"Job creation" is a weird, political talking point.
Like Beckert (#1562148, #1562148)) being obsessed and snarky about "the economy" as a feeling, moral agent (it isn't; nobody sane thinks it is: #1571119), "creating jobs" isn't a good, admirable, desirable thing.
To paraphrase Milton Friedman's quip, we can create lots of jobs by replacing all the construction machines with shovels, and then replace the shovels with toothpicks. But of course, the intelligentsia doesn't understand what "a job" is (#1457057) or value or, frankly, anything to do with economics. #1568327, #1573563
Here's our friend, the pseudo- or semi-economist Soumaya Keynes (yes, relation), asking...
"Where do new jobs come from?"
Joseph Schumpeter: "waves of technology destroy old work, making way for the newer sort."
video stores, punch-card operators, hoooorrrses! #1486740
And maybe younger readers wonder whether they will live long enough for the only jobs left to be AI engineers.
Adam Smith: specialization.
Jobs are essentially bundles of tasks. Big markets allow people to unbundle those tasks, divide them up, and produce much more.
WHO WINS?!
The answer is complicated by the two stories being hard to separate. Technology could enable specialisation, as when the computing revolution made it easier to co-ordinate between faraway factories. Advances in knowledge could push specialisation too, as it becomes unreasonable to expect one person to be an expert in everything.
Economic historians have tried to see through the fog, analysing past labour market revolutions to get a sense of which dynamics mattered.
- court record study favored the Smithian account: clock makers from 1550 -> tens of specialized jobs
- recent U.S. study used LLMs to show whether jobs were tech, "Of the new employment introduced between 2011 and 2023, they calculated that the tech-related share was roughly a third. Substantial, though not a majority."
- new Swedish study, 1880-2019, used AI to designate occupations on Smithian vs Schumpeterian scale
"With the caveat that the lines between technology and specialisation can be blurry, the researchers get some interesting results"
Over time, jobs with both high Smithian and Schumpeterian scores have become more important as a share of the total. But the jobs associated with the division of labour have contributed the bulk of the change. Between 1990 and 2019, strongly Smithian jobs accounted for 60 per cent of employment, compared to just a third for strongly Schumpeterian ones.
soooo... Smith wins? "Smithian jobs seem much more likely to last than their highly Schumpeterian counterparts."
Although we might fret now about AI sweeping away jobs in a wave of creative destruction, thinking about how it will affect specialisation could be more instructive. The Sweden study’s authors calculated that around 70 per cent of employment between 1990 and 2019 was in occupations that existed at the end of the 19th century. What matters now is which tasks stay bundled within jobs and which move elsewhere.
As Luis Garicano, an author of the book Messy Jobs, put it to me, some bundles of tasks are only weakly held together, and are therefore ripe for AI disruption.
He also noted that past trends could reverse if AI makes specialist subjects so accessible that it becomes easier to bring them in-house, rather than suffer the cost of co-ordinating with other experts. (Think of an economics columnist deciding to build her own website rather than pay a freelancer.)
I wonder: do you see the wordsmith profession moving more towards specialization too? For IT this has always been the case, also with every hype cycle, including the current AI one.
Background:
Some decade and a bit ago I used to work with a lady that was by profession a speech writer, but she didn't like DC and went to work commercially as a proposal writer instead, for a tech company of course. The most amazing magic she had weren't the RFP responses though - it was resumes. I am 200% sure that she will write a better resume than an LLM (after all the "slop resume" accusations in the media the past 18 months, I checked what all that was about and LLMs are imho horrible resume writers, just not as horrible as the normie writing their resume.) So there's a specialization path for this particular person and skillset, even if it means let the LLM write the bullet points and then she throws away the slop and makes it a real thing (though from experience with this writer in particular, that wasn't the operating process either, because the input was an interview, not a list of bullet points)
Anyway... I wonder how that is felt for you, and whether you see that as a Smithian development?
Fuck, good questions. Will think.
Quick on-the-fly response:
It's definitely also speeding up my work. Instead of pondering with a draft on an article for days or week, I'll just shove everything I have, unstructured, into an LLM and tell it to write a suitable article in my voice — specifically including x, y, z, and this direct quote from NN. Usually produces decent drafts I can work with
I like how you differentiate the subjective here. And maybe this all is subjective in the end. Just because the top 10 in one's field got a raise and are doing better than ever, doesn't mean that the bottom 10% won't lose their jobs (or, as Jack showed us, the "bottom 60%".) You don't need AI for the bottom n-percent to lose their jobs though, that's perfectly capable of happening without AI too, and without automation, and without information technology[1]
I was talking to a young developer the other day, they started their career in 2021. Now they have Claude Code. And they complained: I can do anything now, but when you shoot on my code, I have to ask Claude what you mean. And that's where it becomes harder for me to accept. Because you had this dev that was good within their limits, now pushed the limits to "produce" things that were out of reach before. But because those things not only were, but technically still are out of reach, they can only manage it with a dependency on a subscription to a lab's model. Take that away, and that whole ranking into percentiles will flip. That's also my fear. Why I struggle at times to go through all the shit I need to review that I could ask an LLM to review for me, but the moment I truly depend on it other than for things I know I master without it, I start losing instead of winning?
And perhaps, that's the ultimate challenge, to keep the competitive advantage versus the LLM where we can keep on saying
I once was hired by a 150+y/o company that every 20 years went through a crisis and would fire half the personnel. Can't say it didn't work for them if they still had money to hire me after 150 years. But it illustrates that you don't need AI to need reorgs. ↩