Recently I suggested that one of the hottest jobs over the next five years will be that of  prompt engineer--someone  who acts as an intermediary, interpreter and translator between the real world’s inquiries and the universe of large language models (LLMs), neural networks, and chatbots of all kinds. The reactions that I received were mainly directed to my suggestion that (a) the skill sets required to be a successful prompt engineer were almost entirely non-technical and (b) that I wasn’t even certain that an expensive four-year college education was required.  I come to this perspective having taught at the executive, graduate and college level for more than three decades. 

This may be a case where the smartest idea for a recent grad is to jump off the cliff and hope that you’ll grow sufficient wings on the way down to survive the fall. Take an entry level position at a reasonable salary where you can learn the tools and do the tasks that will drive the future. The fact that you don’t cost the company a fortune to bring on is a blessing - more freedom, more flexibility, and no one wondering why they’re paying you a ton of money to sit around and “play games” with the computer. A year or two from now they’ll wonder how they ever got along without you.

The talents and skills I felt are most likely to assist in the process are much more mundane. You’d need some serious confidence in your own abilities, a desire to take on the unknown, and an appetite for risk and challenges. And you’ll need a lot of common sense, an extensive vocabulary, the ability to constantly iterate and tighten queries on a wide range of subjects, and a passion for problem solving, crossword puzzles, Scrabble and Wordle. Readers, writers, liberal arts grads, and kids straight out of high school would all be welcome. The job is all about aptitude and knowledge, not college; what you know, not where you go. It would also probably help to be hyper-literal and a little anal as well, but that wasn’t essential. Think of this as an on-the-job vocational training program rather than a couple of years spent being taught yesterday’s technologies by people far afield from the firing line.

The real lesson is that narrowly-trained engineers and computer science students are likely to be too modest and conservative -- too measured and concrete -- to have the ability to expand the scope of their inquiries.  And,  as trite as it has become, to think outside of the box and beyond the typical constraints of the problem. They are largely focused on optimizing and streamlining existing systems and solutions, not leapfrogging to the next ideas. This type of focus narrows attention to incremental improvements rather than quantum leaps. It makes you think you’re approaching an apex when there may be huge additional opportunities that lay beyond that point.

Success in this particular space is going to be all about the ability and the willingness to bite off far more than you can chew. Moonshots, crazy guesses and bets, and wild ideas are going to be the everyday starting point for some of the challenges we’ll be facing. Falling flat on your face - when the machine calls you an idiot - will be par for the course. The saving grace is that iteration and experimentation is at the heart of prompting and what the whole process is about.  It’s exactly how the queries narrow and tighten over time until they deliver real results, including many findings, conclusions and revelations that no one initiating the process even imagined.

As it happens, I wasn’t thinking broadly enough myself. Prompt engineer is only one of the many new careers that the AI revolution will create while it eliminates thousands of others. As “evil” as automation is sometimes regarded by those whose routine and rote positions will be replaced or eliminated, AI-driven augmentation will make millions of other workers far more effective and productive, as well as creating virtual assistants, and ultimately AI will demand the creation of millions of new positions. We’ll need robot trainers, designers for cyber ecosystems and worlds, architects for new virtual cultures, and AI ethicists across the board to try to keep the many unforeseen genies in the bottle.

But, even more importantly, just as with prompt engineers, there will be a growing premium of new skill sets that our current colleges and universities do a crappy or non-existent job of teaching. Just as virtually no MBA program in the country teach sales, which is the absolute heart of every business, the old-line academics at the elite technical schools especially frown upon the very soft skills that will be essential in the new work world. We’re preparing students for careers that don’t yet exist, where they’ll be required to use using technologies that haven’t yet been invented, and asked to solve problems we don’t even know are problems yet.

Broadly, the new and required skill sets fall into three areas:

People Skills: including management, negotiation, service orientation, coordination with others, and emotional intelligence.

Problem Skills: including complex problem solving, critical thinking, sound judgment and decision making.

Polymath skills: including creativity, conceptual capacity and cognitive flexibility.

Hopefully, new schools and centers will be created to address these needs before the flood of demand. I’m not that optimistic because I’ve spent far too long with hidebound professors and lecturers who simply don’t appreciate or understand these “squishy” talents and abilities. That’s in part because they’re so hard to quantify, measure, and evaluate  and these educators have no interest, aptitude or ability to teach them or desire to change the old and ineffective--but easy-- ways they’ve always tried to keep score.

Unfortunately, by virtue of seniority combined with the juvenile politics that infect so many academic departments and endeavors, it’s far more likely that the ultimate solutions will come from new groups and enterprises. We need disruptive and innovative approaches, and new providers rather than the brand-name, unbearably expensive, and totally backward elite technical institutions that presently dominate the space.