My father’s first job after graduating University of Pennsylvania law school in the 1950s was not in law, but computer programming. In 1956, the U.S. Social Security Administration installed
a massive Univac computer—housed in a 10-story, city-block-spanning building in Baltimore—to automate Americans' earnings and benefits records. However, computers were so new that
employers found no natural fields from which to hire the needed programmers beyond mathematicians.
Thus, to find the best talent in such a nascent field, the U.S. government experimented by
hiring recent graduates in fields like linguistics, statistics, music, law and engineering, and training them from scratch.
My father loved the intellectual challenges of computers, but left
the field after a few years to pursue his passion for law and return to his hometown roots in the western Pennsylvania coal region. Of course, he enjoyed bragging for the rest of his life to all who
would listen about his early days programming computers.
advertisement
advertisement
Today, we are in the same position the U.S. government was in the 1950s: wondering where to find future AI talent. Here are my thoughts
on where to look:
Be open, and experiment. We will need to look to many fields to find the foundational training that will give our future AI talent the tools for success. Success
stories like Grace Hopper (who helped develop the Univac) and Dorothy Vaughan (of NASA, and “Hidden Figures” fame, who helped develop FORTRAN) show how open-minded government programs can
surface era-defining talent that would otherwise be held back.
Hire for curiosity and a history of hard work. The past few decades of Internet and digital-driven disruption have shown
the world the power of smart and social talents focused as teams on solving big, hard problems. However, as all of us who have built start-ups in the space know, nothing beats curiosity and hard work
to solve problems no one has solved before -- nothing.
Embrace the nonintuitive and counterintuitive. The field of behavioral economics as we know it today was built on research that
Danny Kahneman and Amos Tversky did to identify the best prospects for risk-taking roles in the Israeli Defense Forces. Their research showed that in many situations people made wrong decisions, and
they made those wrong decisions in very predictable ways. Basically, conventional wisdom was frequently, and predictably wrong in certain situations. Essentially, you can win in many areas by
following the data in ways that are exactly opposite to following your gut.
This issue is top of mind for me, as I just posted several entry-level Generative AI analyst roles for my company. I
am following the approaches I suggested above, certain that companies finding and developing superior AI talent at the entry level will have massive competitive advantages for years. How about
you?