https://www.vox.com/a/new-economy-future/robert-gordon-interview
For the present day:
https://fortune.com/2026/05/27/ai-productivity-internet-boom-solow-paradox/
Published before the present "AI boom" or "AI bubble".
What do you think about Robert Gordon? Should you have a flair for him? Regardless, I chose Daron Acemoglu because he is a techno-pessimist on automation.
I think job destruction through advanced artificial intelligence (which I don't think we're even close to achieving) is a necessary but not sufficient condition for more prosperity. That technology will increase total factor productivity.
So, in a way, to increase productivity, jobs have to be destroyed. I would imagine such a world to have some positives not yet manifested in the real world, such as being able to get into a car that can take your anywhere and not drive it. That would require an AI with almost human-level capability that can function in complex environments.
I said this in 2023:
The forecast of the Metaculus pros seems to corroborate my pessimistic framework. There is no Solow's paradox concerning the anemic growth of the 2010s and 2020s; it is just that smartphones and generative AI are not capable of generating torrid economic and productivity growth.
I wouldn't be that against neoliberalism or whatever if John Maynard Keynes was correct about people only working 15 hours a week in 2030.
Posted by AdmiralKurita
2 Comments
[https://www.vox.com/a/new-economy-future/robert-gordon-interview](https://www.vox.com/a/new-economy-future/robert-gordon-interview)
For the present day:
[https://fortune.com/2026/05/27/ai-productivity-internet-boom-solow-paradox/](https://fortune.com/2026/05/27/ai-productivity-internet-boom-solow-paradox/)
Published before the present “AI boom” or “AI bubble”.
What do you think about Robert Gordon? Should you have a flair for him? Regardless, I chose Daron Acemoglu because he is a techno-pessimist on automation.
I think job destruction through advanced artificial intelligence (which I don’t think we’re even close to achieving) is a necessary but not sufficient condition for more prosperity. That technology will increase total factor productivity.
So, in a way, to increase productivity, jobs have to be destroyed. I would imagine such a world to have some positives not yet manifested in the real world, such as being able to get into a car that can take your anywhere and not drive it. That would require an AI with almost human-level capability that can function in complex environments.
I wouldn’t be that against neoliberalism or whatever if John Maynard Keynes was correct about people only working 15 hours a week in 2030.
rate this hot take: the death of corporate mentorship is unironically probably what stopped the AI hype train, ie if corporations screened more on being able to teach those below you when it came to promotions it would have cultivated the skillset such that senior engineer prompting claude for agents could have written their skills and behaviors better, and they would be better an translating ambiguous orders from their superiors into more narrow actionable goals for their agents. Thus had corporate mentorship been preserved, AI would have been way more hyped and transformative. As it stands the organizations need to reorganize to take advantage of AI