# Average Is Over  ## Metadata - Author: [[Tyler Cowen]] - Full Title: Average Is Over - Category: #books ## Highlights - These trends stem from some fairly basic and hard-to-reverse forces: the increasing productivity of intelligent machines, economic globalization, and the split of modern economies into both very stagnant sectors and some very dynamic sectors. ([Location 71](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=71)) - Tags: [[blue]] - This imbalance in technological growth will have some surprising implications. For instance, workers more and more will come to be classified into two categories. The key questions will be: Are you good at working with intelligent machines or not? Are your skills a complement to the skills of the computer, or is the computer doing better without you? Worst of all, are you competing against the computer? Are computers helping people in China and India compete against you? If you and your skills are a complement to the computer, your wage and labor market prospects are likely to be cheery. If your skills do not complement the computer, you may want to address that mismatch. Ever more people are starting to fall on one side of the divide or the other. That’s why average is over. ([Location 76](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=76)) - Tags: [[blue]] - What is happening is an increase in the ability of machines to substitute for intelligent human labor, whether we wish to call those machines “AI,” “software,” “smart phones,” “superior hardware and storage,” “better integrated systems,” or any combination of the above. This is the wave that will lift you or that will dump you. ([Location 91](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=91)) - Tags: [[blue]] - Writers and teachers need to consider what aspects of their work are better done by intelligent-machine analysis and look closely at the irreplaceable value they do provide. ([Location 134](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=134)) - Tags: [[blue]] - let’s ask a few questions. First, in which major areas do we see ongoing technological advances exceeding expectations from just a few years ago? Second, in which areas do we see a lot of new and promising technological works in progress? Third, in which areas can we expect the general forces propelling innovation (say globalization or Moore’s law that processing power for computers will continue rising at rapid rates) to remain powerful? Finally, can we see evidence that these areas are already influencing economic statistics measuring our nation’s well-being? I’ll get into more detail on all of these questions, but for now the point is that the areas of the economy identified in the answers to these questions all overlap on one technology: mechanized intelligence. And its effect on economic statistics is trending up. ([Location 148](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=148)) - Tags: [[blue]] - Technological progress slows down when there are too many people who have the right to say no, but software in general gets around a lot of the traditional veto points. The key work is done in the individual mind and with relatively small teams and in computer code, and it’s hard to hold back the innovations by law or custom. ([Location 243](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=243)) - Tags: [[blue]] - let’s say that machine intelligence helps us make a lot more things more cheaply, as indeed it is doing. Where will most of the benefits go? In accord with economic reasoning, they will go to that which is scarce. In today’s global economy here is what is scarce: 1. Quality land and natural resources 2. Intellectual property, or good ideas about what should be produced 3. Quality labor with unique skills Here is what is not scarce these days: 1. Unskilled labor, as more countries join the global economy 2. Money in the bank or held in government securities, which you can think of as simple capital, not attached to any special ownership rights (we know there is a lot of it because it has been earning zero or negative real rates of return) We see high returns to resource owners, such as the new-resource millionaires and billionaires of Brazil, Russia, Canada, and Australia, and similarly huge revenues for intellectual property giants such as Apple and other innovators in that sector. The current array of scarce and plentiful resources now means high wages or capital gains to talented and inventive workers, and pretty low returns on ordinary labor and ordinary savings. ([Location 248](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=248)) - Tags: [[blue]] - That means humans with strong math and analytic skills, humans who are comfortable working with computers because they understand their operation, and humans who intuitively grasp how computers can be used for marketing and for other non-techie tasks. It’s not just about programming skills; it is also often about developing the hardware connected with software, understanding what kind of internet ads connect with their human viewers, or understanding what shape and color makes an iPhone attractive in a given market. Computer nerds will indeed do well, but not everyone will have to become a computer nerd. ([Location 270](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=270)) - Tags: [[blue]] - The ability to mix technical knowledge with solving real-world problems is the key, not sheer number-crunching or programming for its own sake. Number-crunching skills will be turned over to the machines sooner or later. ([Location 278](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=278)) - Tags: [[blue]] - Despite all the talk about STEM fields, I see marketing as the seminal sector for our future economy. A salesperson can use knowledge of computers or engineering to sell a complex technical product to a technically sophisticated user, and in fact such knowledge might these days be required to sell effectively. That’s based in some STEM skills, but it’s also marketing. ([Location 280](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=280)) - Tags: [[blue]] - Note: If the barrier to create goes down significantly (due to LLMs coding), then the value of the non-coding startup tasks (strategy, marketing, product taste) goes up substantially - It might appear that a masseuse is not much affected by computers, at least provided you are skeptical about these robots that now offer massages. Nonetheless, masseuses increasingly market themselves on Google and the internet. These masseuses fit the basic model that favors people who can blend computer expertise with an understanding of how to communicate with other people. Again, it is about blending the cognitive strengths of humans and computers. ([Location 283](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=283)) - Tags: [[blue]] - The growing importance of marketing integrates two seemingly unrelated features of the modern world: income inequality and increasing pressures on our attention. The more that earnings rise at the upper end of the distribution, the more competition there will be for the attention of the high earners and thus the greater the importance of marketing. ([Location 296](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=296)) - Tags: [[blue]] - But don’t just focus on those computers; it’s also about management. The CEOs and higher-level managers are paid handsomely to assemble and direct the individuals who work every day with mechanized intelligent analysis. If you have an unusual ability to spot, recruit, and direct those who work well with computers, even if you don’t work well with computers yourself, the contemporary world will make you rich. If we look at the increase in the share of income going to the top tenth of a percent from 1979 to 2005, executives, managers, supervisors, and financial professionals captured 70 percent of those gains. ([Location 321](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=321)) - Tags: [[blue]] - To put this point in a broader business context, until another good manager is hired, there is no point in employing another assistant. It’s the manager who is the scarce input, and that is one way to think about why managerial salaries have been going up so much. Managers play a role of growing importance in coordinating complex, large-scale production processes. ([Location 358](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=358)) - Tags: [[blue]] - to be blunt—while I know I can’t prove this—I wonder how much of the middle class consists of people in government or protected service-sector jobs who don’t actually produce nearly as much as their pay. ([Location 482](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=482)) - Tags: [[blue]] - As a general rule, the age structure of achievement is being ratcheted upward due to specialization and the growth of knowledge. Mathematicians used to prove theorems at age twenty, but now it happens at age thirty because there is so much more to learn along the way. If you are a talented twenty-two-year-old, just out of Harvard, you probably cannot walk into a furniture factory and quickly design a better machine. Young people have made fundamental contributions in some of the internet and social networking sectors, precisely because of the immaturity of those sectors. Mark Zuckerberg needed a good grasp of Myspace, but he didn’t have to master decades of previous efforts on online social networks. He was close to starting from scratch. In those cases, young people tend to dominate the sector, but of course that won’t cover the furniture factory. ([Location 526](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=526)) - Tags: [[blue]] - The financial crash was a very bad one-time event that revealed, rather suddenly, this more fundamental long-term structural problem, namely that a lot of workers had been overemployed relative to their skills. ([Location 678](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=678)) - Tags: [[blue]] - The sad reality is that many of these workers you don’t want at all, even if the business plan involves additional labor. Some workers simply aren’t worth the trouble unless the demand for extra labor is truly pressing. I believe these “zero marginal product” workers account for a small but growing percentage of our workforce. At the very least they make it unlikely that we will return to 4 percent unemployment in the foreseeable future. ([Location 701](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=701)) - Tags: [[blue]] - Note: Unemployment HAS fallen far below 4% in the intervening 10 years. Is it just that many people dropped out of the labor force entirely? Or did Cowen misanalyze the economic drivers as well? - Overall, these job market trends are bringing higher pay for bosses, more focus on morale in the workplace, greater demands for conscientious and obedient workers, greater inequality at the top, big gains for the cognitive elite, a lot of freelancing in the services sector, and some tough scrambles for workers without a lot of skills. Those are essential characteristics of the coming American labor markets, the new world of work. ([Location 788](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=788)) - Tags: [[blue]] - Top American grandmaster Hikaru Nakamura was not a huge hit when he tried Freestyle chess, even though he was working with the programs. His problem? Not enough trust in the machines. He once boasted, “I use my brain, because it’s better than Rybka on six out of seven days of the week.” He was wrong. ([Location 945](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=945)) - Tags: [[blue]] - Some critics of computer chess think that computer reliance on the opening book is wrong or unfair. Indeed, most programs allow you to turn off the book, in which case the computer has to think through the opening moves on its own. Even the very best computers aren’t so good at doing that, because in the early stages of the game there are an especially large number of variations, and long-term strategy is usually paramount over tactics. For the first fifteen moves or so of most games, a reasonably strong master (even more, a top grandmaster) is a better player than a top-class program without its opening book. ([Location 988](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=988)) - Tags: [[blue]] - Note: With so many possibilities, it is impossible to search through them all. Human knowledge sets the bounds of the search space at the beginning (using openings which govern the first 30 or so moves), and this substantially constrains what the program has to look for. Transferring to LLMs, we can constrain the search space by knowing a lot about system design, management, unit testing, etc. If we are able to do this cognition before prompting the LLM, we can constrain the search space to a much smaller area (eg coding a single function whose name, parameters, and docstring are already provided) - in plenty of real-world situations the immediate command over factual or analytical material brings a big edge. Discussions in meetings, strategies and reactions during sales calls, lawyers arguing in front of a jury, and managers in volatile, voices-raised personnel situations all try to draw upon preprocessed information at a moment’s notice. In all those cases, it matters more and more what workers have learned from the computer, or not, and how well they remember computer-derived information and advice. ([Location 1010](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=1010)) - Tags: [[blue]] - We also can use the concept of man–machine collaboration to define the difference between a worthless or “zero marginal product” worker and a potentially valuable worker. The worthless worker is one whose cooperation with the machine makes the final outcome worse rather than better. A potentially valuable worker offers the promise of improving on the machine, taken alone. In the language of economics, we can say that the productive worker and the smart machine are, in today’s labor markets, stronger complements than before. ([Location 1101](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=1101)) - Tags: [[blue]] - When economists investigate human rationality, they are often too dependent on arbitrary stipulations about what is rational and what is not, expressed in the form of models. An economist might write down some mathematical axioms and then find that human behavior falls short of those axioms. But how convincing were those axioms in the first place for complex and multidimensional human problem solving? A lot of the research in this tradition isn’t convincing, no matter how brilliant the investigator. ([Location 1170](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=1170)) - Tags: [[blue]] - Most of all, Ken is impressed by the overall reliability of human decision making. Rybka is of course better at chess than humans, but both the good moves and the mistakes of the humans fit regular, understandable patterns. There is a sense of rationality and order to human error. These same human beings are making decisions about love, business, and which new car to buy, and there is something comforting about what shows up in these patterns. For instance, players are at their toughest when it matters most. A player is least likely to make a major error when the game is tight, and if anything, players do their absolute best when they are faced with a slight disadvantage in their position. When players are decisively up or down, they don’t seem to think or concentrate with the same facility. Again, this is a sign of human rationality, at least if there is some need for a conservation of effort. Before these investigations, Ken expected to find evidence for a Nassim Taleb “Black Swan” model of cognitive failure. That is, a lot of errors coming out of the blue. But in fact, radically surprising “Black Swan” errors don’t play much of a role in the final outcome. Most games are decided on the basis of the accumulation of advantages, and the level of error is fairly well predicted by the relative skills of the players. Ken finds these results all the way down to the level of a 1,600-rated player, which would be a middling club player in most cities (he has yet to look at the games of worse players). Players also have consistent styles. For instance, Vladimir Kramnik, world champion from 2000 to 2007, plays with an especially high level of accuracy relative to what the computer suggests. His expected error is extremely low—for a human being, that is. Yet Kramnik also does not pose his opponents many problems over the board. He does not cause an inordinate number of mistakes, and he does not crush his opponents’ egos or cause them to choke up and collapse in the face of mind-numbingly complex positions. It’s not that hard to play fairly well against Kramnik. He creates calm positions with a lot of good, long-term strategic moves. Under Ken’s measure of the most “nettlesome” player—the player whose play most pushes his opponents into an above-average error rate—prodigy Magnus Carlsen scores the highest. If you play against Carlsen, you’re much more likely to make major blunders, as the game more rapidly becomes a minefield of complications. ([Location 1204](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=1204)) - Tags: [[blue]] - One thing that Robin shows with his model is that you can have a long period where—for reasons of complementarity—machines boost wages, but eventually machines substitute for intelligent labor and wages can fall rapidly. ([Location 1607](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=1607)) - Tags: [[blue]] - The ancient arts of memory, in their most general form, are techniques to improve your mind. These arts were not just about memorization and many of their advocates drew an explicit distinction between the memory arts and memorization. The memory arts were about learning how to order ideas in new ways, and thus the memory arts were a path to composition and innovation and the generation of novelty. It was about taking older and simpler parts and from those parts making new things, be they hymns, poems, prayers, books, or a new appreciation of the wonders of God. The point was to make your ideas “searchable,” as a computer literate person would say. That is, you could start with a question or some simple starting points and be led, by the algorithmic nature of the memory arts, to more complex ideas and truths. When it comes to Google, the right magic keys can get you to many new places, and those keys consist in the relatively manageable art of knowing the right search words. Google is the successful embodiment, through technology, of the earlier dream of the memory palace. ([Location 1819](https://readwise.io/to_kindle?action=open&asin=B00C1N5WOI&location=1819)) - Tags: [[blue]]