• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    AI

    Economist Repeats Claim of No AI Job Impact

    GovAI researcher states he and other economists see no AI effect in job data.

    SM
    1 Source, 35d ago, first seen 35d ago

    TLDR

    Sam Manning, a Senior Research Fellow at GovAI who previously worked at OpenAI, wrote on X that he and other economists have claimed AI shows no discernible impact on jobs in aggregate statistics. He said he made the same point in the latest International AI Safety Report and in recent talks. Manning added that the observation holds for most measures because they reveal no strong relationship between AI and employment numbers. The post records his public restatement of this position without new data or analysis.

    Combined views

    6.2K

    1 Source, first seen 35d ago

    Combined views

    6.2K

    1 Source, first seen 35d ago

    63 likes
    63 likes
    9 comments
    37 saves
    12 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    9 comments
    37 saves
    12 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet

    1 Source

    @sj_manningI've heard a number of economists claim that AI has had no discernible impact on jobs in aggregate statistics. I myself have made this claim, for example in the latest International AI Safety Report and in a few recent talks. This is true in one sense because, for most measures, we don't see any strong relationship between AI usage or exposure and aggregate headcount across jobs. You can see some of the most recent data on this in work by the Yale Budget Lab, for example. I think this is probably misleading though and causes people to underappreciate the probable impact of AI progress on the US labor market to date. If we think about the labor market today relative to a counterfactual of no AI progress post GPT-2 or GPT-3, my best guess is that AI has likely had positive effects on wages and potentially on reducing unemployment as well (though I'm less confident in the latter). Some reasons why AI progress has probably had a meaningful impact on jobs: - An entirely new industry has popped up around AI. There are the big model developers, but there are also a ton of new companies that do AI integration services for businesses, AI security services, legacy companies hiring in AI-focused divisions, and a bunch of downstream developers starting or extending businesses that only exist because of what LLMs can currently do. I know multiple people in my random non-tech-hub suburb who do sales for "AI companies" I've never heard of and earn six-figure salaries. - All the investment in chips, data centers and energy has driven demand along those supply chains, creating new jobs or pushing up wages in construction, manufacturing, logistics, etc. - The effects above + stock market gains largely driven by AI have made a good share of the US population richer, leaving them with more disposable income to spend, sustaining employment or pushing up wages in service sectors, health care, etc. If we hadn't had AI progress post GPT-2 or 3, there may have counterfactually been some other place where a bunch of investment would have gone that would have had some similar effects, but it seems unlikely to me that the effects would have been of a similar magnitude. Wonder if this seems right to people?

    1 Source

    @sj_manningI've heard a number of economists claim that AI has had no discernible impact on jobs in aggregate statistics. I myself have made this claim, for example in the latest International AI Safety Report and in a few recent talks. This is true in one sense because, for most measures, we don't see any strong relationship between AI usage or exposure and aggregate headcount across jobs. You can see some of the most recent data on this in work by the Yale Budget Lab, for example. I think this is probably misleading though and causes people to underappreciate the probable impact of AI progress on the US labor market to date. If we think about the labor market today relative to a counterfactual of no AI progress post GPT-2 or GPT-3, my best guess is that AI has likely had positive effects on wages and potentially on reducing unemployment as well (though I'm less confident in the latter). Some reasons why AI progress has probably had a meaningful impact on jobs: - An entirely new industry has popped up around AI. There are the big model developers, but there are also a ton of new companies that do AI integration services for businesses, AI security services, legacy companies hiring in AI-focused divisions, and a bunch of downstream developers starting or extending businesses that only exist because of what LLMs can currently do. I know multiple people in my random non-tech-hub suburb who do sales for "AI companies" I've never heard of and earn six-figure salaries. - All the investment in chips, data centers and energy has driven demand along those supply chains, creating new jobs or pushing up wages in construction, manufacturing, logistics, etc. - The effects above + stock market gains largely driven by AI have made a good share of the US population richer, leaving them with more disposable income to spend, sustaining employment or pushing up wages in service sectors, health care, etc. If we hadn't had AI progress post GPT-2 or 3, there may have counterfactually been some other place where a bunch of investment would have gone that would have had some similar effects, but it seems unlikely to me that the effects would have been of a similar magnitude. Wonder if this seems right to people?