Fears of job loss due to AI are probably overblown

Without fail pundits are trotting out their proclamations about how AI will make {fill in the blank} job obsolete. This year it’s OpenAI. Last year it was DALL·E that was supposed to make artists obsolete. Even as far back as a decade ago, AI was predicted to make lawyers obsolete, among other professions:

Armies of Expensive Lawyers, Replaced by Cheaper Software (2011)
Will Robots Steal Your Job?
Can an Algorithm Write a Better News Story Than a Human Reporter? (2012)
Rise of robot reporters: when software writes the news (2014)
Stephen Hawking warns artificial intelligence could end mankind (2014)

Yet again, those artists, journalists, and other content creators and careers persist. AI just refuses to destroy those jobs. The reality is, much of these recent AI tools are just content creation tools, whether it’s pictures or text. But content has always been cheap. Long before DALL·E and Open Ai, it cost next to nothing to hire foreign workers to rewrite text or produce graphics. Affordable content creation has never been a pressing problem, at least not one that necessitates AI.

At best, AI and other automation only replaces a part of a job, not the job in its entirety: being a lawyer is a lot more than just scanning text. Being a doctor is a lot more than matching symptoms with conditions. A journalist doesn’t just write the story, but must also gather information, investigate sources, and interview people, which cannot be automated as easily. Same for self-checkout machines, which have not displaced cashiers, but only led to some cashiers attending to the machines instead.

Distribution and promotion is way harder, much more expensive, and not nearly as amenable to AI, if at all. Google is worth almost a trillion dollars because companies spend so much money trying to get their content seen, not the actual production of said content. By comparison, Shutterstock, one of the world’s largest paid repositories of images, is worth just over $1 billion. The world is already overflowing in content. Making content visible and stand out from the competition is where the money is at, and is a much harder problem than producing it.

Even the highest paid of content producers, like authors, journalists, and graphics designers, are paid much less compared to people involved in strategy, execution, and design, such as coaches of professional sports teams, developers, consultants, and executives. A decade or so ago I recall reading that the highest paid journalist at the time was the WSJ tech columnist Walter Mossberg, who earned $500k/year, a large sum for sure, but this just scratches the surface of compensation compared to other fields like law and tech.

Same for quality. Sure, AI can compose a passable essay based on some prompts, but can it ever produce something of Pulitzer caliber? Unlikely. Can it replicate the nuance that separates top writers from average ones? Also unlikely. There will always be a need for the top 5% or so of creators, whose work cannot be as easily automated.