How to Decide if College is Worth It for You in an Age of AI (2026)

I teach college admissions exams for a living, so obviously I’m expected to say that college is worth it. For the students I tutor one-on-one, going to college is a no-brainer for them. But for a lot of people I talk to outside my job, when I say what I do, they ask… do you really think college is worth it in the age of AI?

My answer is more nuanced than I am able to reply in a few sentences, so I wrote this article to send to them. No, college isn’t always worth it. But instead of diminishing the value of a college education, AI actually enhanced it.

The misleading college ROI

Finances are important. At some point, when looking into colleges, someone, whether it’s the student or the parents, or both, looks at college ROI. (Here’s a great website to do that!)

In purely monetary terms, many people have noticed that some college degrees and some colleges aren’t worth it. (You can’t put a price tag on happiness ha-ha). Here’s the rub. Looking at the ROI for a specific college or a specific degree and expecting current trends to hold is mistaking past performance for future guarantees. That’s to say, STEM fields have done well in the past, so they’ll continue to do well. While logic is somewhat sound, it’s uncertain. The more interesting examples I’ll get to, though, are about the value of the humanities - which could deductively be inferred to see an increase in value because of pervasive LLM usage.

Worked ROI Example

The problem isn’t ROI calculators. The ROI calculators have good data - I mean that a data hygiene perspective. I’ll show a few examples of college ROI math to expand on what I mean by the ROI calculators aren’t the problem: it’s how we use them that is extremely misleading.

Let’s consider an English degree:

Ziprecruiter says that the average English major earns 58k. Again, the data is actually really good - it goes into median, noting that the median spread is wide, meaning English majors are all over the place in terms of expected ROI. This is good, well-parsed data.

Before I can get to why using ROI calculators to make decisions may be unadvisable, we have to first talk a little about statistics. A lot of people think of the term “mean” as in “average US income is $69,770” but the average US worker makes that much in a year like they have 2.5 kids: they don’t.

The average US worker doesn’t earn the average income. I know it can sound confusing, but those are two different uses of the word average. The average worker - means the most common, everyman. Average income means statistical average. Those two things are not the same.

With income, the average is misleading because there are a few very high earners who move the average up a lot. The median is more useful in this casen: the median US worker earns $46,960 - which is a lot closer to what the “average” US worker can expect to earn. That’s a difference of $22,810. If you try to make life plans based on an average - which people do all the time - you will be sorely disappointed when you reach the median.

This article will answer the question whether college is worth it: the answer is not quantitative: it’s qualitative. College can be immensely lucrative. But there’s a whole separate plane on which it’s wrong to compare ROI by percentages spent versus percentages earned. Finances are paramount… they just don’t really speak to the math behind college. The reason is because ROI doesn’t tell you what someone is motivated to do. Or how things evolve.

Black Swan Events

It’s hard to make an apples to oranges comparison, which is what people are trying to do when they say “average ROI for college degree is x.” We don’t have to look back very far in history to see why it would be inadvisable to plan to obtain a certain college degree based on ROI.

Let’s consider the advent of computer engineering as a major - an undisputed “high ROI” college major. An official study of starting salaries of college grads from 1960 to 2015 couldn’t even include computer engineering because the discipline didn’t exist in the 1960s.

Computer Engineering as a field emerged in the 70s, and grew throughout the 80 and early 90s. But the discipline didn’t see explosive growth, nearly doubling during the time period of approximately 1994-2004. The following people graduated with computer science or computer engineering degrees in the mid-90s, right at the start of the field’s biggest expansion. None of them could have looked up a college ROI calculator and predicted how well they’d do.

• Marc Andreessen: BS computer science, University of Illinois, 1993. He co-created Mosaic, co-founded Netscape, and is now a mega-VC.
• Sergey Brin: BS math and computer science, University of Maryland, 1993. Google co-founder.
• Larry Page: BS computer engineering, University of Michigan, 1995. Google co-founder.
• Brian Acton: BS computer science, Stanford, 1994. Founded WhatsApp, which sold to Facebook for about $19B.
• Max Levchin: BS computer science, University of Illinois, 1997. PayPal co-founder.

( Noticeably absent from this list is Steve Jobs because he studied English before dropping out of Reed College)

While finances are very important, a college degree is more than just an ROI. It’s a qualitative decision. And it’s one you and your child need to make.

Isn’t AI Going to Take Our Jobs?

First of all, it’s tricky to make the connection between going to college and getting a job in the field you studied. English majors don’t become English professors; sociology majors don’t become sociologists; acting majors don’t always become actors, and so on. I want to explore that connection more in detail. But first, having said the above…

No, AI isn’t going to take our jobs. According to research by Stanford, AI’s impact on the work force has been small.

“There is also little evidence of AI depressing employment or job postings in the most highly exposed occupations. Employment trends in occupations with high exposure to AI are fairly stable. While employment growth in coding-heavy occupations has slowed somewhat, it remains positive. There is no evidence that AI adoption has negatively impacted firms’ job postings. Indeed, online job postings for software developers - a very highly exposed occupation - have been growing faster than for other occupations over the last year. Among firms that adopted enterprise AI, employment grew by 10 percent in the two years following adoption, an effect driven by firms with the highest per capita AI spending.”

The aggregate numbers don’t show AI destroying jobs. But new graduates are entering exactly the occupations AI handles best, and early-career hiring there has weakened. Researchers can’t yet say how much of that is AI and how much is interest rates, but it’s something to watch.

How does AI impact the decision to go to college?

That entry-level weakness is the strongest argument for what college is actually for. I’m going to pick up the thread about the connection between one’s college major and their job with a very bold claim. It’s more important than ever not to go to college for a job. You go to college to develop critical reasoning skills. If you need technical training, then you can go to post-secondary school: law school, med school, masters/PhD training. But the point of college is and has always been to develop critical reasoning skills. Let me explain.

Over the past 30 years, college costs have essentially doubled. There are a number of reasons for this - not many of them good. Enumerating and discussing them is outside the scope of this essay. As previously stated, finances are important. The increased costs of college compared to the expected ROI has made many question the value of a college education in the past 30 years – especially when there’s no hands-on job training. That’s understandable.

To improve college ROI, well-intentioned people have then added hands-on job training as part of the college curriculum (meaning job training outside of internships). But I contend that this move has been misguided. As technology develops faster, as evidenced by the advent of LLMs, any hands-on applications with high technology are likely to become quickly outdated. It’s common knowledge that software engineers learn principles, not programming languages. Similarly, hands on training is actually a bad use of a college degree. Because the one thing that is very difficult to develop and that colleges can actually do a pretty good job at cultivating is: critical reasoning skills. That’s the ultimate job skill that you can’t quantify or measure but will, and does, reap rewards many times over.

Why College (and a liberal art degree) is More Worth It than Ever

To return the question posed in the beginning: is college worth it? Especially given AI. My answer is: if college is a good fit for your son or daughter, there’s never been a better time to attend. But the problem isn’t actually college ROI as such - it’s that too many people are going to college for the wrong reason. I think people with humanities degrees will be extremely sought-after over the next few years - perhaps like the good software engineers are now. The reason has everything to do with how LLMs work (or don’t work.)

LLMs don’t know anything. They’re basically glorified auto-correct machines. When your phone thinks you’re going to spell “thinks” based on the you typing in “th” - that’s the same principle LLMs operate on. Except they do so at such a large scale, people (neurotypicals) ascribe human thoughts to them - such as the ability to reason.

But LLMs have no capacity to reason. They know which response it most likely to satisfy the user based on the user’s input - and that’s it. But they don’t know whether the user should be doing what they’re trying to do. Ethically, morally, financially. There are some guard rails in place, but fundamentally AI doesn’t know anything.

Some people have tried to distinguish LLMs (ChatGPT, Claude) from AGI: AI from Artificial General Intelligence. The word “general” is meant to distinguish AGI from AI by denoting AGI that can reason. The distinction is needed because AI can’t reason. According to the top researchers (AI through the eyes of experts, the book), we don’t currently know what AGI will look like or how we will get there. The LLMs are not AGI, and they can’t reason.

But people keep thinking AI can reason. And that’s going to be a problem from time to come. For a comparable analogy, ask people who were in business in the 90s what it was like when Lotus was more popular than Excel. They’ll tell you that people would put their business model into Excel (or Lotus) and find out that they were going to become overnight millionaires because the spreadsheet told them so. Sound familiar? The only difference between then and now is AI is far more convincing than a spreadsheet - and that’s because the AI companies keep trying to make us think LLMs are so much smarter than a spreadsheet.

But the point of this essay is to help parents or students articulate the value of a college education. The value of college is that it helps you make sense of the world and be on the cutting edge of solving problems. It’s not a golden ticket to a brilliant career, nor is it the end of all of your hard work. A college education is valuable if you want to do the things a college degree can set you up to do.

The Qualitative Decision

I’m asking what kind of life you want to have. Do you want one where you work really hard but don’t need to drown in conflicting opinions and research to make what seems like a simple decision? If you answered yes to that question, then don’t go to college. But also understand what you’re potentially forfeiting.

If you want to work really hard, carefully study problems, learn how to constantly shift your perspective to understand different points of view, expose yourself to more topics than you can possibly master - then college may be the right choice for you.

The important distinction between these two decisions is they’re each a different lifestyle - not that one necessarily works harder than another. Nor is it true that one’s better than the other. It’s also untrue that one option is universally better for every person. The option needs to be aligned with the person. That’s why this guide is helping you choose the right option for you. The question you have to answer is: Is college worth it for you?