#18: AI + College

The Case for the Liberal Arts Degree

Writing this piece feels wildly contradictory. I don’t like what I’m about to say.

I went to a pure business college. My graduating class was made up of nearly all business majors (finance, accounting, marketing, economics) and the small minority studied business-adjacent fields like math or public policy. The core curriculum was heavy on those four business subject matters with little room for flexibility until you reached second semester of junior year. That’s a long time to tough it out if you want to study something other than business, so a relatively high percentage of students transfer after their first year. So yeah, the “nearly all” comment includes some survivorship bias.

In the mid-2010s, the mantra echoed in my ear was: learn something useful in college so you can get a job. Don’t waste your four years of undergrad to walk off the graduation stage without a plan. Employers want you to be able to hit the ground running on day one, so you better be prepared or you’ll be left behind.

I followed this advice. And I’d say it’s worked out for me. Work is nothing like school, but my rather technical business education did help flatten the learning curve when I started my first job in Corporate America.

Today, I believe the interpretation of this mantra is different. In my college years, “useful” meant learning technical skills like accounting or finance. Immediately applicable skills that I can apply right away to a business role. Nowadays, a lot of this technical knowledge (particularly for entry level positions) can be executed by AI applications. You can ask AI any question you have, ranging from help with a formula, to instructions on how to use a software program, and even write the software for you! So, why memorize a finance formula when you can simply ask AI? For the sake of an exam? Come on now, that’s silly.

Since I can ask ChatGPT any question in the world and it’ll spit back out a single, curated answer, is learning technical skills still necessary?

Shopify’s CEO published a memo last week asking employees to justify why AI can’t do the job before hiring an external candidate for a role. Additionally, several large companies (ie Citi, Nokia) have developed their own GPTs to power internal problem solving (likely opting to build proprietary models over relying fully on external LLMs, mainly due to security concerns). Before we know it, employers will go from encouraging AI adoption to requiring it. I’d guess we are in the grace period right now, while companies are still figuring out the most relevant AI use cases for their respective businesses.

In my opinion, the definition of day one job productivity has changed. It’s becoming increasingly important to know how to communicate with AI. See, prompt engineering doesn’t require too much technical knowledge (despite the phrase having the word “engineering” in it), however you do need to structure your thoughts logically and clearly. Top prompt engineers possess strong thinking and clear communication, over a knowledge-based skill set.

Reading and writing won’t go out of style. If anything, they become even more in vogue because the ability to leverage critical thinking skills and structure your thoughts in written form becomes necessary to receive desired results from an LLM. Also, we still work with others so effective oral communication is crucial to success. Companies may become more labor efficient, but they still will employ people. Understanding how to effectively collaborate with others will set you apart from your peers, particularly as you progress in your career. As mentioned in AI + School, communication (both with technology and humans) will become the paramount skill in an AI-driven world.

And that’s where the Liberal Arts education earns its keep. From English to psychology to sociology, building a strong critical thinking base is essential. Today, the “what” is easy to find (thanks to AI). The “why” is where we still need people. Technical skills help bridge the gap to “what”, but a strong understanding of how to think and communicate allows you to arrive at “why”.

Although, those that are truly exceptional at a technical skill should stick with further developing that skill. If you’re a top 1% graphic designer or software engineer thanks to such in-depth understanding of your field, you can rely on your talents. Don’t worry about what you studied in college as you’ve figured it out. Being exceptional is still the best moat against big macro shifts, however I will argue that the definition of exceptionality will begin to include knowledge of relevant AI applications over time.

A Liberal Arts education not only contributes to your ability to effectively interpret an LLM output, but it also helps with analyzing real world situations. Your college education should prepare you for the workforce, especially the workforce at the time you plan to enter it.

With all of this “prepare for a world of AI talk”, I’d don’t want to overburden you. I’m not saying you need to become an AI savant. That’s not necessary today. But you do need to be self-aware of how your strengths and weaknesses ladder up to what’s needed in the workforce. And then adapt.

I’ll leave you with this thought.

In a world where information is infinite and access is nearly frictionless, what kind of knowledge actually matters?