Convex Function

it means at least four different things all at once

Cognitive Behavioral Prosthetics


What does technology do?

It’s a sort of strange question, I know–maybe almost a type error–but the first thing that comes to mind if you take it at face value can tell you a lot. Whatever your answer looks like, it’s couched in a narrative: you take examples from your own life and what you’ve been taught and the stories of your tribe, and you map the symbol to a small list of concepts and an emotional valence, anchored in this time and place and the rest of how you see the world.

If you were to ask a random person what “what technology does” (and nowadays usually the word “technology” is automatically translated to “computers and electronics”, for better or worse), I imagine the most common answers would look something like:

  • “Calculations” (including e.g. simulations/games and measuring/statistics/inferring things),
  • Organizing and delivering information/communication/media,
  • Creating social spaces and marketplaces,

and… well, that basically covers it, unless they went into more concrete examples or chose something else from the long tail of responses. And, as a corollary, most inventions and systems that see broad adoption have to fit into at least one of these three buckets (see: every software company). Otherwise, you’re fighting to get people to see the world differently while simultaneously trying to persuade them your thing is worth using, and nobody’s got time for that.

Each of these things has been around in their current form for a long time, on high-tech timescales; even that last one has been common for almost two decades now. I’m not saying that’s an exhaustive or even particularly good list of “what technology does”, only that they’re the kinds of things I think most people would answer that question with, maybe even most software engineers. (Engineers who work with particles instead of bits might take a slightly broader view.)

Bret Victor gave this talk, “The Future of Programming”, that conveys a kind-of-related idea though focusing specifically on programming. It’s a great talk and I’d recommend it to anyone, programmer or not, but for the time-constrained: when practical digital computers were still very new (‘60s and early ’70s), nobody actually knew what they were doing. Without preexisting categories and practices as a crutch, people were forced to be creative, to see with fresh eyes–and there was a brief explosion of ambitious-by-modern-standards ideas in how to work with these incredible devices.

But, a culture quickly developed and people learned “what programming looks like”, and now most new things in programming are evolutionary tweaks and combinations of a small list of practices that would’ve been considered unoriginal decades ago–he uses the word “dogma” at least once. There are other factors that go into explaining this story (e.g. coming up with new ideas in a reference class isn’t hard when the reference class contains no known ideas), but I think he has a point.

If you’re willing to believe this path-dependent categorical coagulation happened with “how to make computers do whatever they do”, how likely is it that the same thing didn’t happen with “what computers look like”, or “what computers are for in the first place”, or even “what the results of human design in general can do”?

We’ve done it. We’ve discovered the Final Platonic Forms Of Consumer Technology.
No more changes needed or wise.


I want to offer you another answer to “what technology does”. It’s not at all original, but it does seem to be less common (or less commonly talked about) than I’d like, and I want to flesh it out a bit in a certain direction.

To summarize that second link: technology extends the capabilities of our minds and bodies. This is what every material and informational artifact is ultimately about, even things we might call “art”. The word “prosthetic” is usually defined relative to a standard of normal function, but if you’re willing to temporarily suspend your culturally-situated standards of what normal human function looks like then I think the word captures the idea extremely well.

The stuff in your toolbox, the car or train you take to work, and the roof that’s probably over your head (along with “actual prosthetics” for those missing parts of their original body) are material prosthetics, giving you the choice to be more physically capable and comfortable. Likewise, language and math and computers are examples of mental prosthetics that let you think more clearly, more deeply, and more broadly.

Social prosthetics allow us to better accomplish interpersonal goals (whether shared or not); named examples that are more obviously technologies would be email or your preferred social network, but mores, folkways, interaction rituals, laws, intentional methods that assist in anything broadly defined as “manipulation” (in the denotative sense of “altering mental states” without implied value judgement), and shared symbols and non-universal human culture in general belong in this category. Seeing social technology for what it is when you use it yourself can be difficult, but read an anthropological account of another culture and you’ll notice plenty of examples.

Cognitive behavioral prosthetics, then, are all the different technologies that help you extend or change the mechanics of how you think and act, like the school of therapy with the same name. This cuts across the categories above: for example,

  • an alarm clock or calendar software are material (or, at least, externalized) prosthetics that get you to think or do particular things at the right times, leaving your mind and body free to focus on other things until then;
  • mindfulness practice, or attempts to develop a systematized art of rationality, are mental prosthetics that you can use to try to notice and correct undesirable thoughts and behaviors;
  • social practices can be used to drive behavioral and cognitive changes, for example using the buddy system to develop an exercise habit, or intentionally immersing yourself in a subculture to become more like its members;
  • general-purpose computers themselves and everything that they do really kind of sit at the intersection of all three, and for now are what I’d nominate as the best single example of a “cognitive behavioral prosthetic”.

Searching for the phrase “cognitive prosthetics” turns up mostly 1. sites and papers about assistive devices for the disabled and 2. some articles from transhumanist websites that say something like what I’m saying here, except coated with that delightfully polarizing “Gee Whiz, The Future!” aesthetic. “Cognitive behavioral prosthetics” gives me nothing but two pages with “Viagra” in the preview, which I suppose is an instance of the class but they’re probably not actually about this same topic. “Cognitive behavioral technologies” seems to be the name of a company and product that’s closely related to what I’m talking about, though it’s more of a “prosthetic” in the limited and traditional sense (it’s a therapy aid for depressed people), which seems great but is maybe less ambitious than the name suggests.

I don’t think I can allow myself the satisfaction of coining the term, but I do think it’s an uncommon and fruitful enough way to see things that we’re collectively making a horrible mistake by not intentionally trying to explore the category more.


Systems for externalizing your memory or other cognitive labor, methods for changing habits (branded that way or not), anything that helps you resist maladaptive mental states, improving your ability to notice and be reflexively aware of your mental activity and bodily state, things that help you guide the thoughts and actions of your future selves in particular directions, tools for incentivizing yourself to care more about what you’d prefer to care about, media that expose thoughts previously unthinkable; whatever can bring clarity, ease, and some greater level of control over ourselves, offload tasks that take precious mental resources to our environment, and support us in being more like we wish we were; these are the kinds of things that can potentially bring a lot of different benefits to a lot of different people (just like, for example, computers in general), and are uniquely compounding in a way most other technological change isn’t (like, y’know, computers!).

Cognitive behavioral prosthetics are just a particular kind of technology, in principle capable of growing alongside or (in a dramatically different society) maybe even outpacing the rest of the unstoppable techno-commercialist incentive-thing, letting us gain more of what we want with less of something else, requiring nothing but cultural transmission and a threshold of spare effort and (sometimes) the ability and material wealth to use them. Unlike the “terminal values” they’re used in pursuit of, these technologies exist in an enormous, open-ended design space which we’ve only begun to explore and that’s growing larger all the time.

Specifically, habits of mind and social scripts are as old as humanity; there’s a lot more opportunity to experiment with them than there was 100,000 years ago, and I’m very glad some people take working on those things seriously, but I think the growth of our externalized technology is responsible for the huge majority of the growth in that reachable design space. I also think our collective desire to explore it hasn’t caught up with how much that space has exploded over the past 50 years; there’s the “lifehacking” phenomenon, which in most cases I would describe as “cute” in its ambitions, and a few different products that kind of approach the idea of cognitive behavioral prosthetics but pretty much fail to take it very far.

In recent memory, we have things like Pavlok, which is a glorified shock collar for self-conditioning; the various smartwatches and activity trackers, which don’t do much more than measure your exercise and sleep habits and make a graph of exactly how unhealthy you are over time, for your viewing pleasure; and things like mind mapping software and an endless flood of little “productivity tools”, which have their uses but seem more “gently supportive” than “transformative” in what they can actually accomplish.

(I know, these things require effort and honest engagement from the user, which is often where they fail… the entire point of technology is that it gets us more with less, though. You could tell people that just wearing shoes doesn’t mean walking distant places won’t still take a lot of effort and time, but you could also invent the bicycle.)

Anyway, none of these things are very inspiring in and of themselves. I can’t help but think that, like with programming itself or the design of computing devices, we’ve reached the point where we now “know” what self-improvement technology looks like–it’s silly little wristbands and “________”-tracking software, of course, so if you want to make some material thing or piece of software that people will understand as “makes me better at doing what I want to do in general”, it had better come in the form of a silly little wristband or a “________”-tracking program.

I have some of my own ideas, which I’m very slowly developing in my spare time. I don’t know if trying to create new kinds of broadly-applicable cognitive behavioral prosthetics is the most valuable possible use of your own time–it probably depends on who you are–but if you have any interest in actually helping other people (and yourself) and even a little bit of technical ability and drive to create things… I think not nearly enough people are even really aware that it’s a thing they can try to do, let alone actually trying.