptional. Based on this evidence, I have come to the following conclusion:

A person’s degree of exceptionalism is almost entirely determined by the degree of a person’s relative emotional response to their work. And a person’s emotional response to their work is only at an exceptional degree when people have a deep, personal reason why they do the thing they do

There are a few potentially confusing things I want to clear up first.

One, an “relative emotional response to their work” does not have to be broadcasted or expressed. Some exceptional people I have met are extremely emotive; you can see their eye’s light up when they do something right, and look sad when things go wrong. This does not hold for everyone, though. Some people I have met simply don’t have that much emotion to begin with, so the simple frustration of failure manages to get them pretty far. Others do have really deep connections with their work, allowing them to go all the way to the top, but their body simply doesn’t respond to that emotion.

Another issue is that I am talking about individual reasons relatively unconnected to other people. In many ways, this is a personal bias, as my own motivation is deeply personal. In reality, most people get most of their motivation to work from other people. Family, friends, role models help us all find the motivation to work on the things we do. My argument is not that these are real, or that they are powerful enough for people to make it close to the top. Instead, my argument is simply that in order to be truly exceptional, your emotions must be strong enough to endure separation from specific individuals. Say you have a deep desire to protect people, and so you become extremely sensitive to dangerous situations. Say you became this way because you are a parent. Then I argue that if your emotion will almost certainly continue after the nest is emptied. You will find other people to protect.

Another confusion is what the nature of this “deep, personal reason” is. I believe that the strongest reasons only occur when we align our experiences with our nature. I believe that I was born deeply afraid of this world, and particularly the people in it. A lifetime of caring support has so far failed to comfort me to the degree where I can relax around people I don’t know extremely well. I have only met a small number of people with a similar level of natural social anxiety. This is an example of a inborn reason, the nature side of the possibility. However, I strongly believe that nurture has equal, if not greater power to make deep and powerful motivations. One of the most astonishing people I have ever met, in fact, probably the most exceptional person I have gotten to meet up close is a math teacher at Reed College, named Jerry Shurman. He has little of what you expect from a great teacher in terms of personality or demeanor. He is tense, nervous, unpredictable, and widely accepted as the best math teacher imaginable. How is this possible? Well, Jerry (we call teachers by their first name at Reed) has a story he likes to tell. Actually he has several of them, one for each level of class. I only took a 300 level class from him, so I head the grad school story. The story is this. In undergraduate school, Jerry found that he loved the mathematical puzzles in number theory. He loved that things were always right or wrong, and that he could always make sure he was right. He pulled an all nighter the day a assignment was given out, and then turned it in perfect and complete the next day. With this natural enthusiasm, he got into Princeton grad school for mathematics. If you know anything about high level math you can probably guess what happened. Jerry found himself lost and confused. The problems he loved solving weren’t even on people’s radar. They were 18th century puzzles, now we are in the 21st century. He was given 3 hour tests where he couldn’t solve a single problem. He buried himself for weeks in checking the work of the paper he read, while his fellow students skimmed over it and had ideas of how it connects to other work in a matter of hours. It took him 10 years to graduate, and even then he felt like he had not truly caught up. For someone who loves doing things right, you can imagine that this was an extremely traumatizing experience. Traumatizing enough, that even now, over 20 years later, you still see the horror he experienced in his eyes when telling this story. And so every day, he stands in front of class, and he has a desperate desire that we do not experience the same troubles. So what does he do differently? He watches us, closely. During lecture, when he is tired from lack of sleep, and unfocused, he still notices when we don’t seem to be getting what we are saying. And that horrifies him, so the next day, he explains the parts that confused us perfectly, showing exactly what we needed to understand what was confusing us. Many good teachers write up lecture notes. Many textbooks were born from those notes. Most teachers do not constantly obsess about specific wording and idea organization for hours. Most teachers do not feel compelled to write several explanations to help students of different backgrounds and levels. Most teachers do not feel that every slight potential confusion, or difficult to read section is a horror. And so Jerry’s notes are astonishingly good, far better than most published textbooks. Jerry is still not a great math researcher. There are other Princeton graduates in the math department faculty who are far more brilliant. But none of them have the same powerful drive to teach, as none of them struggled endlessly with the horror of being taught poorly.

Ok, now that story time is over, what can we learn from this? What can we say about humanity, and intelligence, and AI.

In a way, this observation shouldn’t be that surprising, from an intelligence perspective. If emotions drive how we learn and what we do, then clearly those who have stronger emotional responses will learn faster, and work harder. They will be willing to put themselves through more pain.

But there is a lot we can learn from this. For example, I have noticed that older people in general have stronger connections than younger people. Possibly because they found meaningful work? Possibly an inertial effect? Who knows.

a