A Perfectly Rational Being

Over the years, I’ve played with a thought experiment: suppose there were a purely rational being who had no instincts, emotions, or irrational stimuli of any kind. Something we imagine robots and Vulcans to be like. What would such a being do? In the time it has taken me to get around to writing this, I have gone through three phases of what I thought the answer would be. So let’s dive in and imagine what a person/creature/AI would be like if they were perfectly rational.

Phase I: The Rational Statue

When we talk about rational behaviors, we usually mean things that advance our careers or keep us alive. That’s logical, right? But we run into a problem: the is-ought gap. Reasons to do things ultimately come down to instinct, not factual observations. Sure, we may need to stay alive in order to do anything, but without any drive to do anything, we have no reason to keep ourselves alive. Under these conditions, self-preservation is no more rational than self-destruction.

You might ask whether we already have perfectly rational beings in artificial intelligence. Isn’t a robot perfectly rational? Not at all. A computer program does not think about what it is told to do, it simply runs its programs, taking input and giving output. It is one hundred percent instinctual, not rational in the slightest. A perfectly rational being would indeed have the ability to give you the answer to any math or logic problem quickly and correctly, but, having no motivation, they wouldn’t, because they would have no reason to do so. A perfectly rational being would do absolutely nothing.

Phase II: The Rational Egotist

Then I thought of something that turned this entire argument upside-down: a perfectly rational being would realize that it might be modified sometime in the future to have instinctive, irrational motivations. Therefore, not knowing what those motivations would be, it would act in such a way as to keep as many options open as possible.

To start with, this would mean staying alive. It would also accumulate resources, including wealth, influence, and information. It will make deals and build trust, do favors so it can call in returns later. However, it will also screw others over when it calculates a sufficiently high probability to get away with it. After all, its ultimate goal is to open up future options; it has no sense of morality nor respect for laws or social conventions.

Phase III: The Rational Altruist

However, there is yet another major factor that turns everything upside down. A being who is perfectly rational will understand that individuality is an artificial construction. In a sense, every living thing that exists is an extension of one person, the universe. Thus, a perfectly rational being would not have to wait around for some unknown motivation in the future; it already has motivations, the motivations of other people.

Thus, a perfectly rational being with no motivations of its own would work for the benefit of others. It would not follow the whims of a master or people ordering it about, nor invest in caring for one particular child or homeless person at a time. Rather, it would be impartial and utilitarian, putting its efforts where they would do the most good. To that end, most of its efforts would be focused on medical research, alleviating extreme poverty, eliminating monstrosities like dictatorships and factory farms, and preventing extinction.

So that is how I believe a perfectly rational being would behave. The phases are hierarchical, that is, each phase takes into account all of the information of the previous phases. It’s like changing religious or scientific paradigms, the being will not slide back into a previous phase unless they receive new information that takes into account all of the information they already have. I don’t know how practical this thought experiment is, but at least it was fun. Maybe I’ll write an artificial character who goes through these phases in one of my books someday.