pull down to refresh
174 sats \ 7 replies \ @Undisciplined 12 May \ on: Many people now think that AI is a mortal threat to jobs - are they right? econ
I don't know the specifics of tech jobs, but normal attrition can often be in the single percents. So, even if AI only eliminated a few tenths of a percent of positions, that would make the job market much tighter.
For the sake of argument let's say 2% of workers in tech are normally leaving their job during the year and there are normally enough open positions for them to take. If 0.1% of positions are cut, then there will now be 2.1% of unemployed chasing 1.9% of positions.
Not only is that a 5% increase in unemployed tech workers, but now almost 10% of unemployed tech workers have no available position to take.
I imagine that 2% will think a lot harder about leaving their job!
You're an econ guy...what do you think about all this?
Right now it feels like AI is doing a "shock and awe" type campaign. I use it all the time, and it can just do SO MUCH. It's a little unnerving.
I was doing some work recently in Gumroad.com, and their tech support AI was absolutely outstanding. It was extremely helpful, with an easy link at the bottom to "contact a human" if you needed to (which I did once). I started my tech career in customer support - wonder if that opportunity will still exist in a few years.
(BTW Gumroad open sourced it - it's Helper, on github - https://github.com/antiwork)
reply
I think this movement of few jobs in this area has many factors and AI can be the smallest of them. Some colleagues I have in the area complain that in recent years it has had an explodation of poorly trained professionals and accepted any value to do inferior quality work. With the AI becoming more accurate it may be that keeping people insufficient and cheap is no longer advantageous. And in my opinion, based on some data, I believe that there is a "Shock and Awe", with more people using and forcing the Age soon that it is not as efficient as an experienced professional.
reply
I'm reasonably confident in my short-term and long-term projections, but I have no confidence at all in any particular medium-term predictions.
Short-term: lots of change, which is very painful and messy
Long-term: enormously enhanced productivity will greatly improve standards of living
reply
I think it's going to take longer for young people to launch out of subsistence living, unless they're willing to get out of the knowledge-producing industry and go back to physical labor.
I just don't see what value a young person with no domain-expertise brings to a knowledge industry anymore. I'd venture to guess that AI will outperform 99% of undergraduates on knowledge tasks in their first 1 or 2 years on the job. So why bother?
So it'll be harder for young folks to get a foot in the door. They may have to work longer at subsistence wages to build up a portfolio to show that they can do self-directed work, including orchestrating AI outputs to align with a business use case, before they'd get hired by a company. Or they might have to go the entrepreneurial route, which is probably a good scenario if more people were to do that.
reply
In times of rapid change, more entrepreneurship is really important. We need people hunting out the new profit opportunities.
I think we'll see another similar change, as we saw with electronic computing. Being good at arithmetic hasn't been a marketable skill for a long time. However, being a creative technical thinker became even more valuable.
Now, my guess is that being a walking encyclopedia will lose it's value, but productively using knowledge will be even more valuable.
Honestly, with how fat and sick our society has gotten, a tilt towards physical professions is not the worst thing in the world.
reply
It's interesting to remember that something like "being good at arithmetic" WAS actually a marketable skill years ago.
Now it isn't, of course.
It would be interesting to see how the job market changed with the advent of personal computing - when computers got cheap enough that most people and certainly all businesses could afford one. Were there people back then, very worried about how computers would replace them?
I imagine there were, but I don't really remember that. I think there was a big push to retrain laid-off factory workers as IT people, but I think that didn't work at all.
reply
I'm not old enough to remember it first hand, but I have heard people saying the conversation and concerns were similar to today with AI.