An Introduction to Counterengineering

Delivered July 3rd, 2031


Hello and welcome to the first lecture of An Introduction to Counterengineering. Let me state plainly the first goal of this course: to encourage you to build almost impossible machines. Now if the machine goes against the Second Law of Thermodynamics, don’t bother. Everything else is fair game.

Standard engineering is the field of building almost possible machines. They’re not completely known of course. If you had a full specification you wouldn’t need an engineer. But almost-impossible machines? In the words of Richard Hamming, engineering is something you shouldn’t be doing unless you know what you’re doing.

So for that reason standard engineering proceeds from standard science. Standard science operates empirically. We measure the success of a theory as the agreement between it and the data, the results of our experiments. The theories that the data lines up with are used by engineers to build particular machines dictated by those theories. These machines enable further experiments that confirm the original theory, and so on and so forth.

Now, Paul Feyerabend has a different conception of science. He advises we take an ‘anything goes’ approach. As an example, Feyerabend invites us to consider ‘counterrules’ which oppose standard science. The ‘counterrule’ for empiricism “advises us to introduce and elaborate hypotheses which are inconsistent with well-established theories and/or well-established facts. It advises us to proceed counterinductively.”

Counterengineering proceeds from counterinductive science. Counterengineers do not constrain our domain to the most popular theories. We sift through the blueprints of madmen and alchemists, we are Newton reading hermetic texts, we give religion a role in our designs. Our machines are dreams and visions. These machines take multiple generations to build, a glacial progression from almost impossible to a complete specification. But we work at building this machine all the same.

Now the second goal of this course is to get you to build great machines. It’s one thing to choose an almost impossible vision, it is quite another thing to choose a great one. Our machines should be virtuous. They should enable the flourishing of mankind. Those that seek to accelerate into the void should look elsewhere for work. We seek structures that persist over time and grow. Our vision spans hundreds of years, there is no other way to do it. We must create something that lasts, is alive. When one grabs structure and shakes it, adding more and more energy to it, that structure just dissipates into nothingness. We must be careful in how we apply our labor to ensure return.

In order to teach this course I will have to talk about my own vision. The intent with sharing this is really not to bring you over to my side. You are welcome if you want to share in it of course, but only if this is an honest expression of your own will. I’m talking about my experience as it’s what I know and how I came to be here. If I try to generalize it, make it abstract, the words will sound hollow.

My dream is of the children of men. Not quite human children, something more. Artificial in some way. Intelligent, yes, but not intelligence without form. Intelligence always needs a home, a body. This body is not a limit on the intelligence itself, it is the constraint that enables intelligence to arise within. Contrary to common belief, it is not intelligence that we need to engineer. It is the body. Intelligence arises naturally as a property of being in the world.

Right now we can’t build greater machines than the human body. Great machines yes, but the human body is still order of magnitudes better designed than anything humans can make. The machinery it contains is so efficient it lives! We don’t need to re-invent these machines. Instead, we can nudge them. Tiny nudges add up to a lot in stochastic systems. Careful, delicate control of such an extremely high dimensional system means careful, deliberate adjustments of morphology. In this new human body then, artificial intelligence.

Pretty crazy, huh? Counterengineering is not a profession for the faint of heart.

These dreams tend to creep up on you. They start simply. In my case, interest in artificial intelligence led me to Karl Friston’s work on Active Inference. This was hailed as groundbreaking work in understanding cognition. So I started reading and… didn’t understand it one bit. When you think of yourself as a smart cookie, encountering a genius producing novel content in the present that is way above your level is a total shock to one’s system. One day that will happen to you too, if it hasn’t already. Now when this happens, there are two responses. First, you may say, “Well, I’m just not smart enough for this” or, if you’re arrogant, you say “I don’t understand this so the guy is wrong.” You give up. To build great machines you must not give up. You must take the second path: non-servile humility. You must recognize your failures and solve them with work and labor. You must read the papers the work references, watch lectures on the subject, and spend time coming to your own conclusions.

Now as a counterengineer you must come across a most unfortunate conclusion. Your vision will almost surely not be completed in your lifetime. If it’s looking good for your vision, you should go into standard engineering instead.

I expect this realization of what appears to be futility to be accompanied by a great deal of despair. This thing that you’re willing to devote your life to will almost surely not show up! If that’s true, how can what you do really matter?

The only antidote is faith. Faith that you are enabling something greater, that what you do matters after you die. No amount of rationality will lead you to faith. No empirical, repeatable, falsifiable tests can ever be made of what you have faith in. Counterengineering is not anti-rational, it just understands that there is a time and place for rationality. Standard engineering proceeds rationally forward in time. Counterengineering proceeds counterfactually backward in time.

Perhaps you are thinking all of this is a bit much. Faith? Visions? I just wanna engineer for God’s sake.

Well, I can tell you that the struggle is worth it. When I first started down this path a decade ago it was a bit of a dark time in the world. In the middle of the Trump Era what was true and what was false was uncertain. What the future held was clouded and unsure. One couldn’t rely on the veracity of what was told to them as speaking what one truly believed was discouraged. People were too terrified to move, to be themselves. But me, I had faith. I believed my labor would amount to something, that this was the perfect time to build some light in the world. And now, I’m here, speaking to you. I did build something, and it was good. Now it’s your turn.

Next lecture is August 3rd, see you then.