The Hidden Profits in AI
Recently, I happened across a bit of news on the “robotization” of yet another industry.
I’ll tell you right now — I don’t feel sorry for these particular workers in the least. They were overpaid and underworked.
They deserved to get put on the unemployment line. Good riddance.
Burger flippers? No.
Fruit pickers or warehouse workers? Nah.
It’s not who you think…
Try Wall Street traders — and soon, their investment banking cousins too.
A recent article in MIT Technology Review sheds light on the emerging opportunities in artificial intelligence (AI) — a trend that Profits Unlimited Editor Paul Mampilly has made a centerpiece of his investment strategy.
Back in 2000, Goldman Sachs had 600 traders buying and selling big blocks of stocks for its institutional clients. Six-figure salaries were a minimum.
Today? Goldman Sachs has just two people doing the same job. Automated trading programs handle the rest.
Just the Start
A casual reader of The Wall Street Journal or Michael Lewis’ Flash Boys might have guessed the change (for a variety of reasons besides just “computers”). But that’s missing the point — the trend toward this kind of automation — AI, really — is just getting started.
As the article makes clear, investment bankers (average salary: $700,000) are in the crosshairs next.
Don’t get me wrong. The “relationship” aspects of the job — the boardroom pitch sessions, holding the hand of a CEO to make sure he doesn’t lose his nerve (because he’s about to spend billions in borrowed cash on a merger) — those will always be part of a top investment banker’s tool kit.
But all the junior types back at headquarters? Many of their jobs will soon be toast.
It used to be that an investment bank needed lots of MBA-toting associates. A computer couldn’t catch the potential errors embedded inside a 250-page legalese-filled prospectus (such as whether all the numbers on Figure 5-A, Section 2 on Page 47 all corresponded properly to the related material on Figure 2-C, Section 5 on Page 73 — you get the point).
But no longer … thanks to continuing advances in AI software.
“Goldman has already mapped 146 distinct steps taken in any initial public offering of stock,” writes MIT Technology Review’s Nanette Byrnes, “and many are ‘begging to be automated.’”
Mohit Joshi, the head of Infosys’ financial services division, recently laid out all the places where AI will go to work:
- Customer service (with intelligent digital assistants).
- Data-backed lending decisions.
- Fraud detection.
- Personalized wealth-planning recommendations.
- Biometric identification.
And perhaps it’s no surprise that banks and other finance-oriented institutions are jumping on the AI bandwagon. For instance, Royal Bank of Canada recently created an AI laboratory, hiring a top AI expert as the lab’s adviser.
My colleague Paul Mampilly already laid out a great long-term pick in Profits Unlimited recently for a company that ought to be a huge beneficiary of AI. But the opportunities now and in the years to come are going to be numerous.