Technological Progress and The Flying Car Problem
The best way to make progress is to have a problem that can't be solved without it
Tanner Greer wrote a piece that I would’ve commented on had I not gotten sick, under the title “Has Technological Progress Stalled?” He’s clearly not a journalist, because this violates Betteridge’s Law of Headlines rather badly— he believes the answer is “Yes,” and offers both a bunch of other people’s reasons for saying so, and adds a few of his own.
This is, in large part, a question of definitions and expectations, specfically “What sorts of technological advances get to count?” and “What rate of progress should we expect?” Greer and the folks he quotes are of the opinion that software isn’t real enough to count; from near the very end of the piece, he offers this:
Double-digit GDP growth means transforming the physical basis of an entire society. It means mud to concrete. Wood to steel. Sweat to dynamos. Shovels to dynamite. Wicks to lightbulbs. Carts to cars to areoplanes.
A future boom will not come from improvements in organization and information. It will be from inventing new materials to build from, new ways to move what we build, or new sources of energy to power our building. This is my yardstick for evaluating the “revolutionary” potential of new technologies. For this reason I am bearish on the long run economic impact of supercomputers. Like other information age technologies, all they do is coordinate, organize, and calibrate existing modes of production. Something like nanotechnology or bioengineered materials are more promising, for they promise to reshape the physical basis of our built environments.
(Emphasis in original)
That’s at least a defensible choice of a standard, while I’m not sure I would fully endorse it.
The expectation question is a tricky one, and one that a lot of the post is devoted to. It’s pretty clear that there was at least one extremely rapid period of technological growth in the relatively recent past, and then the question becomes when that started and whether it’s ended. Should we expect that rate of change to be sustained, in which case the current era is Very Bad Indeed, or have we returned to a different state of slower growth, in which case we’re not necessarily underperforming what we ought to expect?
This is always the point where I have a problem with these discussions, because they seem to suffer from a sort of a variant of the base rate fallacy. That is, I think it’s very difficult to have a sensible discussion about whether scientific and technical progress is “stalled” without some idea of what the rate of progress “ought” to be, and I just don’t know that there’s any good way of doing that. Which tends to lead me to shrugging and moving along when a lot of these questions come up.
What got me to keep Greer’s tab open this time around was the thought that there’s maybe a related question here that starts to get to a more reasonable answer about what to expect. That is “What problem(s) should we be looking for technology to solve?”
This is, in a sense, a variant of the educational semi-joke about learning to code, namely “The best way to learn to write computer code is to have a problem that you need a computer program to solve.” In much the same way, I think we tend not to see all that much innovation happpening totally in the abstract— we tend to get concentrated progress in areas where people are working on solving particular problems.
Seen in that light, I wonder if the issue with the “stalled progress” in things like construction and transportation isn’t coming from a lack of ingenuity on the research side or procedural barriers put in the way of clever people, but rather a lack of compelling problems that need solving. That is, our current modes of transport and methods of construction are very, very good right now, and it’s not clear to me that there are many urgent tasks that they’re not up to. Which means there’s not a huge amount of urgency to invest the resources and effort needed to push toward further breakthroughs in those areas.
Put another way, moving from carts to cars is a revolutionary change, and from cars (and trains) to planes is another, but once you’ve gone from it taking a month to cross a continent or ocean to it taking a day, I’m not sure that next order of magnitude offers as big a win. It’s the Flying Car Problem: the reason we don’t have flying cars is that the problems that would be solved by people having flying cars are less annoying than the headaches that would come from trying to make flying cars widely available.
Wood to steel is a huge upgrade in construction material, but having done that, I’m not sure what generally useful thing there is that we can’t build now because of material limits. A space elevator, maybe, but that just pushes the problem back to “What problem do we have right now that would be solved by access to a space elevator that would justify the expense of developing the needed tech?” (There’s a sort of infinite regress here, where people offer asteroid mining as a reason for a space eleveator, and then we go to “What project did you have in mind that needs a billion tons of nickel?” and so on…)
Looked at from that angle, it’s probably not all that surprising that we haven’t seen dramatic improvements in those areas: we’re not facing any problems right now that our existing technologies aren’t up to solving. The area where there is a relatively pressing problem is energy production, and not coincidentally that’s the area where we do see relatively rapid progress happening, in things like the plummeting price of photovoltaic electricity generation, and the rapid improvements in storage technology.
(At this point, you start to tip back into questions of definitions, and I have a bit of a problem with one of Greer’s sources waving off the electrical grid as not really counting because Edison would’ve understood the physics. That’s also where the “software doesn’t count” thing becomes a problem, because the kinds of active grid management you can conceive of doing with computers at some point approach a qualitative change in the system, but that’s getting a bit afield…)
So, anyway, that sort of pushes me toward thinking that a lot of the discussion of “stalled” progress in technology are a little off base, in that there isn’t really a problem that would be driving faster progress. I might feel differently if there were some urgent problem that we’re being held back from solving for lack of hypersonic flights to Asia or materials with strength-to-weight ratios well beyond what we’ve got now, but I just don’t see many of those, if any.
There’s probably a piss-off-the-string-theorists post to be written here about how the situation is more dire in fundamental theory, where there really are problems that physicists have been beating their heads against for a generation with no success. But this has gotten a little rambly already, so I’ll leave it there…
If you’d like to see if I ever do come around to pissing off the string theorists, here’s a button:
If you’ve got suggestions of problems that could be solved with a flying car or a billion tons of nickel, the comments will be open.
As the saying goes, "necessity is the mother of invention." What technologies should we expect to see progress in? Well, probably the ones need to solve pressing problems. Energy being the obvious one, as you've pointed out. Beyond that... I'd say atmospheric engineering, but we as a species haven't figured out that it's an enormous problem yet (and probably won't until it's too late). As we seen, there's been tremendous progress in unmanned aerial vehicles in warfare, because we want to project power in ways that don't involve putting soldiers in direct danger (as that is politically untenable these days). Better ways of delivering medical therapeutics (good thing mRNA vaccine technology came along when it did). But a lot of things we're working on - such as flying cars - strike me as solutions in search of problems.
>The best way to learn to write computer code is to have a problem that you need a computer program to solve.
Well, that is definitely not true. The way to learn to write code is by *playing*. Telling the computer to do weird, crazy stuff, figuring out why it doesn't work, realizing you're an idiot (always by far the most fun part) etc. but crucially never afraid of what might happen if you fail.
When you have a problem that you need solved, you don't have time to play. At that point, you either use what you already learned *by playing*, or just go copy/pasting random code and hoping it works. Programmer here, by the way, if it wasn't obvious.
And I think this ties into your broader point. It's not that we don't need better transportation (or construction), rather that we're already close to what's possible at our current level of energy generation, chemistry and engineering.
However, the fundamental advances that define that level weren't made by someone trying to build a faster train, but by some dudes in a garage wondering whether this flying thing could be made to work, or by a patent clerk wondering about the speed of light in his spare time. Einstein didn't "have a problem that can't be solved without relativity", did he?
>“A space elevator [...] “What project did you have in mind that needs a billion tons of nickel?””
Well, turning all Earth's strip mines into nature reserves sounds nice. Also, the elevator's owner gets to stop worrying about being friends with some African country (or Russia) who controls the only deposits of something or other. And of course, with negligible launch costs, orbital solar platforms start looking good. And this is just some of the obvious stuff. Zero-g factories, hospitals and bio-engineering? Low-gravity retreats for disabled/elderly? Effective ballistic missile defense?
Honestly, I think space elevators are the best example of a really useful thing we'd *obviously* like to do but can't.