The natural category grouping all these people together isn’t “high achiever” or “creative revolutionary,” but “legendary figure,” as you indicate in your last paragraph. Bill James was a key innovator in the science of baseball, and yet Babe Ruth is included in this riff and Bill James is not. Babe Ruth is central to the legend of baseball, and James is not.
To become a legendary figure, it’s important to have some sort of achievement in a way that serves the national interest or makes for colorful storytelling, and it doesn’t hurt to have a distinctive personality. Einstein’s innovations helped us fight wars, Darwin was an ocean explorer, Armstrong kicked Russia’s ass in the space race, Beethoven was a cranky deaf incel, Galileo was forced to recant by the church, Socrates was forced to drink hemlock.
The whole “low hanging fruit/tapped out fields” theory of scientific stagnation seems inadequately operationalized. It conjurs up the image of an apple tree, with fruit pickers standing on the ground reaching ever-higher for the remaining fruit.
An alternative image is that the scientific fruit-pickers are building a scaffold, so that more fruit is always within arm’s reach. Unless we’re arguing that the scaffold is nearing the top of the tree of knowledge, there’s always low-hanging fruit to be picked. Somebody will pick it, and whether or not they become a legendary figure depends on factors other than just how juicy their apple turned out to be. The reaching-for-fruit action is always equally effortful, but the act of scaffold-building gets more efficient every year. In literal terms, previous scientific discoveries and capital investments permit achievements that would have been inaccessible to earlier researchers, and we’re getting better at it all the time.
Improved infrastructure doesn’t always coincide with inaccessibility, either: consider the trend in the accessibility of supercomputer or or DNA sequencing over time. Think about how incredibly hard it was to learn mathematics when mathematicians were in the habit of jealously guarding their procedures for solving cubic equations, or when your best resource for learning algebra was Cardano’s Ars Magna, of which Jan Gullberg as written “No single publication has promoted interest in algebra like Cardano’s Ars magna, which, however, provides very boring reading to a present-day peruser by consistently devoting pages of verbose rhetoric to a solution… With the untiring industry of an organ grinder, Cardano monotonously reiterates the same solution for a dozen or more near-identical problems where just one would do.”
Consider also that either this heuristic was either false at one point (i.e. it used to be the right decision to go into science to achieve “greatness,” but isn’t anymore), or else the heuristic is itself wrong (because so many obvious candidates for “great innovator” were all in academic science and working in established fields with fairly clear career tracks). If it used to be true that going into academic science was the right move to achieve scientific greatness, but isn’t anymore, then when and why did that stop being true? How do we know?
A potential objection is to argue in favor of the existence of scientific stagnation. Isn’t there empirical evidence that the apples are, indeed, getting harder to pluck?
Perhaps there is. Let’s accept that premise. The “low-hanging fruit” perspective is that the right way to cope is by finding your own tree, one that nobody else has picked. The “build a scaffold” perspective suggests that all the fruit, all the time, is a priori equally hard to reach, because you’re always reaching up from a scaffold somewhere. In fact, if any trees do exist without a scaffold, there’s a risk that either nobody built a scaffold there because it’s an unproductive tree, or that the lowest-hanging fruit there is still so high up that you’ll have to do all the work of building a scaffold there before any harvest can begin. No guarantee that all that investment will pay off with a rich harvest of apples, either.
I’m really skeptical of the stagnation argument as well. Take machine learning. This a critically important technology, one with sweeping implications in every field (not just in science), and that may be the foundation of GAI—certainly one of, if not the most important technology that has ever, and perhaps will ever, be invented. Complaining that “it’s not clear they’re more significant breakthroughs than the reordering of reality uncovered in the 1920s,” as Collison and Nielsen do, seems both subjective and small-minded. I’d say the same thing about the computing revolution in general.
Can anybody point to one individual who deserves the bulk of the credit for ML, or the home computing revolution? It’s a combination of the development of the underlying mathematical theory, its instantiation in code, the development of computing infrastructure, algorithmic refinements, and applications to particular problems. The impact is massive, and the credit is diffuse. Probably Benjamin Black and Chris Pinkham, original visionaries of Amazon Web Services, get some of the credit for developing core infrastructure on which the world relies for developing and deploying ML. But nobody knows who they are, and they certainly won’t be getting a Nobel.
Tracking individual citations, reputation, fame, and other measures of credit as measures of counterfactual marginal impact (CMI), seems to me like a misbegotten enterprise, or at the very most the crudest of possible measures. It’s clear to me that CMI is a meaningful concept, and I understand that we now have some theory to explore counterfactuals with sound statistical methods. But let’s just say that if this heuristic is correct, then “measuring scientific progress” is a tree with some very low hanging fruit, and you should definitely consider making a career in it.
> An alternative image is that the scientific fruit-pickers are building a scaffold, so that more fruit is always within arm’s reach. Unless we’re arguing that the scaffold is nearing the top of the tree of knowledge, there’s always low-hanging fruit to be picked. Somebody will pick it, and whether or not they become a legendary figure depends on factors other than just how juicy their apple turned out to be. The reaching-for-fruit action is always equally effortful, but the act of scaffold-building gets more efficient every year. In literal terms, previous scientific discoveries and capital investments permit achievements that would have been inaccessible to earlier researchers, and we’re getting better at it all the time.
I don’t agree with this, for reasons discussed here. I think that empirically, it seems to get harder over time (at least per capita) to produce acclaimed works. I agree that there are other factors in who ends up “legendary,” but I think that’s one of them.
> Consider also that either this heuristic was either false at one point (i.e. it used to be the right decision to go into science to achieve “greatness,” but isn’t anymore), or else the heuristic is itself wrong (because so many obvious candidates for “great innovator” were all in academic science and working in established fields with fairly clear career tracks). If it used to be true that going into academic science was the right move to achieve scientific greatness, but isn’t anymore, then when and why did that stop being true? How do we know?
The heuristic is “to match the greats, don’t follow in their footsteps.” I think the most acclaimed scientists disproportionately followed this general heuristic—they disproportionately asked important/interesting questions that hadn’t gotten much attention, rather than working on the kinds of things that had well-established and -understood traditions and could easily impress their acquaintances. For much of the history of science, this was consistent with doing traditional academic science (which wasn’t yet particularly traditional); today, I think it is much less so.
As science progresses, it unlocks engineering possibilities combinatorially. This induces demand for and creates a supply of a healthier/more educated population with a larger number of researchers, many of whom are employed in building and managing our infrastructure rather than finding seminal ideas.
So what we’re witnessing isn’t scientific innovations becoming harder to find. Instead, scientific innovations produce so many engineering opportunities that they induce demand for an enormous number of engineers. The result is that an ever-decreasing fraction of the population is working in science, and science has to compete with an ever-larger and more lucrative engineering industry to attract the best and brightest.
In support of this, notice that the decline you find in scientific progress roughly begins in the early days of the industrial revolution (mid-late 1700s).
On top of that, I think we have to consider the tech-tree narrative of science history writing. Do you imagine that, even if you’d found that science today was becoming apparently more efficient relative to effective population, that scientific history writers would say “we’re going to have to cut out some of those Enlightenment figures to make room for all this amazing stuff going on in biotech!”?
It’s not that historians are biased in favor of the past, and give short shrift to the greatness of modern science and technology. It’s that their approach to historiography is one of tracing the development of ideas over time.
Page limits are page limits, and they’re not going to compress the past indefinitely in order to give adequate room for the present. They’re writing histories, after all. The end result, though, is that Murray’s sources will simply run out of room to cover modern science in the depth it deserves. This isn’t about “bad taste,” but about page limits and historiographical traditions.
These two forces explain both the numerator and the denominator in your charts of scientific efficiency.
This isn’t a story of bias, decline, or “peak science.” It’s a story of how ever-accelerating investment in engineering, along with histories written to educate about the history of ideas rather than to make a time-neutral quantification of eminence, combine to give a superficial impression of stagnation.
If anything, under this thesis, scientific progress and the academy is suffering precisely because of the perception you’re articulating here. The excitement and lucre associated with industry steers investment and talent away from investment in basic academic science. The faster this happens, the more stagnant the university looks as an ever-larger fraction of technical progress (including both engineering and science) happens outside the academy. And the world eats its seedcorn.
The natural category grouping all these people together isn’t “high achiever” or “creative revolutionary,” but “legendary figure,” as you indicate in your last paragraph. Bill James was a key innovator in the science of baseball, and yet Babe Ruth is included in this riff and Bill James is not. Babe Ruth is central to the legend of baseball, and James is not.
To become a legendary figure, it’s important to have some sort of achievement in a way that serves the national interest or makes for colorful storytelling, and it doesn’t hurt to have a distinctive personality. Einstein’s innovations helped us fight wars, Darwin was an ocean explorer, Armstrong kicked Russia’s ass in the space race, Beethoven was a cranky deaf incel, Galileo was forced to recant by the church, Socrates was forced to drink hemlock.
The whole “low hanging fruit/tapped out fields” theory of scientific stagnation seems inadequately operationalized. It conjurs up the image of an apple tree, with fruit pickers standing on the ground reaching ever-higher for the remaining fruit.
An alternative image is that the scientific fruit-pickers are building a scaffold, so that more fruit is always within arm’s reach. Unless we’re arguing that the scaffold is nearing the top of the tree of knowledge, there’s always low-hanging fruit to be picked. Somebody will pick it, and whether or not they become a legendary figure depends on factors other than just how juicy their apple turned out to be. The reaching-for-fruit action is always equally effortful, but the act of scaffold-building gets more efficient every year. In literal terms, previous scientific discoveries and capital investments permit achievements that would have been inaccessible to earlier researchers, and we’re getting better at it all the time.
Improved infrastructure doesn’t always coincide with inaccessibility, either: consider the trend in the accessibility of supercomputer or or DNA sequencing over time. Think about how incredibly hard it was to learn mathematics when mathematicians were in the habit of jealously guarding their procedures for solving cubic equations, or when your best resource for learning algebra was Cardano’s Ars Magna, of which Jan Gullberg as written “No single publication has promoted interest in algebra like Cardano’s Ars magna, which, however, provides very boring reading to a present-day peruser by consistently devoting pages of verbose rhetoric to a solution… With the untiring industry of an organ grinder, Cardano monotonously reiterates the same solution for a dozen or more near-identical problems where just one would do.”
Consider also that either this heuristic was either false at one point (i.e. it used to be the right decision to go into science to achieve “greatness,” but isn’t anymore), or else the heuristic is itself wrong (because so many obvious candidates for “great innovator” were all in academic science and working in established fields with fairly clear career tracks). If it used to be true that going into academic science was the right move to achieve scientific greatness, but isn’t anymore, then when and why did that stop being true? How do we know?
A potential objection is to argue in favor of the existence of scientific stagnation. Isn’t there empirical evidence that the apples are, indeed, getting harder to pluck?
Perhaps there is. Let’s accept that premise. The “low-hanging fruit” perspective is that the right way to cope is by finding your own tree, one that nobody else has picked. The “build a scaffold” perspective suggests that all the fruit, all the time, is a priori equally hard to reach, because you’re always reaching up from a scaffold somewhere. In fact, if any trees do exist without a scaffold, there’s a risk that either nobody built a scaffold there because it’s an unproductive tree, or that the lowest-hanging fruit there is still so high up that you’ll have to do all the work of building a scaffold there before any harvest can begin. No guarantee that all that investment will pay off with a rich harvest of apples, either.
I’m really skeptical of the stagnation argument as well. Take machine learning. This a critically important technology, one with sweeping implications in every field (not just in science), and that may be the foundation of GAI—certainly one of, if not the most important technology that has ever, and perhaps will ever, be invented. Complaining that “it’s not clear they’re more significant breakthroughs than the reordering of reality uncovered in the 1920s,” as Collison and Nielsen do, seems both subjective and small-minded. I’d say the same thing about the computing revolution in general.
Can anybody point to one individual who deserves the bulk of the credit for ML, or the home computing revolution? It’s a combination of the development of the underlying mathematical theory, its instantiation in code, the development of computing infrastructure, algorithmic refinements, and applications to particular problems. The impact is massive, and the credit is diffuse. Probably Benjamin Black and Chris Pinkham, original visionaries of Amazon Web Services, get some of the credit for developing core infrastructure on which the world relies for developing and deploying ML. But nobody knows who they are, and they certainly won’t be getting a Nobel.
Tracking individual citations, reputation, fame, and other measures of credit as measures of counterfactual marginal impact (CMI), seems to me like a misbegotten enterprise, or at the very most the crudest of possible measures. It’s clear to me that CMI is a meaningful concept, and I understand that we now have some theory to explore counterfactuals with sound statistical methods. But let’s just say that if this heuristic is correct, then “measuring scientific progress” is a tree with some very low hanging fruit, and you should definitely consider making a career in it.
> An alternative image is that the scientific fruit-pickers are building a scaffold, so that more fruit is always within arm’s reach. Unless we’re arguing that the scaffold is nearing the top of the tree of knowledge, there’s always low-hanging fruit to be picked. Somebody will pick it, and whether or not they become a legendary figure depends on factors other than just how juicy their apple turned out to be. The reaching-for-fruit action is always equally effortful, but the act of scaffold-building gets more efficient every year. In literal terms, previous scientific discoveries and capital investments permit achievements that would have been inaccessible to earlier researchers, and we’re getting better at it all the time.
I don’t agree with this, for reasons discussed here. I think that empirically, it seems to get harder over time (at least per capita) to produce acclaimed works. I agree that there are other factors in who ends up “legendary,” but I think that’s one of them.
> Consider also that either this heuristic was either false at one point (i.e. it used to be the right decision to go into science to achieve “greatness,” but isn’t anymore), or else the heuristic is itself wrong (because so many obvious candidates for “great innovator” were all in academic science and working in established fields with fairly clear career tracks). If it used to be true that going into academic science was the right move to achieve scientific greatness, but isn’t anymore, then when and why did that stop being true? How do we know?
The heuristic is “to match the greats, don’t follow in their footsteps.” I think the most acclaimed scientists disproportionately followed this general heuristic—they disproportionately asked important/interesting questions that hadn’t gotten much attention, rather than working on the kinds of things that had well-established and -understood traditions and could easily impress their acquaintances. For much of the history of science, this was consistent with doing traditional academic science (which wasn’t yet particularly traditional); today, I think it is much less so.
As science progresses, it unlocks engineering possibilities combinatorially. This induces demand for and creates a supply of a healthier/more educated population with a larger number of researchers, many of whom are employed in building and managing our infrastructure rather than finding seminal ideas.
So what we’re witnessing isn’t scientific innovations becoming harder to find. Instead, scientific innovations produce so many engineering opportunities that they induce demand for an enormous number of engineers. The result is that an ever-decreasing fraction of the population is working in science, and science has to compete with an ever-larger and more lucrative engineering industry to attract the best and brightest.
In support of this, notice that the decline you find in scientific progress roughly begins in the early days of the industrial revolution (mid-late 1700s).
On top of that, I think we have to consider the tech-tree narrative of science history writing. Do you imagine that, even if you’d found that science today was becoming apparently more efficient relative to effective population, that scientific history writers would say “we’re going to have to cut out some of those Enlightenment figures to make room for all this amazing stuff going on in biotech!”?
It’s not that historians are biased in favor of the past, and give short shrift to the greatness of modern science and technology. It’s that their approach to historiography is one of tracing the development of ideas over time.
Page limits are page limits, and they’re not going to compress the past indefinitely in order to give adequate room for the present. They’re writing histories, after all. The end result, though, is that Murray’s sources will simply run out of room to cover modern science in the depth it deserves. This isn’t about “bad taste,” but about page limits and historiographical traditions.
These two forces explain both the numerator and the denominator in your charts of scientific efficiency.
This isn’t a story of bias, decline, or “peak science.” It’s a story of how ever-accelerating investment in engineering, along with histories written to educate about the history of ideas rather than to make a time-neutral quantification of eminence, combine to give a superficial impression of stagnation.
If anything, under this thesis, scientific progress and the academy is suffering precisely because of the perception you’re articulating here. The excitement and lucre associated with industry steers investment and talent away from investment in basic academic science. The faster this happens, the more stagnant the university looks as an ever-larger fraction of technical progress (including both engineering and science) happens outside the academy. And the world eats its seedcorn.