4 January 2025 · 1 min
The more likely outcome of the rise of generative artificial intelligence is higher value for the best experts… where ‘expert’ means ‘someone with experience solving real problems’.
“While one may worry that AI will simply render expertise redundant and experts superfluous, history and economic logic suggest otherwise. AI is a tool, like a calculator or a chainsaw, and tools generally aren’t substitutes for expertise but rather levers for its application.
By shortening the distance from intention to result, tools enable workers with proper training and judgment to accomplish tasks that were previously time-consuming, failure-prone or infeasible. Conversely, tools are useless at best — and hazardous at worst — to those lacking relevant training and experience. A pneumatic nail gun is an indispensable time-saver for a roofer and a looming impalement hazard for a home hobbyist.
For workers with foundational training and experience, AI can help to leverage expertise so they can do higher-value work. AI will certainly also automate existing work, rendering certain existing areas of expertise irrelevant. It will further instantiate new human capabilities, new goods and services that create demand for expertise we have yet to foresee.” (Autor, 2024)
Daniel Lemire, "Artificial Intelligence as the Expert’s Lever: Elevating Human Expertise in the Age of AI," in Daniel Lemire's blog, January 4, 2025, https://lemire.me/blog/2025/01/04/artificial-intelligence-as-the-experts-lever-elevating-human-expertise-in-the-age-of-ai/.
[BibTeX]
Only if the people who need and use the output care about the difference between good, bad, and mediocre. I know a tech writer who got fired and replaced by AI, and the AI is producing worse text but the people in charge of the text can’t tell or don’t care. Instead they need to check off different checkboxes like “every option parameter to every command is documented” instead of “the docs make sense and are human-readable, and the text flows well, and it’s all true and correct”.
I worry about the same effect for computer code. I like to think I’m very good at writing it, and with an LLM I’m even better! But I’m way more expensive than a bad coder who uses AI. So if you have no concept of code quality, then I’m definitely too expensive.
Or, put differently, a friend recently noted that AI is dangerous because the way it tends to seem correct tends to be the same as the way bureaucracies perceive correctness. So without a person with discernment and expertise in the evaluation seat, AI slop will taste just fine to an organization. We need the people with the power of the purse to be able to tell good from bad from middling, and right now I’m not sure they can.
Did you read the title and go straight to the comment section?
The quoted paper is: David Autor, “Applying AI to Rebuild Middle Class Jobs” (2024) http://www.nber.org/papers/w32140