Technological stagnation: Why I came around

[jasoncrawford]

Link post

“We wanted flying cars, instead we got 140 characters,” says Peter Thiel’s Founders Fund, expressing a sort of jaded disappointment with technological progress. (The fact that the 140 characters have become 280, a 100% increase, does not seem to have impressed him.)

Thiel, along with economists such as Tyler Cowen (The Great Stagnation) and Robert Gordon (The Rise and Fall of American Growth), promotes a “stagnation hypothesis”: that there has been a significant slowdown in scientific, technological, and economic progress in recent decades—say, for a round number, since about 1970, or the last ~50 years.

When I first heard the stagnation hypothesis, I was skeptical. The arguments weren’t convincing to me. But as I studied the history of progress (and looked at the numbers), I slowly came around, and now I’m fairly convinced. So convinced, in fact, that I now seem to be more pessimistic about ending stagnation than some of its original proponents.

In this essay I’ll try to capture both why I was originally skeptical, and also why I changed my mind. If you have heard some of the same arguments that I did, and are skeptical for the same reasons, maybe my framing of the issue will help.

Stagnation is relative

To get one misconception out of the way first: “stagnation” does not mean zero progress. No one is claiming that. There wasn’t zero progress even before the Industrial Revolution (or the civilizations of Europe and Asia would have looked no different in 1700 than they did in the days of nomadic hunter-gatherers, tens of thousands of years ago).

Stagnation just means slower progress. And not even slower than that pre-industrial era, but slower than, roughly, the late 1800s to mid-1900s, when growth rates are said to have peaked.

Because of this, we can’t resolve the issue by pointing to isolated advances. The microwave, the air conditioner, the electronic pacemaker, a new cancer drug—these are great, but they don’t disprove stagnation.

Stagnation is relative, and so to evaluate the hypothesis we must find some way to compare magnitudes. This is difficult.

Only 140 characters?

“We wanted flying cars, instead we got a supercomputer in everyone’s pocket and a global communications network to connect everyone on the planet to each other and to the whole of the world’s knowledge, art, philosophy and culture.” When you put it that way, it doesn’t sound so bad.

Indeed, the digital revolution has been absolutely amazing. It’s up there with electricity, the internal combustion engine, or mass manufacturing: one of the great, fundamental, transformative technologies of the industrial age. (Although admittedly it’s hard to see the effect of computers in the productivity statistics, and I don’t know why.)

But we don’t need to downplay the magnitude of the digital revolution to see stagnation; conversely, proving its importance will not defeat the stagnation hypothesis. Again, stagnation is relative, and we must find some way to compare the current period to those that came before.

Argumentum ad living room

Eric Weinstein proposes a test: “Go into a room and subtract off all of the screens. How do you know you’re not in 1973, but for issues of design?”

This too I found unconvincing. It felt like a weak thought experiment that relied too much on intuition, revealing one’s own priors more than anything else. And why should we necessarily expect progress to be visible directly from the home or office? Maybe it is happening in specialized environments that the layman wouldn’t have much intuition about: in the factory, the power plant, the agricultural field, the hospital, the oil rig, the cargo ship, the research lab.

No progress except for all the progress

There’s also that sleight of hand: “subtract the screens”. A starker form of this argument is: “except for computers and the Internet, our economy has been relatively stagnant.” Well, sure: if you carve out all the progress, what remains is stagnation.

Would we even expect progress to be evenly distributed across all domains? Any one technology follows an S-curve: a slow start, followed by rapid expansion, then a leveling off in maturity. It’s not a sign of stagnation that after the world became electrified, electrical power technology wasn’t a high-growth area like it had been in the early 1900s. That’s not how progress works. Instead, we are constantly turning our attention to new frontiers. If that’s the case, you can’t carve out the frontiers and then say, “well, except for the frontiers, we’re stagnating”.

Bit bigotry?

In an interview with Cowen, Thiel says stagnation is “in the world of atoms, not bits”:

I think we’ve had a lot of innovation in computers, information technology, Internet, mobile Internet in the world of bits. Not so much in the world of atoms, supersonic travel, space travel, new forms of energy, new forms of medicine, new medical devices, etc.

But again, why should we expect it to be different? Maybe bits are just the current frontier. And what’s the matter with bits, anyway? Are they less important than atoms? Progress in any field is still progress.

The quantitative case

So, we need more than isolated anecdotes, or appeals to intuition. A more rigorous case for stagnation can be made quantitatively. A paper by Cowen and Ben Southwood quotes Gordon: “U.S. economic growth slowed by more than half from 3.2 percent per year during 1970-2006 to only 1.4 percent during 2006-2016.” Or look at this chart from the same paper:

Growth in US GDP per capita. Cowen & Southwood

Gordon’s own book points out that growth in output per hour has slowed from an average annual rate of 2.82% in the period 1920-1970, to 1.62% in 1970-2014. He also analyzes TFP (total factor productivity, a residual calculated by subtracting out increases in capital and labor from GDP growth; what remains is assumed to represent productivity growth from technology). Annual TFP growth was 1.89% from 1920-1970, but less than 1% in every decade since. (More detail in my review of Gordon’s book.)

Analyzing growth quantitatively is hard, and these conclusions are disputed. GDP is problematic (and these authors acknowledge this). In particular, it does not capture consumer surplus: since you don’t pay for articles on Wikipedia, searches on Google, or entertainment on YouTube, a shift to these services away from paid ones actually shrinks GDP, but it represents progress and consumer benefit.

Gordon, however, points out that GDP has never captured consumer surplus, and there has been plenty of surplus in the past. So if you want to argue that unmeasured surplus is the cause of an apparent (but not a real) decline in growth rates, then you have to argue that GDP has been systematically increasingly mismeasured over time.

So far, I’ve only heard one only argument that even hints in this direction: the shift from manufacturing to services. If services are more mismeasured than manufactured products, then in logic at least this could account for an illusory slowdown. But I’ve never seen this argument fully developed.

In any case, the quantitative argument is not what convinced me of the stagnation hypothesis nearly as much as the qualitative one.

Sustaining multiple fronts

I remember the first time I thought there might really be something to the stagnation hypothesis: it was when I started mapping out a timeline of major inventions in each main area of industry.

At a high level, I think of technology/​industry in six major categories:

• Manufacturing & construction

• Agriculture

• Energy

• Transportation

• Information

• Medicine

Almost every significant advance or technology can be classified in one of these buckets (with a few exceptions, such as finance and perhaps retail).

The first phase of the industrial era, sometimes called “the first Industrial Revolution”, from the 1700s through the mid-1800s, consisted mainly of two fundamental advances: mechanization, and the steam engine. The factory system evolved in part out of the former, and the locomotive was based on the latter. Together, these revolutionized manufacturing, energy, and transportation, and began to transform agriculture as well.

The “second Industrial Revolution”, from the mid-1800s to the mid-1900s, is characterized by a greater influence of science: mainly chemistry, electromagnetism, and microbiology. Applied chemistry gave us better materials, from Bessemer steel to plastic, and synthetic fertilizers and pesticides. It also gave us processes to refine petroleum, enabling the oil boom; this led to the internal combustion engine, and the vehicles based on it—cars, planes, and oil-burning ships—that still dominate transportation today. Physics gave us the electrical industry, including generators, motors, and the light bulb; and electronic communications, from the telegraph and telephone through radio and television. And biology gave us the germ theory, which dramatically reduced infectious disease mortality rates through improvements in sanitation, new vaccines, and towards the end of this period, antibiotics. So every single one of the six major categories was completely transformed.

Since then, the “third Industrial Revolution”, starting in the mid-1900s, has mostly seen fundamental advances in a single area: electronic computing and communications. If you date it from 1970, there has really been nothing comparable in manufacturing, agriculture, energy, transportation, or medicine—again, not that these areas have seen zero progress, simply that they’ve seen less-than-revolutionary progress. Computers have completely transformed all of information processing and communications, while there have been no new types of materials, vehicles, fuels, engines, etc. The closest candidates I can see are containerization in shipping, which revolutionized cargo but did nothing for passenger travel; and genetic engineering, which has given us a few big wins but hasn’t reached nearly its full potential yet.

The digital revolution has had echoes, derivative effects, in the other areas, of course: computers now help to control machines in all of those areas, and to plan and optimize processes. But those secondary effects existed in previous eras, too, along with primary effects. In the third Industrial Revolution we only have primary effects in one area.

So, making a very rough count of revolutionary technologies, there were:

• 3 in IR1: mechanization, steam power, the locomotive

• 5 in IR2: oil + internal combustion, electric power, electronic communications, industrial chemistry, germ theory

• 1 in IR3 (so far): computing + digital communications

It’s not that bits don’t matter, or that the computer revolution isn’t transformative. It’s that in previous eras we saw breakthroughs across the board. It’s that we went from five simultaneous technology revolutions to one.

The missing revolutions

The picture becomes starker when we look at the technologies that were promised, but never arrived or haven’t come to fruition yet; or those that were launched, but aborted or stunted. If manufacturing, agriculture, etc. weren’t transformed, then how could they have been?

Energy: The most obvious stunted technology is nuclear power. In the 1950s, everyone expected a nuclear future. Today, nuclear supplies less than 20% of US electricity and only about 8% of its total energy (and about half those figures in the world at large). Arguably, we should have had nuclear homes, cars and batteries by now.

Transportation: In 1969, Apollo 11 landed on the Moon and Concorde took its first supersonic test flight. But they were not followed by a thriving space transportation industry or broadly available supersonic passenger travel. The last Apollo mission flew in 1972, a mere three years later. Concorde was only ever available as a luxury for the elite, was never highly profitable, and was shut down in 2003, after less than thirty years in service. Meanwhile, passenger travel speeds are unchanged over 50 years (actually slightly reduced). And of course, flying cars are still the stuff of science fiction. Self-driving cars may be just around the corner, but haven’t arrived yet.

Medicine: Cancer and heart disease are still the top causes of death. Solving even one of these, the way we have mostly solved infectious disease and vitamin deficiencies, would have counted as a major breakthrough. Genetic engineering, again, has shown a few excellent early results, but hasn’t yet transformed medicine.

Manufacturing: In materials, carbon nanotubes and other nanomaterials are still mostly a research project, and we still have no material to build a space elevator or a space pier. As for processes, atomically precise manufacturing is even more science-fiction than flying cars.

If we had gotten even a few of the above, the last 50 years would seem a lot less stagnant.

One to zero

This year, the computer turns 75 years old, and the microprocessor turns 50. Digital technology is due to level off in its maturity phase.

So what comes next? The only thing worse than going from five simultaneous technological revolutions to one, is going from one to zero.

I am hopeful for genetic engineering. The ability to fully understand and control biology obviously has enormous potential. With it, we could cure disease, extend human lifespan, and augment strength and intelligence. We’ve made a good start with recombinant DNA technology, which gave us synthetic biologics such as insulin, and CRISPR is a major advance on top of that. The rapid success of two different mRNA-based covid vaccines is also a breakthrough, and a sign that a real genetic revolution might be just around the corner.

But genetic engineering is also subject to many of the forces of stagnation: research funding via a centralized bureaucracy, a hyper-cautious regulatory environment, and a cultural perception of something scary and dangerous. So it is not guaranteed to arrive. Without the right support and protection, we might be looking back on biotech from the year 2070 the way we look back on nuclear energy now, wondering why we never got a genetic cure for cancer and why life expectancy has plateaued.

Aiming higher

None of this is to downplay the importance or impact of any specific innovation, nor to discourage any inventor, present or future. Quite the opposite! It is to encourage us to set our sights still higher.

Now that I understand what was possible around the turn of the last century, I can’t settle for anything less. We need breadth in progress, as well as depth. We need revolutions on all fronts at once: not only biotech but manufacturing, energy and transportation as well. We need progress in bits, atoms, cells, and joules.

[Viliam]

Edit to shorten (more focus on arguments, less rhetorics), and include the top comment by jbash as a response /​ second part. The topic is important, but the article seems to have a bottom line already written.

[jbash]

Frankly, I think the biggest cause of the “stagnation” you’re seeing is unwillingness to burn resources as the world population climbs toward 8 billion. We could build the 1970s idea of a flying car right now; it just wouldn’t be permitted to fly because (a) it would (noisily) waste so much fuel and (b) it turns out that most people really aren’t up to being pilots, especially if you have as many little aircraft flying around from random point A to random point B as you have cars. A lot of those old SciFi ideas simply weren’t practical to begin with.

… and I think that the other cause is that of course it’s easier to pick low hanging fruit.

It may not be possible to build a space elevator with any material, ever, period, especially if it has to actually stay in service under real environmental conditions. You’re not seeing radically new engine types because it’s very likely that we’ve already explored all the broad types of engines that are physically possible. The laws of physics aren’t under any obligation to support infinite expansion, or to let anybody realize every pipe dream.

In fact, the trick of getting rapid improvement seems to be finding a new direction in which to expand, so that you can start at the bottom of the logistic curve. You got recent improvement in electronics and computing because microelectronics were a new direction. You didn’t get more improvement in engines because they were an old direction.

Your six categories are now all old directions (except maybe manufacturing, because that can mean anything at all). In 1970, you might not have included “information”… because wasn’t so prominent in people’s minds until a bunch of new stuff showed up to give it salience.

At the turn of the last century, you had much more of a “green field” in the all of the areas you list. You’re going to have to settle for less in those areas.

And there’s no guarantee that there are any truly new directions left to go in, either. Eventually you reach the omega point.

That said, I think you’re underestimating the progress in some of those areas.

Manufacturing

The real cost of basically everything is way down from 1970. Any given thing is made with less raw material, less energy, and less environmental impact.

I build stuff for fun, and the parts and materials available to me are very, very noticeably better than what I could have gotten in the 1970s.

Materials are much more specialized and they are universally better. Plastic in 1970 was pretty much synonymous with “cheap crap that falls apart easily”. In 2021, plastics are often better than any other material you can find. 2021 permanent magnets are in 1970s science fiction territory (and more useful than flying cars). Lubricants and sealants are vastly better. There’s a much wider variety of better controlled, more consistent metal alloys in far wider use, and they are conditioned to perform better using a much wider variety of heat treatments, mechanical processing, surface treatments, etc. Things that would have been “advanced aerospace materials” in 1970 are commonplace in 2021.

Mechanisms in general are much more reliable and durable, and require much less maintenance and adjustment.

I don’t believe 1970 had significant deployment of laser cutting, waterjet cutting, EDM, or probably a bunch of other process I’m forgetting about. They existed, but there were rare then, and they are everywhere now. 1970 had no additive manufacturing unless you count pottery.

It’s true that there’s no real change in how major bulk inputs are handled… because that stuff is really old (and was really old before 1970). There’s not much dramatic improvement still available, and not even that many “tweaks”.

Yeah, you don’t have MNT. Although there’s a lot of “invisible” improvement in the understanding of chemistry and the ability to manipulate things at small scales… and MNT was always supposed to be something that would suddenly pop up when those things got good enough. It might qualify as a “new direction”, but there are no guarantees about exactly when such a direction will open up.

Construction

Construction has always been conservative and has never moved fast. Given a comparable budget, 1970 construction wasn’t all that different than 1870 construction, the big exceptions being framed structure instead of post-and-beam and prefab gypsum board instead of in-place plastering.

As for 1970 to 2021, in 1970 you would have used much more wood to frame a house. Nobody used roof or floor trusses in residential construction. There was also a lot more lead and asbestos floating around… and they needed lead and asbestos, because without them their paints and insulation would not have remotely approached 2021 performance. For the most part they weren’t as good even with them. There’s also much wider deployment of plastic in construction (because plastic doesn’t suck any more). Fasteners are better, too, or at least it’s better understood which fastener to apply when and where.

I can tell at a glance that I’m not in a 1970 living room because the plugs are grounded. Also, unless it’s a rich person’s living room, the furniture is prefab flat pack particle board with veneer finishes instead of stick-built wood.

Agriculture

When I was a kid in the 1970s, the fresh food available in your average supermarket was dramatically less varied than it is now, and at the same time dramatically less palatable. Even the preserved food was more degraded. We actually ate canned vegetables at a significant rate.

If you didn’t live through maybe the 1980s to the 1990s or early 2000s, you can’t really have an idea how much better the food available to the average urban consumer has gotten.

A big part of that was better crop varieties, and I think another very big part was better management and logistics.

Energy

Energy is doing quite well, thanks, with several major, qualitative changes.

We have working renewables. Solar cells in 1970 were just plain unusable for any real purpose. Wind was a pain in the ass because of the mechanical unreliability of the generators (and was less efficient because of significantly worse turbine geometries). We’re also better at not wasting so much energy.

Batteries in 1970 were absolute garbage in terms of capacity, energy density, energy per unit weight, cycle count, you name it. Primary cells were horrible, and rechargeables were worse. You simply did not use a ton of little battery-powered gadgets of any kind. That’s partly because all the electrical devices we have now are much less power hungry, but it’s also because batteries have actually started not to suck. People in 1970 would have looked at you like you were crazy if you suggested a cordless drill, and that has nothing to do with the efficiency of the motor. By the way, that progress in batteries is based on a crapton of major materials science advances.

Yeah, nuclear didn’t happen, but that was for political reasons. One notable political issue was that fission plants are easy to use to make material for nuclear bombs. Nobody quite caught on to the whole CO2 issue until it was too late. And “nuclear homes, cars and batteries” were never a very practical idea, so it’s not surprising that they haven’t happened. You don’t want every bozo handling fissionables… and controlled fusion for power is probably impossible at a small scale, even assuming it’s possible at a large scale.

Transportation

The limits on transportation technology are energy and the Pauli exclusion principal. These are not things that you can easily change. You can’t expect new transport modes because the physical environment doesn’t change. You can’t expect a bunch of new engine types because there are a limited number of physically possible engine types.

For actual deployed infrastructure, you have to add political limits (which are probably the main reasons you don’t have much more efficiency by now)… and limits on what people want.

Doing a lot of space flight is a massive energy sink, and there is no urgent reason to waste that energy at the moment. Yes, I have heard the X-risk arguments, and no, they do not move me at all. Neither does asteroid mining. And the manifest destiny space colonization stuff sure doesn’t. Maybe the people who want all that space flight are simply a minority?

Supersonic transport is also not worth it. Speeds have gone down because nobody wants to waste that much energy or deal with that much noise (or move the material to dig huge systems of evacuated tunnels).

Medicine

Yeah, it’s a hard problem, see, because you have to hack on this really badly engineered system, which you’re not allowed to shut down or modify.

That said, cancer isn’t a single disease, and “the cure for cancer” was never going to be a thing. I think that actual medical people understood that even in 1970. There’ve actually been very significant advances against specific kinds of cancer. There are also improvements in prevention; screenings, HPV vaccine, whatever.

“Heart disease” isn’t really a single disease, either. But there’s a lot less of it around, with less impact, and not just because people stopped smoking. Even if you eat all the time and never exercise (which we’re worse about than in 1970), ya got yer statins, yer much better blood pressure meds, yer thrombolytics, yer better surgery, yer better implantable devices...

Oh, and they turned around a vaccine against a relatively novel pandemic virus in under a year. They identified that virus, sequenced its genome, and did a ton of other characterization on its structure and action, in time that would definitely have sounded like science fiction in 1970. They actually know a lot about how it works… detailed chemical explanations for stuff that would, in 1970, have been handwaved at a level just about one step above vitalism.

[Sanghyeon Seo]

In 1970, you might not have included “information”… because wasn’t so prominent in people’s minds until a bunch of new stuff showed up to give it salience.

I disagree. Any self-respecting history of technology includes invention of writing and printing press. Those two are like among the most important technology ever invented. You’d have “information” category just to include writing and printing press.

[jasoncrawford]

Also the phonograph, telegraph, telephone, radio, and television! If “information” wasn’t a category before the late 1800s, it was by then.

[lsusr]

Manufacturing

• Production per hour labor has increased in the USA and elsewhere.

• I have 3D printer in my living room.

Transportation

Home delivery is way cheaper than it used to be. Mail order has been a thing for a long time, but good mail-order products are new. It’s not impractical to buy everything you need without going to a store.

Manufacturing + Transportation

When you combine the developments of manufacturing plus transportation you get brand new phenomena, like being able to start a consumer hardware startup out of your living room. This doesn’t affect everyone’s living room, but my living room is full of shipping supplies and my basement has a small factory in it.

War

The B-2 stealth bomber reached initial operational capability in 1997. GPS and satellite mapping are ubiquitous. Guided missiles, real-time satellite surveillance and Predator drones are more capable today than in 1980s science fiction. Laser defenses, Gauss cannons and exoskeletons are on the way.

We also do nuclear tests with computer simulations instead of physical atom bombs. This particular example illustrates how physical technologies can turn into digital technologies. Physical technologies turning into digital technologies can create the illusion that physical technology is stagnating.

[jaspax]

The fact that some advances exist is entirely consistent with the thesis that the overall rate of advance has slowed down, as the original article points out.

[Aaro Salosensaari]

Home delivery is way cheaper than it used to be.

I am going to push back a little on this one, and ask for context and numbers?

As some of my older relatives commented when Wolt became popular here, before people started going to supermarkets, it was common for shops to have a delivery /​ errand boy (this would have been 1950s, and more prevalent before the WW2). It is one thing that strikes out reading biographies; teenage Harpo Marx dropped out from school and did odd jobs as an errand boy; they are ubiquitous part of the background in Anne Frank’s diaries; and so on.

Maybe it was proportionally more expensive (relative to cost of purchase), but on the other hand, from the descriptions it looks like the deliveries were done by teenage/​young men who were paid peanuts.

[lsusr]

When I think about home delivery, my reference point is the dao xiao mian 刀削面 knife I bought in 2020 from AliExpress for $3.57 including shipping and delivery to my door. In the 1990s, the simplest way to get an exotic product like that was to fly to China.

I’m not just thinking about the ease of sending something from one house to another within my city. I’m thinking about the ease of sending something from an arbitrary residence on Earth to an arbitrary residence on Earth.

[CronoDAS]

Shipping small packages from China to the US via the USPS has been subsidized—the price of mailing a Beanie Baby from Beijing to New York has been lower than the price charged to mail that same Beanie Baby from Los Angeles to New York. One of the few things the Trump administration did right was renegotiate the international postal system treaties so that China doesn’t get “developing nation” subsidies any more.

https://​​reason.com/​​2019/​​11/​​11/​​american-taxpayers-are-subsidizing-ultra-cheap-shipping-from-china/​​

[shminux]

While I don’t agree with everything, this is a top-quality comment deserving to be its own post, consider posting it as top-level?

[Viliam]

Even better (but with more work), an SSC-style adversarial collaboration.

[FireStormOOO]

Seconded; this should really be a reply post and is a good rebuttal. Much (though far from all) of the original argument is down to not really appreciating how much sci-fi tech we do have since the 70′s

[greylag]

(Epistemic status: lame pun)

I don’t believe 1970 had significant deployment of … EDM, or probably a bunch of other process I’m forgetting about

It was called “disco” in the 70s

[habryka]

I disagree with a bunch of this, but still upvoted for being a great comment with a lot of great data. Thanks for writing it!

[MalcolmOcean]

I don’t have the detailed knowledge needed to flesh this out, but it occurred to me that there might be a structure of an argument someone could make that would be shaped something like “we got a lot of meaningful changes in the last 70 years, but they didn’t create as many nonlinear tipping points as in the previous industrial revolutions.”

Fwiw, flying cars probably wouldn’t hit any such tipping point, though self-driving cars probably would.

Widespread nuclear energy might’ve meant little concern about global warming at this point, but solar & wind have been trucking along slowly enough that there’s tons of concern.

I think the internet is doing something important for the possibility of running your own 1-2 person business, which is a meaningful tipping point. There are various other tipping points happening as a result of computers and the internet, which is why I think it stands out as @jasoncrawford’s only named revolutionary technologies.

Anyway, hoping someone can steelman this for me, considering the nonlinear cascades in each era & from each technology, and seeing whether there’s indeed something different about pre-1970 and after. I’m not confident there is, to be clear, but I have some intuition that says this might be part of what people are seeing.

[ChristianKl]

There are also improvements in prevention; screenings, HPV vaccine, whatever.

Given that we have relatively constant cancer death rates it’s unclear whether the changes in cancer screenings are improvements. They might very well about taking organs from people who would otherwise live healthy lives if they wouldn’t be screened.

Oh, and they turned around a vaccine against a relatively novel pandemic virus in under a year.

You could do that with inactivated viruses the way Sinopharm and Sinovac do, also in 1970. The last year saw a lot of burocracy that made vaccine production harder and there’s a good chance that people in 1970 would have been better at producing and administering vaccine’s then we are today.

They identified that virus, sequenced its genome, and did a ton of other characterization on its structure and action, in time that would definitely have sounded like science fiction in 1970.

We have a relatively poor understanding of action. We don’t have crucial information about the long-term effects of getting infected. Our instiutions took an embarrasingly long time to recognize that masks are a good idea.

[dr_s]

Some of those issues though are political, not technological. Albeit it can certainly be argued that one of the causes of stagnation is that it’s the political and social institutions that have become inadequate at incentivising true innovation.

[Gerald Monroe]

If you examine your first 5, limited AI agents, similar to the kind demonstrated for autonomous cars, is capable of lifting the limits.

Manufacturing—self replicating robotics would drive prices through the floor

MNT—build tool designing AI agents to crack this problem, once you make the equipment for working at this scale cheap by producing it autonomously. Tool designing agents seem to be feasible per some of Open AI’s recent results.

Construction—same robots can build the buildings at hyperspeed, pre-fabrication with human workers is already vastly faster

Agriculture—falls with the same robotics case

Energy—presently it’s governed by the need for mass solar/​battery production

Transportation—building a new type of car/​engine/​overhead transit pods is a case of the manufacturing/​robotics problem. Since it’s too expensive right now to try anything but what we already have.

Medicine is a harder problem, I have a vague idea of using very advanced robotics and AI agents to build a “bottom up” understanding of biology so that it is possible to make new interventions in living humans and know they are going to work beforehand.

Alas there are government/​institution throttling issue with some of these advances. For construction, corrupt local jurisdictions can block construction of modular buildings, forcing expensive custom designs. For energy, solar/​battery systems need government support for there to be demand management/​grid backfeeding/​permits. For transportation, even if a new modality can be found (overhead maglev tracks, underground tunnels), a government has to permit the installation.

And of course medicine is the big one. We can posit an AI agent that could design a custom edit to a single patient’s genome. Or even invent a new treatment in realtime, during the period a single individual is in the process of dying. The old model of “RCT on enough people for statistical significance and pay 1 billion dollars in fees and salaries” does not allow for such rapid iteration to be possible.

[ChristianKl]

For construction, corrupt local jurisdictions can block construction of modular buildings, forcing expensive custom designs.

It’s the opposite. Local jurisdictions aren’t corrupt anymore to allow a billionaire to just pay bribes to get to build the modular buildings he wants to build.

We can posit an AI agent that could design a custom edit to a single patient’s genome. Or even invent a new treatment in realtime, during the period a single individual is in the process of dying.

While an AI theoretically could do that, we have a burocracy that outlaws such progress. As a result it’s now a lot more expensive to develop new treatments then back in 1970.

[Gerald Monroe]

Can you clarify the second point? The first point is—and ‘corrupt’ is a relative term. But for the overall society, inexpensive and large scale indoor space allows for progress. It makes a city more productive, a country more efficient, it makes the overall pace of technological development slightly faster. San Francisco blocking construction when they are arguably America’s most productive city as it is is likely harming the city, the city residents, the state they are in, the country they are in, and to a small extent, the world.

However the benefits of blocking construction do accrue to present landowners in expensive cities who get more certain ROIs on their investments and get to maintain their views. And in the way cities are allowed to block new construction in the US (versus it being handled at a higher level of government), the only votes come from current residents, many of whom are landowners and thus invested in the current system...

Are you referring to Broad Sustainable Building? https://​​en.wikipedia.org/​​wiki/​​Broad_Sustainable_Building

[ChristianKl]

It’s not possible for the people with money to pay bribes to get the policy outcomes they want. You have a middle class coalition who’s voting for politicians who block construction and the politicians fulfill their mandate.

the only votes come from current residents, many of whom are landowners and thus invested in the current system...

Politicians doing what their voters want is the opposite of them being corrupt.

[Gerald Monroe]

This is arguable. Certain you can winnow to arbitrarily small government districts and see how the interests of a tiny area can run contrary to the interests of everyone else.

For example, a small town sheriff refusing to arrest members of the town’s most powerful extended family—who in small enough town could control the electorate—is clear and simple corruption.

It’s because I see the job of a sheriff is to enforce the laws fairly and equally, not to favor tiny subgroups of people—even if that is the ‘will of the voters’.

Similarly, I see the office of building permits to promptly and efficiently issue permits, or denials, with clearcut and deterministic outcomes. In San Francisco, instead a project can be delayed 10+ years while it gets put through endless ‘reviews’.

In my specific case, I’m one of the many (millions?) of mobile tech workers. I move to wherever the next gig is. Whether that is Bay Area or San Diego or Atlanta, etc. It’s a national market for jobs with a national pool of workers. Should the rules for construction of housing be decided by the entrenched interests of a tiny area, or at the Federal or State levels? Or should I be able to obtain housing, and should my employer be able to obtain additional office space, in a free market at efficient prices.

Arguably, as most jobs today involve interstate commerce, and workers able to move in is interstate commerce, national building codes and a “shall issue” permit system (where a jurisdiction must issue a permit, by a deadline, if the project plans meet the code and all fees are paid) would make the United States more efficient. This is how it works in Japan, where Tokyo is an example of a place that doesn’t have a housing crisis.

[ChristianKl]

It’s because I see the job of a sheriff is to enforce the laws fairly and equally, not to favor tiny subgroups of people—even if that is the ‘will of the voters’.

It depends on your political system what the job of the sheriff is supposed to be. I don’t think that a sheriff should represent the interests of voters and don’t believe that it makes sense to have sheriff be an elected position.

In my specific case, I’m one of the many (millions?) of mobile tech workers.

A majority of people in California don’t think that the interests of mobile tech workers are more important then the interests of other people. In a democracy that means rich tech workers often not getting what they want.

Should the rules for construction of housing be decided by the entrenched interests of a tiny area, or at the Federal or State levels?

In a democracy that question is answered by the law. In California there don’t seem to be majorities for having those rules be made on the State level. Decisions being made at the level where the voting population wants them to be made instead of being made at the level you think best isn’t corruption. It’s just democracy.

[Gerald Monroe]

The decisions can be made at the state level just the laws have not yet passed.

https://​​www.enr.com/​​articles/​​48600-housing-density-effort-fails-in-california

[ChristianKl]

Yes, laws could be passed but at the moment the democratic majorities are not there for passing them on the state level.

[Gerald Monroe]

The linked article says they are but the California government has been lobbied to not care about the will of the voters. Almost like it’s corrupt per my original point upstream.

[ChristianKl]

You always have lobbying on an issue like that, and lobbying itself isn’t corruption.

California happens to have a referendum process and enough rich techis that want to build more housing to run such a referendum if it would pool well enough to have a good chance of getting passed.

[ChristianKl]

Should the rules for construction of housing be decided by the entrenched interests of a tiny area, or at the Federal or State levels?

Being able to distinguish people adopting a policy that you don’t like and have reason to believe to be bad from them being corrupt is very valuable.

Should the rules for construction of housing be decided by the entrenched interests of a tiny area, or at the Federal or State levels?

It’s easily possible to make a referendum for moving it onto the state level in California if the population wanted that.

[kyleschutter]

How does Pauli Exclusion principle limit transportation? Is this a teleportation joke? Thanks

[jbash]

It was mostly a joke and I don’t think it’s technically true. The point was that objects can’t pass through one another, which means that there are a bunch of annoying constraints on the paths you can move things along.

[johnswentworth]

Gordon, however, points out that GDP has never captured consumer surplus, and there has been plenty of surplus in the past. So if you want to argue that unmeasured surplus is the cause of an apparent (but not a real) decline in growth rates, then you have to argue that GDP has been systematically increasingly mismeasured over time.

So far, I’ve only heard one only argument that even hints in this direction: the shift from manufacturing to services. If services are more mismeasured than manufactured products, then in logic at least this could account for an illusory slowdown. But I’ve never seen this argument fully developed.

The argument which jumps to my mind is that value-of-bits is more systematically mismeasured than value-of-atoms. Information is generally cheap to copy. Marginal cost of copying the information is close to zero, therefore its price will be close to zero. That is exactly the condition under which GDP would be an especially poor measure, so we’d expect GDP to be increasingly poor as a larger percentage of growth shifts into information goods.

I still think the overall stagnation case you put forward is strong, though.

[jp]

Your point seems strong, but I’d go further. If I observe measured productivity go down, coinciding with a digital revolution, why should I update that true progress is slower rather than that digital goods are mis-measured?

[MalcolmOcean]

Something missing from the top-level post: why stagnation.

I’ll just put out that one of the tiny things that most gave me a sense of “fuck” in relation to stagnation was reading an essay written in 1972 that was lamenting the “publish or perish” phenomenon. I had previously assumed that that term was way more recent, and that people were trying to fix it but it would just take a few years. To realize it was 50 years old was kinda crushing honestly.

Here’s google ngrams showing how common the phrase “publish or perish” was in books through the last 200 years. It was coined in the 30s and took off in the 60s, peaking in 1968. Interesting & relevant timing!

[dr_s]

I think there’s a case for two different sources: one external, the simple lack of any further low hanging fruit to exploit, and one internal, which is exactly this—the increasing inadequacy of our institutions to create the conditions for innovation, often caused paradoxically by the excessive focus on promoting innovation. “If scientists gave us all this cool stuff by working on their own, imagine how good they will be if we hire a lot more of them and pit them into a competition for funding with each other!” was a catastrophically stupid idea, assuming humans can be made to produce highly creative work on command and like clockwork. It led to short-termism, creating a humongous confusing, amorphous mass of tiny innovations, many of which non replicable or straight up bogus, and efficiently killing off any incentive to actually work on long term, solid, sweeping discoveries.

[jasoncrawford]

That’s a great way of putting it

[jasoncrawford]

Some hypotheses for “why stagnation” in my review of Where Is My Flying Car?

[Raemon]

While reviewing two other progress studies posts for the LW 2019 Review (Concrete and Artificial Fertilizer) I found myself thinking “these seem neat, but I don’t feel like I can directly apply the knowledge here very much. What I really wish Jason Crawford would write (someday) are posts that take a birds eye view of progress, taking advantage of the gestalt of all the different object level “how did we invent X?” research and then deriving insights about the bigger picture.

So, I am quite delighted by this post, which I feel like does exactly that.

[jasoncrawford]

Thanks! I had to start from the ground level to build up to this perspective…

[Davidmanheim]

I think there is a partial explanation that comes from Goodhart’s law and human desire for superstimulation, and the ease with which that type of superstimulation is now available.

Capitalism is effective at fulfilling human desires, which are measured on the level of personal tradeoffs, taken on a very short time scale. The measure is not about human progress write large, but until people’s basic needs are satisfied, the two are very well correlated; enough food, labor saving devices, etc. If we are trying to advance “human progress” than at a certain point, the individual-level, short term motive no longer suffices.

If you want longevity, you need people to be willing to pay for it now—but they aren’t. If you want space travel, you need the beneficiaries to exist now, not in a few generations. But this got worse, because we’ve found a much faster feedback loop for desires than physical progress—screens. They are dopamine devices, and they absorb almost unlimited capacity for desire. That means we’re wireheading, and any longer term goals will be ignored.

I’d be really happy to hear ideas for how this can be routed around, as short of a global pandemic to convince people to actually care about some specific technological advance like a vaccine, I don’t know what motivates society enough to overcome this.

[maximkazhenkov]

I agree with your arguments but disagree with your value judgment—why shouldn’t digital entertainment be considered progress? What’s the point of “physical progress” once people’s basic needs are satisfied? If humanity ever becomes a Kardashev III civilization, what would we do with all that matter and energy besides creating digital Disney parks for septillions of immortal souls? What’s your vision for humanity’s future in the best case?

[Davidmanheim]

I think you’re asserting something I didn’t say. I said that, first, there are at least some types of digital “progress” that are akin to wireheading, and second, that this was likely a reason that progress in other domains had slowed. I did imply that if we value becoming a Kardashev III civilization, as you suggest, we can’t spend all our lives in digital space—someone needs to build the rockets, do the engineering, and colonize the galaxy. And to the extent that wireheading reduces how much we do those things, it could be a net negative.

[Raemon]

Curated.

As noted in a previous comment: I appreciate this post because I think it’s fairly hard to make sense of all the stuff going on in the technological-innovation-space, if you’re just looking at individual “trees” rather than the collective “forest.” I know Jason Crawford has been thinking a lot about this space over a long time, getting to look at a lot of individual trees enough to take stock of how the collective forest is doing.

So, getting to see how he changed his mind about things, and how that exploration fed into updates on the stagnation hypothesis, was useful to me.

[Dustin]

I can’t point to the episode(s) or post(s), but I believe both on his blog and on his podcast Conversations with Tyler, Tyler has expressed the idea that we may be currently coming out of the stagnation of stuff in the Real World driven by stuff like SpaceX, CRISPR, mRNA, etc.

[jasoncrawford]

Yup, here is a roundup of recent “is stagnation over?” commentary: https://​​noahpinion.substack.com/​​p/​​techno-optimism-roundup

I would like to address /​ reply to this soon.

[theme_arrow]

I like Noah Smith and enjoyed that post, but I’m not sure I agree with the conclusions. I think I agree more with the critique from Applied Divinity Studies than I do with Noah. There are definitely areas for optimism, but I haven’t seen anything that looks like we’re actually getting increases in productivity growth in the US or similar countries. Moreover, I have seen no indications that cost disease has slowed (see healthcare costs for example, other than the decrease in care due to COVID lockdowns, there’s no real evidence of a slowdown). And beyond all that, the ineffective response to COVID in the US has shown just how badly our state capacity for effectively confronting problems has deteriorated.

I would really, really like stagnation to be over (and I do my best to support to efforts to make that happen!), but I don’t yet feel very hopeful.

[jasoncrawford]

I’m not convinced by the optimists, either, and ADS made some good points. This post was laying the foundation for my response. With this framework I think you can analyze things in at least a slightly more rigorous way.

[DirectedEvolution]

Since then, the “third Industrial Revolution”, starting in the mid-1900s, has mostly seen fundamental advances in a single area: electronic computing and communications. If you date it from 1970, there has really been nothing comparable in manufacturing, agriculture, energy, transportation, or medicine—again, not that these areas have seen zero progress, simply that they’ve seen less-than-revolutionary progress. Computers have completely transformed all of information processing and communications, while there have been no new types of materials, vehicles, fuels, engines, etc.

One of the challenges here is defining what counts as a “new” category of innovation as opposed to an incremental improvement on an “old” category.

I expect many people would count the invention of Bakelite as a “new”material. Bakelite was the first fully synthetic plastic, invented in 1909. But all plastics are polymers, and Bakelite mainly just replaced these natural polymers with a cheaper, better alternative. It is chemically distinct from the natural polymers it replaced, and its synthetic nature was truly revolutionary.

But that’s just the problem. What makes a material “revolutionary” depends less on how physically distinct it is from other machines and materials, and more on what economic possibilities it unlocks. The compound bow, invented in 1966, might have been “revolutionary” if it had been invented in 10,000 BC. Because it was invented centuries into the gunpowder era, it was not.

So “stagnation” really ought to mean that we are creating economic demand at a lower rate than in the past. The airplane created economic demand for flight, and improvements to the airplane in terms of safety, speed, cost, and destinations increased that demand. Now, the airline industry is a poster child for stagnation. It’s not that people don’t want to fly. It’s just that there’s no economically tractable way to make them want to fly more.

A structural feature of medicine that promotes “stagnancy” in the sense of economic demand is the fact that people take most medicines only because they get a disease. No matter how good a statin is, patients won’t take it unless they’re at risk for something like a heart attack or stroke, and they usually will take it if they are at risk. Only medications for entirely new diseases can create demand where there was none before, and these medicines are mainly (though not exclusively) in the category of rare diseases, which limits the economic impact that any one of them can have. So “stagnancy” in the field of medicine merely means “we have at least one go-to treatment for almost every common illness.”

Note that this isn’t 100% true. In HIV, we have PrEP, which is a new category of medication that people use to prevent acquiring HIV, and thus attacks HIV from a whole new angle, creating a whole new type of demand. COVID vaccines and medications also have this attribute. We had nothing, and now we have something.

This structural feature of “stagnation” is almost, but not quite, an inevitability. Some technologies are “generative,” in the sense that they make further new products possible that would have been meaningless before. iPhones are “generative” because they are not only in demand, but they create demand for many other products and services to improve and exploit them: waterproof and decorated cases, magnetic stands for use in cars, bluetooth headphones, etc. Air travel is “generative” because it creates demand for things like tourist resorts, travel guides, gift shops, neck pillows, and wheeled luggage.

While “stagnation” ought to set in within any given product category for the purely structural reason that we naturally wind up creating at least one product to satisfy each form of demand, I see no reason why we would a priori expect to see a decline in these “generative” technologies that create knock-on forms of demand. But I do accept that a decline in GDP is empirical evidence for such a decline.

There are certain categories of progress that I think we neglect in these discussions.

• I’d tentatively count Ukraine’s so far successful repulsion of Russian invasion, using innovative arms and intelligence supplied by other countries, as an example of progress. As is the ability to punish Russian aggression with international sanctions. In the past, my guess is that the difference in manpower would have counted for a lot more, and it would have been much more difficult to weaponize the global economy against Russia. Who knows what the knock-on effects will be if Ukraine is able to win the war?

• Note that many of these definitions of “progress” and “stagnation,” including the ones I’m offering here, are specifically about progress for the global rich. The global poor have seen massive progress since the 1970s, and I don’t think anybody seriously denies that.

• Another example of “progress” in atoms is the rise of polyamory, gay marriage, greater permissiveness around divorce and contraception, and other forms of social and sexual advancement. These forms of progress aren’t machine-based, but they do seem heavily facilitated by the internet, which makes it much easier for minority interests to connect, form community, and share information. Cultural progress of this kind must somehow be relevant as a form of economic progress, but I have no idea how it would be counted in traditional measures of GDP. It seems to me that this might ultimately prove to be economically generative over time as these communities have a chance to mature. After overcoming a material bottleneck, people turn their attention to social bottlenecks. Maybe eventually they turn their attention back to material bottlenecks, or to something else.

All that said, there’s still something compelling to the question “where’s my X?”

If I could list some things that I really want, and for which it seems no technology exists, I might wish for:

• A better way to collate various DNA precipitation protocols, ensure they work with the materials and equipment in my lab, and train me to execute the protocol successfully.

• Much better couple’s therapy than presently exists. Being able to have a therapist experience in which both members of the couple feel the therapist has the capacity to simultaneously be “exclusively” their individual therapist while also being able to act fully as a therapist for the couple.

• Effective, non-addictive, and legal nootropics.

• VR-controlled robots with fine manual dexterity.

• Cheap miniature heat-and-humidity-resistant stick-on microscope cameras that could be accessed over the internet.

• An affordable laptop with 3-4x the screen space.

• A much easier way to try on clothes and other products.

• Technology that could automatically and reliably find the highest-quality local services that fit your price range and schedule.

But mostly, what I want are things that already exist, but cost more money than I’m willing to pay. I want my own house, access to better lab equipment, more money to buy better-quality food, a Maine Coon cat, a really nice bike, etc. etc. etc. I have a lot of room for material progress in my own life without requiring any new technologies in order to get it. Over time, I fully anticipate that I will realize these forms of material progress.

Unfortunately, material progress will come with declining health and fewer years of life remaining. Eventually, longevity will become my bottleneck to progress, and this is true for every human.

Based on all this, I think that a progress agenda that focuses on increasing the supply and quality of goods and services that already exist, while working toward longevity technology, would address the two areas that I think are most relevant for making material improvements in my own life.

Longevity among the global rich currently appears to be slowing down, but I would not be surprised if that trend changes in my lifetime, and I think it’s mostly a consequence of the time it takes for the science and engineering community to master the suite of new technologies we’ve developed over the last 20 years.

I’m not sure if sheer material abundance is decelerating. Global meat production was accelerating from 1980-2018, but suddenly flatlined in that year. World car production was pretty linear from 2001-2016 and has fallen off a cliff since then. World PC shipments have been fairly steady since 2006. World energy consumption flatlined in 2019, but was visibly decelerating after about 1980. These make me weakly update to “yes,” material progress is decelerating.

As Cowen and Southwood point out, these decelerations are happening in a context of massively increased scientific research. This suggests that scientific innovation is not the driver of economic production that it once was. If I had to guess, I’d blame a combination of coordination problems, regulatory slowdowns, philosophical differences, and underinvestment in public infrastructure. We should have had full nuclear power a long time ago. The way we regulate medicine is onerous. Insane laws like the Foreign Dredge Act accumulate obstacles and linger far too long because it’s so hard in a democratic context to coordinate to repeal them. How is it that fission scientists have been so underfunded for so long? We have the miracle of CRISPR-Cas9, and every article on the topic prefaces it by stating that it’s wildly controversial.

From a bird’s eye view, humanity rarely knows what it wants most, feels embarrassed about those desires it has identified, and even when it’s not too confused or timid to pursue its objects, often has no clear idea of how to cooperate in order to get it. Humanity needs therapy.

From that perspective, a “philosophy of progress” seems like it’s pointing in the right direction. “Have you considered that what you might want is material abundance, longer lives, a community you enjoy, and the technology that allows you to get these things without constantly producing catastrophic side effects?”

[sophia_xu]

effective debate notes: I’ve read main points of every first-level comment in this thread and the author’s clarification.

epistemic status: This argument is mostly about values. I hope we can all agree with the facts I mentioned here, but can also consider this alternative framework which I believe is a “better map of reality”.

I disagree with your conclusion because I disagree with your model of classification of tech frontiers. In the body of this article and most comments, people seem to agree with your division of technology into 6 parts. Here’s why I think this model might be not very capable: I believe it assumes every kind of innovation is “fundamentally” or “canonically” same.

1. Specifically, it ignores different area’s different “transferrability” or “interconnectivity” with other fields. For example, innovations in manufacturing/​agriculture/​energy/​transportation/​medicine generally cannot be transferred to one another directly; while innovations in information can be transferred to other fields easily. Google scholar “machine learning” + any of the five fields and we should be able to find plenty of literature reviews on them.

2. It doesn’t care about how important different fields matter to us on a “meta” level. One defining characteristic for humans is the use of complex language and theory of mind. None of the other 5 field in the original framework contribute directly to languages and the use of languages; while information technology by definition includes the enhanced efficiency and accessibility of a wide variety of discussions and knowledge-sharing. I see that you might agree with this point here by your discussions in the first half of the article. However, the impact of this aspect of information technology can be much bigger than other fields in the original classification, including:

   1. Simply more potential for progress: if you ask scholars 50 years ago in their very specifically subdivided fields, that they would have so many pre-prints to read even if they don’t sleep or eat, they’ll probably ask “what’s a pre-print?” before even saying “no way!”. Internet-based tools are constantly increasing accessibility and thus quantity of research, while simultaneously increasing the speed of research through streamlining the research process from knowledge acquisition to reproducibility to peer review. Results may take a bit to propagate from the science world to tech world, but this level of meta-scientific discovery is really unprecedented since the invention of the printing press (and I know you’re not talking about the scientific frontier here, so this is just an example rather than a point).

   2. From nation states to human civilization: sure, technologies like social media divide us and that’s a huge problem we should think about hard, but instant communication across the globe and very good machine translation services have already transforming a large part of the human population into a group of shared values and fundamentally agreeable ways of thinking (im kinda talking about the type of thinking by the time of enlightenment without coining a specific name for it, since people outside of the western tradition doesn’t call it our way; doesn’t mean they don’t think like this), and it should be trivial that this would bring technological progress.

3. Economically, bits > atoms. The previous sentence is used metaphorically, and what I’m trying to convey is that bits are ideas. Ideas can be copied from person to person, or from machine to machine, at a speed many times as fast as copying physical objects, and at a price of zero (mostly). This makes even a tiny innovation in IT matter the same as huge innovations in traditional industries: you can have an entire field dealing with more efficient ways to predict protein structures while spending countless hours, but a machine learning model (alphafold 2) can match lab performance at much less cost and much better scalability (I know way less about biology than I know about machine learning, so please correct me if I’m wrong!)

   1. Cost and scalability are the central point being made here—one of the most important innovations in my book is the public cloud industry led by the Amazon Web Service. Hiring more research assistants have a diseconomies of scale: coordination, management, communication (in a PHYSICAL environment! be sure not have any OSHA violations—and lawyer up! - and make sure to secure your lab equipment—maybe add a guard - …etc) all become harder the more people you hire. However, if you’re just adding another 100 GPUs to your infrastructure, you wouldn’t need “middle management” for them—maybe their programming counterpart, but they’re much cheaper, and you can get many of them open-source.

So my main point is that, the information revolution should really be a printing-press-level innovation, and comparing it to electricity or steam engine really missed a lot of important fundamental differences of IT, and these unique characteristics are already manifesting themselves everywhere. So here’s my alternative framework for the original categories (roughly):

• All important technological innovation categories (impact* from least to most)

  • Helping us enhance reality

    • Manufacturing & construction

      • concrete, civil engineering, skyscrapers

    • Energy

      • non-renewable energy, fission, renewable energy, fusion

    • Transportation

      • highways, containerization, international shipping, self-driving cars

  • Helping us enhance ourselves (but physical)

    • Agriculture (not as important technology-wise since we can already meet all the needs we have; just a matter of redistribution)

      • genetically-modified corps, genetically-engineered corps

    • Transportation (partly)

      • subways, intercontinental flights, self-driving cars

    • Medicine/​Bio-*

      • polio vaccines, CRISPR, COVID-19 vaccines, immortality

  • Helping us enhance ourselves (but conceptual)

    • Information

      • alphabets, printing press, internet

Prescriptively (more of ‘predicting the future’), my belief is that although in the past years our focus shifted from more heavy-industrial innovations to more “meta” and indirect ones (also including non-technical ones like communication theory), the latter has more potential than the former from the points above.

*: Since no innovation come alone, our value functions for importance and impact of an innovation should not only include immediate impacts, but also potential ones that might take longer to fully manifest but we can already see coming.

Edit 1: add epistemic status

[jasoncrawford]

Interesting, but I think you’re underestimating the impact of other general-purpose technologies, such as in energy or manufacturing. New energy sources can be applied broadly across many areas, for instance.

[ChristianKl]

When it comes to medicine I think the key thing we notice stagnation is in lifespans. Focusing on individual diseases is misleading. Curing cancer is commonly seen as being worth three years in life-span. Health in Hong-Kong is two times curing cancer better then health in the US if you use lifespan as your measuring stick.

Looking at the history it seems interesting that the change in progress in lifespan was more in the early 1960s then in the 1970s. The Kefauver Harris Amendment in 1962 might be one of the key drivers.

In the last decades we had conceptual innovations in medicine like Evidence-Based medicine and meta-studies. We can finally say that Chiropratics do roughly as much for back pain as our other medical interventions for back pain because we can use metastudies which means we have an alternative to rejecting it because we don’t like the theory.

[Prescod]

I think that the reason we don’t use average lifespan as our measuring stick is because people are far more bothered by the variability in that metric than its mean. If 20 year olds were offered the choice of “do you want the average life expectancy or do you want a rock-solid guarantee that you will achieve average minus 5”, I suspect that most people choose the guarantee. Not just because of hyperbolic discounting, but also because the average 73 year old has accomplished what they plan to accomplish in life. Whereas people are very afraid to die while they have small children, before they’ve proven themselves in their careers and before they’ve experienced retirement.

[Liron]

“Go into a room and subtract off all of the screens. How do you know you’re not in 1973, but for issues of design?”

At least if you’re in an average grocery store, you can tell it’s the 2000s from the greatly improved food selection

[Gordon Seidoh Worley]

I speculate that there may be anthropic reasons for some of the stagnation.

In particular, I suspect the slowdown in growth of energy production may be what we would expect to see because in worlds where energy production grew faster we would have also had more abundant and more powerful weapons that would have been more likely to kill more humans, thus making it less likely to find oneself living in a world where such growth didn’t decrease.

[Martin Sustrik]

As for transportation, I would say the average time to get to a place have dropped considerably in past 50 years, not because of any specific invention, but because airplanes has become less toys for the rich and more of buses with wings available to everyone. Similarly, densification of the motorway system made it faster to go places by car.

It’s not clear, of course, whether that kind of thing counts as technological progress. But if not so, what kind of progress is it?

[jasoncrawford]

There was far more progress in aviation from 1920–1970 than from 1970–2020. In 1920, planes were still mostly made of wood and fabric. By 1970 most planes had jet engines and flew at ~600mph. Today planes actually fly a bit slower than they did in 1970. Yes, there has been progress in safety and cost, but it doesn’t compare to the previous 50-year period.

Similar pattern for automobiles and even highways.

[Martin Sustrik]

I was speaking from personal experience.

In 1980′s it took 6 hrs to get to my grandmothers place. Today it is more like 3 hrs. All that not because of better cars but because there’s a highway covering most of the distance.

In 1980 people rarely traveled by plane. A holiday by seaside meant a 12 hour ride by car to Yugoslavia. Today, everyone’s flying to Turkey and Canary Islands.

[jasoncrawford]

Ah, you are from Eastern Europe? To clarify, the stagnation hypothesis is about the frontier of technological development in the wealthiest countries. I don’t think there has necessarily been stagnation in global development.

[Martin Sustrik]

Yes, I am from Eastern Europe. That made me wonder whether the densification of the road system has slowed down in the west.

Here are statistics for the US:

In short, there’s a slowdown, but it starts in ’90.

Source

Air miles per capita seem to tell a different story though:

Source

[ryan_b]

Out of curiosity, where do logistics fit into the categorization you use? I ask because we seem to be measuring only by top-line numbers like mph, but mph was never the point in the first place—now almost everyone in the country can get almost anything made anywhere in the country dropped on their front porch in ~2 days. In real terms this is a radical increase in speed for transport of goods.

My guess is something like the overlap of transport and information; logistics is information applied to transport. It doesn’t change the thesis really, I just notice that a lot of things we now rely on that they didn’t have during the periods of high growth is the notion of on net, which we now apply very widely but in the 19th-20th centuries prior to computers was represented chiefly by manufacturing and notions like vertical integration.

[jasoncrawford]

Yes, I put logistics under transportation. Transportation of cargo has always been more important than passenger travel.

Containerization was huge, but it mostly happened 50+ years ago.

[Timothy Johnson]

I mostly agree with your thesis, but I noticed that you didn’t mention agriculture in the last section, so I looked up some numbers.

The easiest stat I can find to track long-term is the hours of labor required to produce 100 bushels of wheat [1].

1830: 250 − 300 hours

1890: 40 − 50 hours

1930: 15 − 20 hours

1955: 6 − 12 hours

1965: 5 hours

1975: 3.75 hours

1987: 3 hours

That source stops in the 1980s, but I found another source that says the equivalent number today is 2 hours [2]. That roughly matches the more recent data on total agricultural productivity from the USDA, which shows continued improvement, but not on the scale of the mid-1800′s [3].

On the other hand, if the Wall Street Journal is right, being a farmer sounds a lot less strenuous today: https://​​www.wsj.com/​​articles/​​farmers-plow-through-movies-while-plowing-fields-11557508393 (paywalled). Instead of manual labor, farmers can sit in an air-conditioned tractor cab and watch Netflix! That’s progress of a different sort, I suppose...

I also wondered about the impact of GMO food. That sounds possibly revolutionary to me, so why does its impact not show up in the numbers?

The sources I found suggest that recent GMO advances can improve yield by 10% in corn [4] or 20 − 50% across a range of other crops [5]. That’s great, but not as dramatic as I thought it could be.

[1] Farm Machinery and Technology Changes from 1776-1990 (thoughtco.com)

[2] Wheat Trivia—Food Facts & Trivia: Wheat (foodreference.com)

[3] USDA ERS—Agricultural Productivity Growth in the United States: 1948-2015

[4] New genetically modified corn produces up to 10% more than similar types | Science | AAAS (sciencemag.org

[5] GMO crops have been increasing yield for 20 years, with more progress ahead—Alliance for Science (cornell.edu)

[Gerald Monroe]

So for the 2 hours, realize that the reason it’s not zero is there is a ‘residual’ human labor input where currently shipping control systems are not robust enough to replace the human. To summarize the problem (it’s the same problem repeated everywhere): there is a near infinite number of rare ‘edge cases’ that a tractor can experience. Current computer software is not feasible to engineer for all the edge cases, so the tractor has an autopilot that handles the 90-99% or so ‘main happy case’ of driving the tractor, and the person onboard watching netflix has to be ready to take over when it hits an edge.

This is pretty much the same problem repeated for packing boxes at Amazon and all the rest. Too many varied items on the shelves. (the ‘picking’ problem). Or for manufacturing the goods that are being shipped—robots can make the injection molded main pieces, and be hand set up for commonly made goods, but there are all these little ‘edge’ cases where a factory worker has to do some of the steps by hand, making the human labor input more than zero.

[hamnox]

More dependencies tho

[ryan_b]

I wonder about the question of catch-up growth.

There’s a point raised in Is Scientific Progress Slowing Down by Tyler Cowen and Ben Southwood about problems relating to using TFP as an estimate of the value of progress, which is that it is hard to categorize ideas which are adopted late. Duplicating one of the quotes in the comments there:

To consider a simple example, imagine that an American company is inefficient, and then a management consultant comes along and teaches that company better personnel management practices, thereby boosting productivity. Does this count as an improvement in TFP or not? Or is it simply an increase in labor supply, namely that of the consultant? On one hand, some hitherto-neglected idea is introduced into the production process. That might militate in favor of counting it as TFP. On the other hand, the introduced idea is not a new one, and arguably the business firm in question is simply engaged in “catch up” economic growth, relative to more technologically sophisticated firms.

Generalizing a bit, this seems like it could introduce counting problems everywhere. For example, after WWII half of Eurasia had to be rebuilt; does this count as catch-up growth? What about the industrialization of other countries, whether via deliberate economic development strategies or outsourcing—these factories weren’t built with state-of-the-art techniques, but rather because the same old technology was cheaper to build and operate there.

How closely does this global boom in catch-up growth coincide with our 1970s stagnation timeline, and can a part of the explanation be something as simple as capital being redirected from technical progress to expansion of the current technical base?

[jasoncrawford]

This analysis, and the stagnation debate in general, is really about the technological frontier. Global development overall has not necessarily been stagnating—India and China have seen huge growth in the last 50 years.

[Gerald Monroe]

Note that this can be modeled pretty simply.

Evolution is slow and for the sake of argument, assume human intelligence has not changed over the time period you have mapped. If humans haven’t gotten any smarter, and the next incremental step in the technology areas you have mentioned requires increasingly sophisticated solutions and thought, then you would expect progress to slow. Fairly obvious case of diminishing returns.

Second, computers do act to augment human intelligence, and more educated humans alive increase the available brainpower, but this too contrasts with probably exponentially increasing difficulty in certain fields.

For example, in the field I presently work, I see huge teams of people and very sophisticated equipment used to support further improvements in microprocessors. I think in the past the teams needed to be much smaller, and the equipment was simpler, and the gains were easier.

The fields you mention there are specific bottlenecks we can discuss in detail but I feel that would make this post too long. But the TLDR : nuclear turns out to have long tail risks that current human run organizations can’t economically efficiently manage, transportation is gated by energy and human reflexes, medicine is limited by a number of inefficiencies, and manufacturing has seen enormous improvements, just not in the way you think. The rise of China has made manufactured goods of varied quality levels far more abundant than in the past, and has made them accessible to billions more people. Your “typical home, subtract the screens” model is implicitly assuming a nice home in Los Angeles in 1955 or so. But over in China most people did not yet have running water and had minimal access to electricity.

The singularity model is simply, we know that each of the fields mentioned, we can create toy examples of “minimum limit case” where we don’t know the limits of physics, but we do know what we could achieve if we had limitless intelligence to engineer to the limits rapidly.

Below I have listed what are the ‘minimum limit cases’. We don’t know how far the fields can be pushed past the point I mentioned but everything I mention is pretty conclusively supported by evidence as being feasible.

Note that everything listed, including medicine advances, require supporting advances in AI to make them feasible. We don’t have that yet. We have a lot of toy models made at a small scale but not the integrated systems just yet. By AI I mean limited function agents able to choose good (but not perfect) control outputs for a robotic system, not self amplifying superintelligences.

Medicine—we know that aging is governed by processes that accumulate negative changes over time. Stopping/​reversing these changes is possible if we knew exactly which genes to rewrite. At the limit case, medicine would be able to produce artificial replacements for any organ, stabilize any dying patient with high speed AI/​robotics to identify the correct intervention and apply it, and turn off the root cause of most diseases.

Transportation—a packet switched network of vacuum trains/​point to point flying cars/​overhead transit pods/​autonomous cars

Energy—solar panels over all buildings and deserts, batteries at every electrical panel, a network of demand regulation

Agriculture—robotic farms in a sealed pod

Construction—robotics that assemble a building from pre-fabricated building blocks that come on a truck

Manufacturing—robotics that are sophisticated enough to manufacture themselves and to self-clear nearly all faults

[kyleschutter]

Another argument in favor of the slow down is that you don’t hear people arguing that innovation is speeding up. It’s either slowing down or the same as before.

[Stephen Richards]

I’m not sure you are talking about technology advancement but rather the gifts of different hydrocarbon ages. Your IR1 to me is better described as everything we could do with coal. IR2 is everything we could do with oil. IR3 seems different to me, if humanity invented true AI -will we be literally gods?

Also there is the issue of order of magnitude improvements only being noticeable when they pass the human scale and seeming irrelevant either side. Commuting to work in a flying car might be fun, but flying my desk home at the speed of light is far faster and vastly more practical, not to mention shrinking my work desk to a fraction of an SSD.

Finally it seems like a get of my lawn/​kids these days post. Romans had chariots and roads and indoor heating, cars and interstates are impressive in scale and speed, but so is flying my office desk.

What will be the next big driver of technology innovation? (Especially as 8-10 billion people may have to go through more expensive hydrocarbons over the next 50 years slowing down the technological feedstock) My guess is abundant green electrical power, and slightly improved batteries combined with ai and crop improvements to feed us, these will be held back by resources per person declining (oil per capita peaked in the 70s I think, good wood is expensive), and climate change issues.

[MichaelBowlby]

Great article, small (but maybe significant) nitpick. I don’t think there’s great evidence that innovation-by-bureaucracy is bad, and I actually think it’s pretty good. The sheer quantity of innovation produced by the two world wars and space program is spectacular, as is the record of DARPA post WW2. Even the Soviet Union could do could innovation by bureaucracy. At the company level, it’s true that lots of big innovations come from small new companies, but at the same time Bell Labs, Intel, Sony were monsters. I actually think that the retreat of government is a big part of problem.

[jasoncrawford]

Large orgs can get things done, that’s true.

In this post I specifically referred to the centralization and bureaucratization of research funding, e.g., the consolidation of research funding under the NIH/​NSF.

[Jakobovski]

Maybe it’s wishful thinking but I think the stagnation is temporary. I think there is reason to believe that progress can come in waves. Scientific advancements of the early 1900s led to much of the technological advancements of pre and post war era. So to, I think it reasonable that computation advancements of the last 30 or so years will lead to a number of interesting advancements:

I think it is reasonable to assume that many of the items from the list below will happen in the next 10 to 20 years.

• Self driving cars

• 2.5x batteries

  • electric cars

  • electric airplanes

  • independence from fossil fuels

• Genetic engineering revolution (crispr, mRNA vaccines etc..)

• Large advances in AI possibly AGI.

• Fusion ?

If/​when we hit AGI there will be a huge wave of progress.

[jasoncrawford]

Maybe, see Dustin’s comment above for related discussion

[Thomas]

I don’t see how “2.5 batteries” could mean “independence from fossil fuels”.

[CronoDAS]

It’s the grid-level storage problem. Solar panels don’t produce much energy at night, wind doesn’t blow all the time, etc., so you need to produce extra energy and store it for when production is less than demand.

[Olamide Olanrewaju]

Maybe it’s because we’re now using the fiat system. The fiat system was introduced in 1971, and everything has been going downhill from there. It could be correlation or causation. But it’s definitely something. This makes it more vivid:

https://​​wtfhappenedin1971.com/​​

[dr_s]

And what’s the matter with bits, anyway? Are they less important than atoms?

Arguably, yes, because they are less fundamental. A revolution in our understanding of the fundamental laws of physics begets more secondary revolutions down the line; quantum mechanics alone gave us lasers, nuclear power and those very bits—to name but a few. So from a revolution in our understanding of the world comes the promise of more revolutions once that reaches the application stage. Understanding quantum gravity might lead to warp drives. But no matter how great our ability to manipulate bits, the best they generally can do (except for the possibility of AGI, I guess) is help us squeeze more efficiency out of what we have already. We feel the energy problem especially keenly right now, of course, and far from helping all that much, computers only eat up ever more energy, sometimes in rather pointless ways. A clean, cheap source of energy right now would be worth far more than all the social media in the world.

[yds1987]

Seems to me that solving the problem of the safe mass-production, storage and transport of gaseous and liquid hydrogen would be the game-changer, for terrestrial needs as well as space travel. It can be gotten from water or even air. If the danger of handling could be reduced, it would surpass petrol as the primary fuel and then off we go. Solves the carbon problem and the energy resource problem.

Then if someone figured out how to better harness the sun’s energy… Of the energy that reaches earth, what do we capture and put to use now, like .001%? Or hydroelectricity from the tides? Lot more reliable than winds. All the current problems/​roadblocks seems to be energy-related.

[SpectrumDT]

There is one aspect which you almost completely ignore in your post and which I believe is vital: WHY is stagnation bad? WHY is progress good? You make it sound as though you just want fancy gadgets for the sake of having fancy gadgets.

You have a “quantitative” section, but what exactly are you trying to quantify? Why is economic growth good?

You should spend more time pondering: What (quantifiable) factors are valuable in and of themselves? And have these factors improved or stagnated?

For example, I believe there is less famine and fewer people living in extreme poverty worldwide. Lifespans may have stagnated in the developed world, but how about in the developing world? If poor people are less poor and live longer, that is extremely important progress. They need the progress way more than we do.

Conversely, climate change and other looming environmental disasters may make poverty and suffering skyrocket again and make the world a much worse place, even if our gadgets grow ever fancier.

[jasoncrawford]

One clarification: This is about the technological frontier, not about global development. See my followup post.