AI is coming for the squishies first
I recently spoke with an airline engineering exec, who outlined how hard it was to introduce artificial intelligence to optimize their ground operations. They are able to improve turn-around times significantly by scheduling staff, automating checklists, and using sensor inputs more intelligently in preparing the aircraft for take-off.
What stands in their way? Very simply: unions and regulators. Both are formidable adversaries to change. In the case of regulators, they work from texts written from wrong assumptions. In the case of unions, they are fighting for their life and are out for blood. Just see the stevedores - those fights are happening across industries.
And so it got me thinking about where AI-driven automation will have the greatest immediate impact. My current thesis is that, not dissimilar to the internet, AI will choose the path of least resistance. And so it will come for the squishies first.
The internet was truly destructive first to the media industry. It dealt in text and video and it assumed a price of distribution. Well the marginal price of distribution went to quasi-zero with the internet and we're left with a lot of information (some of it quite bad). And the media industry is intact only in remnants (and even those are starting to be poorly edited, partisan, or plain trading on former glory).
Then under the guise of ecommerce, the internet came for retail. Retail benefited for a long-time from protected supply chains, cornered supply, complex operations, and limited store front/foot traffic. But it turned out it was fragmented and competitive and low-margin. And so the internet overwhelmed it with competition based on selection, convenience, and price. Then travel, which was similarly fragmented and opaque.
And only then did the internet come to other industries. Not sure if you noticed, but it is only now coming for the government and its regulators, the industries that are possibly the best protected. Government is putting up quite a fight in order not to be run by Elon Musk.
I'm a child of the web and I think history does rhyme. Hence the thesis that AI will hit hardest in those professions that deal primarily in transforming information (looking at you, wordcels) and that lack regulatory oversight/government control or union organization. This isn't because they are necessarily easier to automate, but because they operate within frameworks that offer the least resistance. Wherever structures are loose and flexibility is high, I expect AI to take over in a surprisingly rapid way.
Copywriters, customer support agents, data entry professionals, content creators, loan officers, risk assessors of all types, paralegals... you saw what happened to the stenographers of the 70s, the typists of the 90s, the stock traders of the 2000s…
Hence, psychotherapists of this world: time to unionize! (If you still can).
As always, if you are founding a business in the area identified above, Heartcore is a great first call.