AI Job Losses--or Not?
Monday, April 28, 2025
AI Job Losses—or Not?
You tell me. I’m not sure.
The AI in question being, of course, merely a language-generating bot of the large-language-model variety.
Back to the point: Does a job count as lost if it might not have existed otherwise? And is terrestrial radio a dying medium anyway, so who cares? Is there a future for flesh-and-blood radio announcers—the disk jockeys of yore—who blab irrelevant nonsense between tracks? Is there any end to this streak of interrogatives?
An intrepid journalist in Australia grew skeptical that an Australian radio host wasn’t an actual person, but instead an LLM with an LLM-built backstory and image that made “her” an Asian woman for the sake of inclusion and diversity bonus points. The story was recounted at StartupDaily.net (for which, kudos to Tyler Cowen at MarginalRevolution.com):
ARN declares it’s “an inclusive workplace embracing diversity in all its forms”, and that appears to include Large Language Models.
The issue was first flagged last week when journalist Stephanie Coombes had a tip off and did some sleuthing on her site, The Carpet, unable to find any other digital presence or backstory for Thy.
“In audio you can’t make an exceptional cup of tea in the office kitchenette without a press release going to trade publications like MediaWeek and Radio Today,” Coombes wrote.
“It seems very odd that CADA hired a new ethnically-diverse woman to their youth station and then just forgot to tell anyone.”
Ha!
Still: with all the chatter about AI taking jobs from carbon-based mortals, how real is this example? These AIs can generate endless streams of coherent-sounding language, although the story in this case involves a human author writing copy for the AI voice to convert to speech. But who would ever know? Does anyone hang on the words of hip-hop DJs?
"Does a job count as lost if it might not have existed otherwise?"
Kevin Williamson sometimes mentions that you can't evaluate the effect of a change simply by considering the situation before the change and the situation after the change. To understand the full impact, you have to compare the situation after the change with *what the situation would have been absent the change.*
Being as how that didn't happen, it's hard to come up with statistics that are fully persuasive, which is why we still have minimum wage laws.
"Does anyone hang on the words of hip-hop DJs?"
My experience with radio-listening youth (mainly Vlad) is that they turn the sound down or change the station if someone is talking instead of playing a song. There's about a 3-second limit for statements of, "That was (song) by (performer) from (year)."