Public predictions about the impact of genAI on workers and the economy are dramatic – but are they right?
Depending on our assumptions on the effectiveness of AI, economic models often suggest outcomes differing from what we see in the public sphere. For example, employment may rise in AI impacted jobs if the level expertise needed for the non-automatable parts of the job is not high.
martingalex2 · 4h ago
My prior is that if Daron Acemoglu thinks the impact of Gen AI on productivity is muted, then it's probably a good unbiased gauge.
Depending on our assumptions on the effectiveness of AI, economic models often suggest outcomes differing from what we see in the public sphere. For example, employment may rise in AI impacted jobs if the level expertise needed for the non-automatable parts of the job is not high.