I Was Sour on AI Before the Jensen Huang Scandal
Let’s be honest, artificial intelligence won’t turn a novice astrophysicist into Neil deGrasse Tyson, but it can turn Neil deGrasse Tyson into a better astrophysicist. Therein lie the technology’s limitations.
I have heard the absolute statements about AI: that it will decentralize everything, allowing everyone to work just a couple of days a week and live like Jeff Bezos; AI will allow you to buy one $100 million mansion to prevent the peasants from seeing your other $100 million mansion; it will take all our jobs; and, if we are not careful, it will eat us all—the list of extraordinary predictions about artificial intelligence from its most ardent proponents can go on to infinity.
I have seen enough to confidently say that AI will structurally transform many aspects of our lives, but it will not fundamentally transform human civilization as it is being promoted.
Once the dust has settled, similar to when the dust settled on the dot-com bubble—by the way, circular financing could wreak havoc before the dust settles—AI will emerge as the best technology for white-collar workers, technocrats, and academics.
What all those things have in common is that all those folks lean on acquiring deep knowledge in their fields first, then leverage AI for deep and continuous learning. AI won’t turn an amateur journalist into Bob Woodward, but it can turn Bob Woodward into a better investigative journalist. It won’t turn a recent college graduate into a finance professional, but it can turn a finance professional into a better finance professional.
Considering the amount of money being spent and invested in AI infrastructure, it’s foolish to take the advice of someone who is semi-technologically literate, but I am telling you, if you look closely, the signs are already there.
Non-white-collar workers are using AI considerably less. Affluent Americans are considerably more likely to use AI chatbots than lower-income Americans. Matter of fact, only 17% of workers without college degrees use AI at work daily or weekly, according to Gallup. What the data are telling us, I believe, is that people often use AI less to acquire knowledge than to deepen knowledge.
As a finance professional, I know how to read an income statement and how information on the income statement carries over to the company’s balance sheet and statement of cash flows. Because I have already put in my 10,000 hours, I know exactly what to look for when analyzing a company’s finances. Is accounts receivable hiding something? How healthy is earnings before interest, taxes, depreciation, and amortization (EBITDA)? What is being amortized? Did the company overvalue the goodwill of a company it recently acquired? That often takes a lot of work, but now, with AI, I can get what I am looking for in seconds—there lies its power. And influential technology executives might justly argue that the nearly $5 trillion projected to be spent on AI infrastructure worldwide from 2025 to 2027 can be justified.
AI cannot put in the 10,000 hours on your behalf. It’s only after you put in the work that you can best leverage AI to deepen your skills.
White-shoe law firms, big consulting companies, and big Wall Street firms are starting to realize the same fact as well. Lots of folks were prognosticating the doom of entry-level positions for those who want to be in the aforementioned fields; however, new hiring data show a mixed picture.
Entry-level hiring has continued in some of those positions, even as it has declined in others. AI can generate new ideas, but it lacks lived experience. Although AI can compile all the research being done on the Capital Asset Pricing Model (CAPM) and yield a result if I ask it to explain it to me, it takes a financial professional with lived experience to recognize its flaws and judge whether a better model will work in practice.
That’s the crux of this argument. There is no evidence that AI is conscious—it does not know the sting of being passed over for a promotion for someone less qualified, it does not feel the emotional drain of Black professionals who have to code-switch, and AI has never had to work in an environment where it has to learn to cope with microaggressions. AI has never fallen in love or been heartbroken. AI can explain to you what a Proustian moment is, but it has never tasted a madeleine cake—those are the lived experiences that make their way into our professional work. AI never can!