NOTE: Most of the data originated in the US, hence the flair.
In summary, the article notes that unemployment for vulnerable sectors (think recent college graduates) has continued rising in accordance to pre-pandemic trends. There has been no “jump” due to ChatGPT.
So the good news: AI seems to have not replaced jobs yet. The bad news: grad unemployment has been rising since 2009.
ucasthrowaway4827429 on
They didn’t mention software engineering specifically in the article which is both a place where unemployment is present recently, and also where AI has some of the most impact and use.
IMO if AI unemployment is a thing it will be a discontinuous transition where AIs are suddenly able to beat the hurdles that make humans better than autonomous AI in those jobs and then things will get crazy, but based on the impressiveness of recent models that looks to not be happening for a while.
throwawaygoawaynz on
Former AI specialist at MSFT, AWS here. Now head of AI at a large MNC.
It was always overblown, but to be fair people are thinking ahead to when we have AGI. For now, it makes us more productive, but it’s not going to replace people wholesale. Similar to the world digitising in the 90s with MS office, some jobs went, but the vast majority stayed but people became more productive. AFAIK technology started having a measurable impact on GDP around the early 2000s due to this.
I see some disruption coming for data entry jobs if compute agents pan out as expected. I already see some companies starting to do mass layoffs in these areas.
Also if you’ve seen the latest Veo3 videos from Google then yeah, I’d be worried if I was in the TV space. But like a lot of stuff out of Google let’s wait and see what the execution is like, as the marketing is usually ahead of the reality (same with OpenAI Sora).
3 Comments
NOTE: Most of the data originated in the US, hence the flair.
In summary, the article notes that unemployment for vulnerable sectors (think recent college graduates) has continued rising in accordance to pre-pandemic trends. There has been no “jump” due to ChatGPT.
So the good news: AI seems to have not replaced jobs yet. The bad news: grad unemployment has been rising since 2009.
They didn’t mention software engineering specifically in the article which is both a place where unemployment is present recently, and also where AI has some of the most impact and use.
IMO if AI unemployment is a thing it will be a discontinuous transition where AIs are suddenly able to beat the hurdles that make humans better than autonomous AI in those jobs and then things will get crazy, but based on the impressiveness of recent models that looks to not be happening for a while.
Former AI specialist at MSFT, AWS here. Now head of AI at a large MNC.
It was always overblown, but to be fair people are thinking ahead to when we have AGI. For now, it makes us more productive, but it’s not going to replace people wholesale. Similar to the world digitising in the 90s with MS office, some jobs went, but the vast majority stayed but people became more productive. AFAIK technology started having a measurable impact on GDP around the early 2000s due to this.
I see some disruption coming for data entry jobs if compute agents pan out as expected. I already see some companies starting to do mass layoffs in these areas.
Also if you’ve seen the latest Veo3 videos from Google then yeah, I’d be worried if I was in the TV space. But like a lot of stuff out of Google let’s wait and see what the execution is like, as the marketing is usually ahead of the reality (same with OpenAI Sora).