Discover more from Noahpinion
Is it time for the Revenge of the Normies?
An optimistic take on technology and inequality.
On the app formerly known as Twitter, I’m known for occasionally going on rants about how it’s good to be normal and average and middle-class. To some degree this is because I believe that the only successful society is an egalitarian one where people don’t have to be exceptional in order to live good and comfortable and fulfilling lives. But some of it is also a reaction against the messages I was inundated with growing up. It seemed like every movie and book and TV show was telling me that nerds like me were special — that because we could do physics or program computers or even just play video games, we were destined to be exceptional. In the late 80s and 90s, it felt like we were on the cusp of a great shift, where the back-slapping jocks who had dominated American society in earlier times were on the verge of losing power and status to the bespectacled freaks and geeks. The Revenge of the Nerds was coming.
It wasn’t just fantasy, either. Over the next thirty years, the nerds really did win the economic competition. The U.S. shifted from manufacturing to knowledge industries like IT, finance, bio, and so on, effectively going from the world’s workshop to the world’s research park. This meant that simply being able to cut deals and manage large workforces were no longer the only important skills you needed to succeed at the highest levels of business. Bespectacled programmers and math nerds became our richest men. From the early 80s to the 2000s, the college earnings premium rose relentlessly, and a degree went from optional to almost mandatory for financial success.
The age of human capital was in full swing, and the general consensus was that “Average Is Over”. And with increased earnings came increased social status and personal confidence; by the time I moved out to San Francisco in 2016, tech people were clearly the masters of the Universe.
The widening gap in the performance of the nerds versus everyone else wasn’t the only cause of the rise in inequality in the U.S. — financialization, globalization, tax changes, the decline of unions, and other factors all probably played a role. But the increasing premium on human capital was impossible to ignore.
That trend lasted so long that most Americans can no longer remember anything else. We’ve become used to the idea that technology brings inequality, by delivering outsized benefits to the 20% of society who are smart and educated enough to take full advantage of it. It’s gotten to the point where we tacitly assume that this is just what technology does, period, so that when a new technology like generative AI comes along, people leap to predict that economic inequality will widen as a result of a new digital divide.
And it’s possible that will happen. I can’t rule it out. But I also have a more optimistic take here — I think it’s possible that the wave of new technologies now arriving in our economy will decrease much of the skills gap that opened up in the decades since 1980.
Generative AI is a power tool for the mind
First, let’s talk about generative AI, the technology that has everyone excited right now. A whole lot of research is being done on the productivity effects of generative AI tools, and they all seem to conclude the same thing: Generative AI gives a much bigger boost to low performers than to high performers.
For example, Brynjolfsson, Li, and Raymond (2023) gave AI tools to customer support staff, and here’s what they found:
AI assistance disproportionately increases the performance less skilled and less experienced workers across all productivity measures we consider…[W]e find that the AI tool helps newer agents move more quickly down the experience curve: treated agents with two months of tenure perform just as well as untreated agents with over six months of tenure. These results contrast, in spirit, with studies that find evidence of skill-biased technical change for earlier waves of computer technology…
[T]he productivity impact of AI assistance is most pronounced for workers in the lowest skill quintile…who see a 35 percent increase in resolutions per hour. In contrast, AI assistance does not lead to any productivity increase for the most skilled workers…[L]ess-skilled agents consistently see the largest gains across our other outcomes…These findings suggest that while lower-skill workers improve from having access to AI recommendations, they may distract the highest-skilled workers, who are already doing their jobs effectively[.]
And when Noy and Zhang (2023) gave generative AI tools to educated professionals to help them with writing tasks, they found the same pattern:
The control group [without ChatGPT assistance] exhibits persistent productivity inequality…In the treatment group [with ChatGPT assistance], initial inequalities are half-erased by the treatment…This reduction in inequality is driven by the fact that participants who scored lower on the first round benefit more from ChatGPT access[.]
And when Peng et al. (2023) studied the effects of GitHub Copilot on programmers’ output, they again found the same pattern:
The heterogeneous effects identified in this study warrant close attention. Our results suggest that less experienced programmers benefit more from Copilot. If this result persists in further studies, the productivity benefits for novice programmers and programmers of older age point to important possibilities for skill initiatives that support job transitions into software development.
And when Choi and Schwarcz (2023) used GPT-4 to help students on law exams, guess what they found:
GPT-4’s impact depended heavily on the student’s starting skill level; students at the bottom of the class saw huge performance gains with AI assistance, while students at the top of the class saw performance declines. This suggests that AI may have an equalizing effect on the legal profession, mitigating inequalities between elite and nonelite lawyers.
And when Doshi and Hauser (2023) used GPT-4 to help writers be more creative, they found that it was an equalizer:
Among the most inherently creative writers…there is little effect of having access to GenAI ideas on the creativity of their stories…In contrast, access to GenAI ideas substantially improves the creativity and select emotional characteristics of stories written by inherently less creative writers…In short, [using GPT-4] effectively equalizes the creativity scores across less and more creative writers.
To my knowledge, there’s no study so far showing that more talented people are able to use generative AI more effectively than less talented people. All of the evidence points to generative AI as an equalizer.
Intuitively, it’s not hard to think of why this might be the case. Whereas previous forms of information technology complemented human cognition, generative AI tends to substitute for human cognition. In order to program computers the traditional way, or even to apply lots of kinds of software, you had to have a mind that could think like a computer; this naturally privileged the kind of people who were able to make their minds do computation. (Even writing has some of this; a lot of the skill of writing is just painstaking mental attention to detail.)
But generative AI is specifically set up to interface with people who don’t think like machines. It does most of the machine stuff for them. Yes, there are some techniques like prompt engineering that may take some amount of difficult cognition, but this is a far cry from the difficult, often repetitive mental work that older IT technologies required.
In other words, it looks increasingly likely that traditional IT acted like a shovel — something that complemented people’s natural abilities — while generative AI acts more like a steam shovel. A steam shovel handles the muscle-power for you; GPT-4 handles the detail-oriented thinking for you. Technologies that substitute for natural ability tend to make natural ability less scarce, and therefore less valuable.
Of course this doesn’t mean generative AI will decrease inequality overall. The computation-intensive nature of these tools means that physical capital — access to large amounts of cheap GPUs or other key hardware — might make a comeback as a source of wealth. But by boosting the performance of the least skilled on cognitive tasks, generative AI looks like it could level the human-capital playing field.
Abundant energy complements average people’s skills
If generative AI is a metaphorical power tool for your mind, we shouldn’t forget the importance of actual power tools for your body. Physical technologies like cars, construction machinery, and many machine tools amplify the power of the average person; a math genius is not likely to be a much better driver or backhoe operator or drill press operator than a random person off the street.
And physical technologies have one main input: energy. From the 1800s through the 1960s, Americans used steadily more energy per capita. Then in the 1970s, the oil shocks put an end to that long party, and our energy use suddenly flatlined:
This is probably a big reason why innovation shifted from “atoms” to “bits” in the late 20th century — why building faster vehicles and new power-hungry appliances became less important than creating faster computer chips and better software. That shift seems likely to have accelerated the rise in the college earnings premium, and skills-based inequality in general.
Well, I have some good news: Energy is about to get cheap again. There are two reasons for this, both of them technological. First, primary energy generation is being revolutionized by a new cheaper technology: solar power. (My apologies for posting this chart yet again, but it’s a really important chart!)
The second key technology is advanced batteries, which are making energy portable in a way that even petroleum often couldn’t manage. I wrote about the implications of universal energy portability in a post a year ago:
Energy that’s both cheap and widely portable will enable all sorts of economic activity that average people will be easily able to master. Battery-powered appliances and industrial tools, robots that can be ordered around with generative AI, cheap chemical manufacturing and earth moving, fast efficient vehicles with long ranges, 3d printers, and so on. The power of every construction worker and factory worker and food delivery worker and nurse will be magnified by the new energy abundance. Increased productivity will lead to higher wages (as it does tend to do, despite certain viral graphs you may have seen floating around).
In other words, cheaper, more portable energy seems like it’ll help put us back on a technology curve more like the one we were on before the 70s.
Hints that the Revenge of the Normies is coming
Of course, all this is purely speculative. But I do think it’s worth noting that the Revenge of the Normies story may already be underway. The college wage premium has actually shrunk in recent years, as the incomes of the less-educated have outpaced that of degree holders.
And this is also reflected in general wage trends, which show wages for the people at the bottom of the distribution growing faster than those at the top over the past decade:
Obviously, generative AI and cheap abundant solar and batteries have yet to make their impacts felt on our economy to a significant degree, so there must be other reasons for this trend. A tight labor market has obviously been involved, and decreased Chinese import competition and a glut of college degrees may be factors as well.
What these basically amount to are tailwinds for the Revenge of the Normies thesis. Benefits from new technologies that level the playing field will come on top of whatever else is driving down wage inequality and the skills gap. So while we can’t know when those trends will end, they’re certainly an encouraging sign.
The Revenge of the Normies thesis is, of course, an exercise in optimism. As flattering as the age of human capital was for my nerdy tribe, such a small slice of society shouldn’t be the only ones who get to thrive. We’ve had a four-decade-long celebration and veneration of talent and excellence in America; we could use an equally long period where the vast middle class and working class are the people who reap the most rewards.