Editor’s Note: The following is a response to Jared Gould’s recent piece, “Welcome to the Unemployment Line, Graduate.” The author—a retired scientist and former recruiter for engineering and research roles in the U.S., Europe, and China—offers a critical perspective on the current higher education model and shares a more optimistic view of how young people might approach the future of work in the age of artificial intelligence (AI).
First of all, far too many young people attend college right out of high school. Most enter college with not the slightest idea of what they want to do with their lives. Attending college to “get the college experience” or to “find your passion”—I see this nonsense in college recruiting advertisements—are very poor reasons to invest four, and more commonly now, six years of one’s life and spend buckets of money or accumulate mountains of debt.
Far too many young people graduate with degrees that have provided them with little in the way of marketable knowledge or skills. I am not the first person to point out that university should be reserved for those students who are prepared for a rigorous education, full of mentally challenging coursework, whether it is in the liberal arts, social sciences, or STEM fields. Perhaps 20 percent of high school graduates are prepared for such an experience, a percentage I suggest hasn’t changed since I was a university student in 1979.
The fact that so many high school students are funneled into university by both parents and high school administrators is, in my opinion, unforgivable. Fortunately, this scam finally seems to be on borrowed time. If the AI revolution accelerates the process, that’s a good thing in the long run, although it may be detrimental to current new graduates.
As for the current AI angst that is parroted endlessly by many members of the mainstream media and the blogosphere’s commentariat, anytime there is disruptive economic and technological change, there is also massive opportunity. As always, opportunity favors the prepared mind. Young people should be asking not what AI will do to them and their prospects, but what they can do with the new toolbox that AI will provide for them.
By “what they can do with AI,” I do not mean using it to write that essay that is due tomorrow or to prepare their resume. Instead, how can they use AI productively as a tool for solving problems and providing insights? Any fool can obtain an answer from AI, but a knowledgeable, dedicated user can utilize it to help expand their knowledge and understanding in a way and at a rate previously unattainable.
Universities are missing the point in worrying about students using AI to “cheat.” This is a self-correcting problem. Such students will have little to offer in the real world of work, and word will get around that it is a route to unemployment or preparing someone’s espresso.
I devoted a significant amount of time to recruiting scientists and engineers—both BSs and PhDs—from major universities in the U.S., Europe, and China. When my organization interviewed people, we always employed “whiteboard” sessions, similar to what most major IT companies do. Either they knew their stuff or they didn’t. AI can’t help an interviewee in this environment. Students who don’t understand this do not belong in the university in the first place.
There is no doubt that AI is replacing conventional coding. In fact, I have been told that the most advanced AI systems can produce very sophisticated code and are being used in many coding environments to improve productivity. Clearly, this change has made “just learn to code” look like the blithering idiocy it was when it was stated.
However, I think the current, popular view of AI is somewhat myopic. Virtually every seemingly intractable, society-scale problem we face in the world today involves complex adaptive systems that cannot be understood using the time-tested scientific method, which relies on reductive experimentation. There are simply far too many variables that interact in multifarious ways. More often than not, reductive experimentation leads to the infamous unintended consequences—economics, bank regulation, epidemiology, as seen in the COVID-19 response, most social welfare programs, climate, geology, population dynamics, genetic and metabolic regulation, etc.—that are one of the principal products of government regulation and intervention. AI is going to be critical in addressing such problems, and there will be enormous opportunity for well-prepared humans in this quest since mountains of new, high-quality data will be needed if any AI system is to make progress.
Humans will be integral to the design and implementation of systems to acquire that data, which AI will then be analyzed under human direction. This is going to be an exciting new world of problem-solving in fields that have resisted or responded negatively to human efforts to make progress in the past. Work in this kind of economy is going to require people with technical skills that are not taught at university, but rather at technical trade schools and through apprenticeship programs, as well as properly prepared university graduates. I believe we will still need liberal arts and social science graduates, but they will work in ways far different than those being taught today. There is no future world that will not require people with deep STEM expertise.
It seems to me that we have raised a generation of young people, many of whom are genuinely afraid of change and of the future. Adults have much to answer for here, particularly helicopter parents and the education establishment. These should be exciting times for young people. I remember Walt Disney Presents “Tomorrowland,” which I watched on a black-and-white Motorola television set as a child in the late 1950s. It presented an exciting and inspiring vision of the future. What we are living in today is hardly Tomorrowland, but we have an opportunity to create our version of Tomorrowland with the AI revolution.
We need young people who are eager to help create Tomorrowland, not those who will be afraid of it or, worse, resent those working on it. I suggest that every young person should be exposed to Marc Andreessen’s “The Techno Optimist Manifesto” or something equivalent in high school. Young people should be optimistic about the future, not afraid of it.
Enough of the Negative Nellie, woe is us vision of their future. They deserve better from the adults in the room.
Image by cherdchai on Adobe Stock; Asset ID#: 1534889372
I remember the early ’70s when each year’s model was longer than the prior year’s and people had to keep expanding their garages — and then things suddenly changed in the winter of 1973-4.
What’s not being said about the projected use of AI is that it will be consuming twice our total existing electric production capacity. Let me rephrase that — not only will it suck up every watt of the electricity we have today, but that we will have to double everything (power plants, power lines, substations) just to feed AI, while we sit in the dark. Millennials won’t even be able to recharge their smartphones…)
Right now we really don’t have enough capacity to produce the power we need while NIMBY activists protest everything from coal trains (feeding power plants) to power lines and substations in their neighborhoods. Electricity is already exorbitantly expensive in much of the country, something being exacerbated by worshipers of the Goddess Carbon, who somehow think that the solution lies in pixie dust and unicorn flatulence.
Microsoft literally buying a nuclear power plant should have caught people’s attention — Microsoft recently bought (and restarted) the reactor on Three Mile Island that didn’t melt down in 1979 — it plans to use the entire plant’s production for just one AI center. Amazon is doing something similar which gives an idea of the obscene amount of electricity that they are planning on consuming.
Where is it all going to come from, and who is going to pay for it? First pay to triple our current electrical production capacity, and then pay for all that electricity, and then pay for all the computers using it, and then pay for a profit for the companies using it — they’re in business to make money and are going to demand their money as well.
And one other thing that no one is mentioning — while the physical size of the internet cables needed to carry an ever-increasing bandwidth hasn’t been an issue yet because a single strand of glass is so much narrower than the dozen pair of twisted copper wires it replaced, we very quickly are going to get into NIMBY issues with increasingly thick and unsightly cables being either strung on poles or requiring neighborhoods to be dug up to install underground conduit runs.
This isn’t going to happen. The necessary resources aren’t going to be provided, and people aren’t going to pay for the product.
The author makes a key point about the general problem of attending college right out of high school. He doesn’t mention what young adults should be doing (and should have already been doing): working. Working at hard, manual labor perhaps: farming, manufacturing, transportation, construction, services, and the military, for example. This also raises of course the question of high school quality which is perhaps a larger education problem than college. Readers may appreciate a note I made in the New York Times: https://www.nytimes.com/2018/08/25/opinion/letters/benefits-college.html
I think high school graduates (they are not adults legally, developmentally, or experientially) should consider other options for careers, such as training that involves certification and training in skilled trades, of which there are many. And high school councilors and parents should spend time over grades 9-12 providing information about such career choices rather than focusing on university as the main option. And, for those who do take the university plunge, I think any work experience, or a stint in the military, and the associated maturity it might bring, would be worthwhile before a majority make a decision as big as deciding to spend 4-6 years and lots of money in a dead-end program, of which there are many. They would be better off making a decision informed by experience and maturity that are not typical of an 18-year-old in the 21st century.
As for the quality of the high school education, I agree. We know that our K-12 system is woefully deficient. I have some (but not much) sympathy for universities in that the product they get from our K-12 system is inadequate and has gotten worse over time. I do not remember there being remedial courses when I entered university back in the 70s. I find it difficult to be optimistic about correcting this national disgrace in the foreseeable future.
“Either they knew their stuff or they didn’t. AI can’t help an interviewee in this environment. Students who don’t understand this do not belong in the university in the first place.”
I guess the bosses can be snarky about AI, and vice versa the students with the bosses (as long as the latter still have their jobs). Robert Humphreys being safely retired gives him greater leeway.
I guess students who still don’t get it about AI “do not belong in the university,” and perhaps never did? — they’ve been exposed to AI for all of a couple of years.
I think everyone is getting used to this new technology. A lot of middle managers are terrified of losing their jobs. So are a lot of beginners — I read just yesterday about how fresh economists are having a hard time. Companies seem to be proud of how many people they can reduce from the workforce with AI — another story this week.
Pros — who actually use AI in reseearch — say that the way to prosper in the age of AI is to keep control of the technology, mainly by cultivating the capability of “critical thinking.” That sounds pretty much like the old advice about how to keep employers happy.
I do think it’s kind of early to be expecting students to understand this when they arrive at college. It is something faculty should try to inculcate in the students — while the faculty are kind of shakily trying to understand this new world themselves.
I watch my young grandchildren on their iPads using Minecraft and playing video games and I marvel at how they adapted to this world (a strange world to me) with virtually no assistance. My not-yet-6 grandson routinely beats me handily in video games and brain games requiring rapid response. Pretty much the same can be said of any high school graduate today. Young people in K-12 today are, without a doubt, the most protected, pampered, and safest generation in the history of the human race. Yet, when I have spoken to high school graduates, including several who are experienced coders already, they seem genuinely worried, perhaps even afraid, of how AI will affect their prospects. There is something incongruous here, I think. Young people were once early adopters of anything new and interesting, especially of it left the adults behind.