Suppose you work in an office with a secretary. This secretary’s job is to manage employee calendars, plan and schedule events, run errands, and do miscellaneous paperwork. Perhaps you don’t really get along with this secretary. Perhaps every time you ask the secretary to do something for you, he becomes upset with you.

You may have found yourself wishing that the secretary could just be replaced by a computer, because you find computers easier to interact with than people. If so, then this blog is for you. The goal of this blog is to explain human interaction, using computer systems as a metaphor.

Let’s try to imagine a computer that does exactly what the secretary does: manage appointments and process paperwork for you. You might think that this situation would be easier, because computers follow rules and logic in terms of their interactions, whereas humans do not.

I think this last claim is false. I have spent years of interacting with human beings. At first, my interactions were clumsy and did not get me the results I wanted or expected.  Using this theory – that people are computers – I began improving my skills. After several years of improving, I have come to the conclusion that in most cases, humans are no more difficult to interact with than computers.

In fact, I have developed my human-interaction skills to the point where a typical human being is much easier for me to interact with than most unix systems. I believe if you can successfully use git without destroying the repository, you’re capable of handling the complexity of human interaction.

If you are very much used to interacting with unix systems, and have trouble interacting with people, this claim might sound totally false. You might think I’m saying communication is easy. I’m not saying that at all. On the contrary, I think communicating with humans is a very difficult problem, in an objective sense. Second, this difficulty is made even more of a problem because of the hardware humans use to think about each other.

Yes, communication is hard

I’m not going to argue that communication is easy. On the contrary, I think communicating with other humans is incredibly difficult, but many humans happen to be really great at it.   A great analogy here is walking on two legs. Walking on two legs is a difficult problem, but you’re probably pretty ok at it. You’re smarter than you realize!

Perhaps you think walking on two legs is easy. Most people think walking on two legs is easy. After all, two year olds can walk on two legs, how hard can it be?

Although this attitude is common, I think it’s wrong.  The implicit belief here is that two year olds are not smart, and so anything two year olds can do is, by definition, ‘easy.’   The problem with defining ‘easy’ in terms of what adult humans can and can’t do is that there’s no meaningful zero.

Imagine asking a bunch of redwood trees how tall I am. They’d just laugh at you. “Hah! This human thinks one of them tall! He’s not even on the scale of what tallness is,” those trees would say. Trees are assholes.

We don’t use a human scale for size. We use a physical scale that isn’t based solely on the limited human perspective. Our scale for height is objective. A key benefit of this objective scale is that we can use the same scale to describe the size of atoms, dishwashers, and solar systems. (Fun fact: the geometric mean of the radius an atom, and the distance from the earth to the sun, is one meter. This means there are roughly the same number of atoms in a dishwasher, as there would be dishwashers in a sphere of dishwashers centered on the sun, extending out to earth’s orbit.)

We can use a similar scale for difficulty: the complexity of the space being navigated. Tic-tac-toe has a very simple configuration space, and thus the game is easily navigated.   The space of checkers is more complex, and thus is harder to navigate. Chess and go are more complex spaces; they are harder still.

The configuration space of human language is far, far bigger.  Bipedal locomotion is probably somewhere between “go” and “smalltalk about the weather.” The fact that toddlers can walk on two feet doesn’t mean ‘walking on two feet is easy’ – it means that two year olds are brilliantly intelligent, and functional adults, even moreso.

If intelligence is like light, then let’s say human beings are about as bright as stars.  Dogs and cats are less bright, mice less bright still. While writing this article, I wagered an initial guess that the relative intelligence of dogs and cats was perhaps the output of a star the size of jupiter.  I figured a mouse might be as bright as a moon-sized star, if such a thing could exist.

Then I went and looked up the numbers, using number of neurons as a rough proxy for intelligence. Instead of brightness, I used the mass of celestial objects, to make these numbers more intuitive for me. If we say humans have intelligence corresponding to the mass of the sun, then a house mouse has intelligence corresponding to the mass of jupiter. Cats are about 10 times the size of jupiter, and dogs 50 times.  I was dramatically underestimating how intelligent mice are.

One of the things that kept coming up for me in developing this theory was that, by seeing humans as computational systems, I began to have a lot more respect and wonder, not just for humans, but for all intelligent life. Rather than seeing humans as far, far more intelligent than animals, I now see us as having crossed some simple inflection point that allowed us to have incredible leverage over them. Dogs don’t appear to think in computational clusters the way humans do, and that difference may be as big as the difference between the brain of one human and the brain of one dog. In a sense, I can think using the brain of every human who’s ever left a written record, whereas dogs can only think with their own brains, and occasional help from nearby dogs.

So yes, communication is extremely hard. The fact that “not very intelligent people” can communicate well – perhaps better than you – should be interpreted as evidence that most people are far more intelligent than we give them credit for. The fact that many children have social skills better than some adults is not evidence that these abilities are trivial. It’s not the case that there is some kind of innate skill or ability that you lack. If you’re reading this blog and you’re like I was – convinced you’re missing something special that would let you interact better with others – It’s probably for reason number two.

You’re Thinking with a Primate Brain

The human computational system – the very thing you use to do your thinking – evolved over millions of years to keep a single primate alive. Much of that evolution occurred in environments that were far more dangerous than the one you operate your machine in today. Your configuration, due to evolution, means that your cognitive machine is hyper-attuned to threats to your survival.  Your brain looks for the slightest hint of danger, and reacts, trying to protect you.

Given that primates are pack animals, social isolation and alienation can be a threat to your survival. Especially when you can’t get a new job or move to a new town. This means that upsetting, irritating, or annoying people causes your bio-computer to panic, as if you were in danger.   Historically, if you failed to interact properly with a human system, you might find yourself physically harmed, or ostracized and cut off from resources. Today, these threats are heavily mitigated by our modern culture. Yet when you fail to interact properly with a human system, your physiology might react as if you had been physically threatened. When you fail to interact properly with a computer, however you generally just get an error message.

Imagine how much harder it would be to learn how to use computers if you received a shock after inputting any invalid commands. In reality, the logical difficulty of the interface would remain exactly the same as it was; the complexity of the space would not change. The addition of pain to the process would make it much harder to remain motivated, and to have the courage to continue trying. If computers inflicted pain on anyone who gave them invalid commands, we might reward the saviest computer operators in the same way we look at the saviest human operators – as celebrities, politicians and leaders.

So yes, I think that interacting with humans is difficult, and not just because humans are so much more complex that computers. It isn’t just the case that you’re trying to navigate a space with extremely high dimensionality. You’re trying to navigate that complex space using a highly threat-sensitive machine, designed to protect a single primate in a violent, gossipy world. You’re using this skittish, miscalibrated machine to interact with other skittish, miscalibrated machines.  Of course that’s a difficult situation.

Fortunately, computers don’t tend to inflict pain on people who give them bad commands. Unfortunately, human cognitive hardware has retained this legacy behavior. The goal of this blog is to give you practice thinking about human beings in the same way you approach computer systems, so that you can become much more proficient at interacting with them in ways that will work for you. You may be acquainted with legacy techniques such as shame and guilt. These techniques are anachronisms in the modern world, and a goal of this blog is to give you far superior tools for grooming and calibrating your consciousness.

Stay tuned!  So far, we’ve simply argued that people are computers, and said “yes this is hard, but you can do it.” A more detailed model of human minds, with suggested techniques and exercises, comes next.

Further Reading

Check out ‘rejection therapy‘ as a technique for nullifying some of the unhelpful negative responses your body presents when you have an unpleasant social encounter.