We Can Use Public Education to Save Us From AI
(Originally published on www.disgruntledrationalist.com)
I recently gave a talk to a group of young people at the Montana Learning Center, and a large part of the discussion centered on artificial intelligence and what that might mean for their future career options. One of the more engaged members of the audience raised her hand and commented that she thought teachers would soon be automated away. I learned later that she is a particularly self-motivated human being and has already taken advantage of internet resources to learn how to code, but my response to her highlighted the as-yet unconquered area of emotional AI.
Think about your favorite teacher of all time. Why did that person stand out? You are probably not thinking to yourself, “Well, she really kept me on task.” Even if you are remembering special, individualized attention — something that AI facilitates — the most likely adjective is probably “inspiring.” So far, Silicon Valley has yet to produce AI that inspires on an individual level. That is not to say it won’t, just that it is a trickier problem than replicating specific tasks and automating pattern recognition. As highlighted in a previous post, many other knowledge-oriented professions will be fully automated before teaching.
A lot of AI optimists have suggested that, instead of replacing workers, algorithms and robots will be used to augment humans. A recent article in The Economist outlines some of the ways that technologists are trying, with varying levels of success, to infiltrate education. Most of the examples show technology allowing teachers to make better use of their time and resources — for instance, by analyzing students’ work, identifying patterns of mistakes, and suggesting specific exercises to target areas of weakness. Full replacement is still a future proposition but a distant one.
It does bring into focus, though, the purpose of education, particularly if it is publicly funded. In the United States, public education had strong support from many Founding Fathers, as they pointed out that democracy only works if the electorate is educated well enough to understand the issues at stake when they vote. As one of my childless, retired cousins recently explained to her cost-conscious neighbor, she is more than happy to pay taxes to support the public education of other people’s kids, because she wants voters who are able to understand abstract concepts like intergenerational transfers. Without public funding for schools, government disintegrates into something more akin to an oligarchy, as those with money to afford education manipulate the votes of those who can’t. It might be argued that governing has grown so complicated and campaign finance so byzantine (thanks to Citizens United) that something similar already occurs, but it is still worth reexamining why public education should exist, especially as we enter an era of increasing automation.
After the launch of the Sputnik satellite in 1957, our schools become weapons in the war against the Soviet Union, and demand for students trained in the STEM fields increased. That has dovetailed with an increasing focus on education as training for an eventual job rather than as preparation for participation in civic life (or as an end in and of itself!). Contrary to many media accounts, the humanities have not experienced declining enrollment numbers, but the prestige of a humanities degree has suffered, especially in the minds of parents writing ever-rising tuition checks. As with many things, education — especially higher education — has gradually become viewed more as a private investment in an individual’s prospects rather than as a public investment in democracy’s future. Not surprisingly, that changes the return-on-investment calculation. Neither the students themselves nor the parents writing the tuition checks are likely to take into consideration the value of universal literacy or broadly held basic math skills, let alone a shared understanding of our country’s founding values and aspirations. This is what some sociologists have termed civil religion, and its power to promote unity in the cultural melting pot of a diverse society like that of the United States should not be underestimated.
Diversity — especially diversity of thought — can be a boon to innovation, but taken too far, it can be polarizing, as has been evident over the most recent election cycle. Despite its early promise for buttressing democracy, the internet and particularly the algorithms that underpin our navigation of it have proven to be more of a force for division. Do we want an atomized society of people addicted to their various screens? Perhaps that doesn’t sound so bad to some people: anyone with social anxiety will be freed from having to interact. However, that level of disengagement is a perfect environment for the rise of tyranny, driven by the interests of the few over the apathy of the many, a situation that never ends well for dissenting voices. The way to avoid this is to foster civic contribution and participation, and publicly funded education is an obvious platform.
More importantly, as we face a world where work may be obsolete, we should be encouraging people to discover “the pleasure of finding things out.” Although we may say, under the guise of promoting creativity, that we encourage it, what we really mean is that employers like workers who can “think outside the box.” In an age of angst about public spending (which, outside of supporting old people, is actually below its long-term average as a percentage of GDP), no policymakers are suggesting funding education that might have consequences that are not measurable.
The real tragedy amidst all this is that our education system, with its insistence on measurement and short-term accountability, is focusing so heavily on the skills of the immediate future rather than thinking longer term. While the much-hyped STEM fields are wonderful for building analytical thinking skills, so is philosophy, and philosophy will still be open to humans long after algorithms and machines have dominated careers in STEM. The time horizon for this shift is open to debate, and engineering majors still command a higher average starting salary than philosophy majors, but if the only pursuits available to humans in the future involve abstract thought and creativity rather than concrete skills and practical applications, then we should start actively fostering those now.
One way to do this while staying within the current STEM-oriented rubric would be to place more weight on basic research and deemphasize applied research. As has been widely reported, overall public funding for research, including both basic and applied, has been flat for more than a decade. More recently, though, private funding from corporations for the former seems to have increased. Although the data aren’t conclusive — not all corporations divide their R&D spending into basic and applied when responding to surveys — some might use this as a counter-argument to my push for public support of abstract thought and creativity. After all, public-private partnerships are all the rage in investment circles, and if purely private funding leads to a breakthrough, everyone will eventually benefit, right? Unfortunately, while that is a possibility, it is far from a certainty. The Martin Shkreli case occasionally devolved into pageantry, but it also highlighted the important of intellectual property ownership. Pushing all basic research into private hands could result in many more Martin Shrkrelis.
Coming back to the role of education, though, as AI advances, it will primarily be aimed at solving specific problems: reducing traffic accidents and pollution via autonomous vehicles, microtargeting agricultural application of fertilizer and water, further streamlining manufacturing processes. As a concrete example, an electrician friend recently installed solar panels at a reservoir that serves the drinking water needs of a nearby community. Rather than needing human intervention, the site now has the ability to churn the water whenever automatic readings indicate that algae buildup needs to be disrupted. Consequently, long before so-called strong AI — computer-driven intelligence rivalling that of humans — arrives, humans will need to find alternative ways of spending their time. This is not only a dilemma for each individual, since one of the most important lessons of the recent upsurge in populism is that humans like to feel useful and appreciated. The civil religion feature of public education that has always been partially about building unity, and thus holding chaos at bay, will be even more important in a world where work is obsolete.
For most of human history, purpose was the purview of metaphysics and theology, something that people looked for at church, not in their daily labors. Work was drudgery at the behest of and for the pleasure of someone more powerful. Capitalism’s triumph has been due, at least in part, to a fusion of purpose and labor. In recent centuries, capitalism has created an economy with more choices, allowing skills and labor to be better matched. One irony has been that the artisan class, which used to enjoy both a measure of prestige and also a more tightly linked sense of purpose and labor, has fallen in relative standing. The people who ply a particular craft to create tangible things (eg, cobblers, furniture makers, tailors) have been displaced first by more efficient manufacturing processes like specialization and then, increasingly, by machines. Nevertheless, some people persevere, fashioning items by hand as a hobby, often selling them at craft shows for less than the cost of materials. There is something more compelling than the capitalist drive for profit propelling them to create.
Perhaps part of the recent angst is the result of a century of both increasing specialization and escalating abstraction. Before the institution of the assembly line, workers were closer to the final product. Nobody can deny that it is easier to derive pride from saying, “I made that car,” than from, “I made the bolts that hold the wheels together,” even though the latter is undeniably important to the overall project. The same goes for amorphous output, as well, and with the growth of the service sector, fewer and fewer people can point to something tangible as the product of their labor. While “helping people” is something that humans do derive pleasure from, the spectrum is broad.
Most service jobs do not involve saving lives, and other jobs that involve making an indelible impact on individuals — teaching, eldercare, home health aid — usually do not pay in proportion to that impact. A capitalist economy rewards scarcity, not just skill. Consequently, we send a signal to the person who can say, “I taught twenty-seven children how to subtract today” that they are not valuable, even though nobody can balance their checkbook without understanding subtraction. At the same time, many service workers don’t have as obvious an effect to hold onto mentally. They operate cash registers or mops, sell unnecessary financial services, or, as in my case, move money around between different groups of rich people. It is harder for those people to determine their purpose, and even if they have an intrinsic sense that what they do matters in some way, it can be hard to convey to others.
That isn’t to say that being unable to concretely explain one’s job to people in other fields means it doesn’t matter. Intangible products can have tangible benefits; insurance, for example, allows people to plan more effectively for the future. However, the person selling the insurance — or managing the portfolio that provides the money to pay claims — doesn’t know, for instance, that the Jones family just bought a house because their recent life insurance purchase gives them confidence that their single-income household won’t fall behind on the mortgage in the horrible event of death. As our society and our economy have grown in complexity, it has become harder for any single individual to point to what they do to “make a difference.”
A previous post argued that literature can help assuage this sort of anxiety, and so can an overall shift in what we emphasize in schools. In addition to greater prominence for the humanities, where human minds are likely to compete with algorithms more successfully than in STEM, reintroducing offerings like home economics and shop class is a way to restore tangible creation to people’s daily lives. A reexamination of the satisfying creative outlets we have surrendered in the name of convenience is long overdue.
As we face a future without work, education needs to transform from a manufacturing process for building skills to a way to teach people how to find meaning and purpose purely from within. This sounds like an advertisement for a yoga retreat or a 1960s hippie commune, but the standard external signals of value — money, status, prestige — are likely to be beyond the reach of the majority of humanity within the lifetimes of today’s children. Although it is possible that technology will render these differentiations moot by raising living standards for everyone, making conspicuous consumption irrelevant, it is more probable that we will descend into a techno-feudalism mimicking the extreme inequality that has been reality for most humans throughout recorded history. Consequently, just as ambition was ineffectual for anyone but an artisan in the past, it will be futile for anyone to strive to accumulate, since we will either all be dukes or almost all be peasants. Differentiation and meaning will necessarily be the result either of craft or of mind, serving no further purpose beyond a sense of personal gratification.
I hope that the self-motivated youngster at the beginning of this piece is truly stirred from within, but for most people, there is usually at least a patina of ambition — to make money or lead a group or achieve renown — that complements or even supersedes a natural desire just to learn for the sake of learning. We need to start shifting education away from the cultivation of specific skills and toward an indulgence in the pure pleasure of finding things out. By doing so, we can shift our future of unemployment from bleak isolation to meaningful endeavor using a social institution we already have in place, even if we eventually decide AI instructors are inspiring after all.
