I am a fan of generative AI because I think it can enhance productivity significantly. And we desperately need to improve productivity to boost growth in an ageing society. Yet, when I read a recently published short note from Daron Acemoglu on the likely growth impact of AI
His paper does 'exist' (must for now) in a world without politics:
1
'the labour share of GDP to decline even further is going to extend further as bargaining power for workers'
What was/is portrayed in a lot of media as a dangerous and vulgar turn (are they marching already?) to the radical/extreme/far/brown right - which didn't play out to the full in a lot of EU nations - is in no small part a move by voters against:
Wage competion (immigration); high energy prices (impact progressive hobbies); the inequality promoting effects of QE and a low interest environment (bankers saving bankers); a lack of affordable housing (gov's aren't building/enabling building); and sheer activist lunacy (the NL's nitrogen laws, national & European judges taking political positions etc).
Though the pay off for an altered gov direction has yet to be seen: Dutch gov officials from the various ministries have been holding meetings since last fall as the center and the right won the elections.
They seem to be looking for ways to oppose if not obstruct new policies - which still have to be translated into programmes. For years soaked in progressive righteousness and self selecting through hiring practices they are now 'very worried' about the future of 'climate policies' and the (EU) legality issues of new gov plans.
The fact that the new gov also wants to cut (probably spelled as 'gut' inside the 4th power) the number of civil workers is obviously also a frightening prospect - job cuts much better fit the commercial sector of course. As one official said 'how can they expect me to execute their policies if they don't care about me?'
2
Bursztyn et al. (2023) run an experiment to evaluate the welfare effects of social media for college students
As the societal impact of social media becomes so evident, thinking about counter measures grows as well. I've seen group discussions by young teen girls - 12/14 yo - who mostly supported (some form of) bans of smartphone (use), social media use etc. and seemed quite unhappy with how especially girls use social media (when boys/men fight they confront each other and typically its over quickly, while girls/women try to destroy each other a bit more vigorously and with impressive stamina).
Here in the NL schools are introducing 'anti smartphone' policies where they have to be put in lockers as schoolday begins. This has a remarkable effect: since a few months you can see children actually running around (again) during the breaks instead of being individually absorbed by their phone (Dutch teens / young adults are the least physically active of all Europeans, viva el ebike!).
Still i think the internet has structurally changed puberty, especially for girls. But i think Jonathan Haidt has more to say about that than Acemoglu since the matter is entirely political.
The result is quite astonishing, but I would disagree with some of the referenced papers on which it is based. As a software and data engineer, I really would estimate the productivity gains to be much higher than the ~50% stated in the study by Peng et al. - rather between something like 100% and 1000% (yes, that is 10 times). This could be also due to the more powerful models available today and increased experience in knowing which tasks to apply them to, and how. (The study's task was also unusual since no software engineer worth their salt would (a) write a web server from scratch, and (b) do so in JavaScript. ;P)
I think the 50% productivity gain is already at the high end. Be aware that the internal combustion engine and the industrial revolution only brought with them productivity gains of about 20% to 50% across the entire economy. There clearly are areas where there is massive productivity gains, such as software engineering or even music (see: https://www.youtube.com/watch?v=zbo6SdyWGns) but in most other sectors the gains are far more modest. In my own field of finance, I am using AI, but so far the productivity gains are only in the range of about 5% to 10% because generative AI is pretty bad at data analysis - even the latest generation. I know this is going to get better, but I doubt that we will see productivity gains of more than 20% in finance.
Plus, don't forget that the vast majority of our economy is based on human to human interaction and physical actions by humans. Think of construction, retail shops, any sales activity. AI can make these processes more efficient, but there are limited efficiency gains in social care, construction, food preparation, teaching, etc.
Finally, you also have to be aware that if we make people redundant due to increased productivity gains from AI, the unemployed will be a drag on national productivity because they depend on the social welfare system. Thus, what you gain in productivity in one part of the economy, you lose in another.
His paper does 'exist' (must for now) in a world without politics:
1
'the labour share of GDP to decline even further is going to extend further as bargaining power for workers'
What was/is portrayed in a lot of media as a dangerous and vulgar turn (are they marching already?) to the radical/extreme/far/brown right - which didn't play out to the full in a lot of EU nations - is in no small part a move by voters against:
Wage competion (immigration); high energy prices (impact progressive hobbies); the inequality promoting effects of QE and a low interest environment (bankers saving bankers); a lack of affordable housing (gov's aren't building/enabling building); and sheer activist lunacy (the NL's nitrogen laws, national & European judges taking political positions etc).
Though the pay off for an altered gov direction has yet to be seen: Dutch gov officials from the various ministries have been holding meetings since last fall as the center and the right won the elections.
They seem to be looking for ways to oppose if not obstruct new policies - which still have to be translated into programmes. For years soaked in progressive righteousness and self selecting through hiring practices they are now 'very worried' about the future of 'climate policies' and the (EU) legality issues of new gov plans.
The fact that the new gov also wants to cut (probably spelled as 'gut' inside the 4th power) the number of civil workers is obviously also a frightening prospect - job cuts much better fit the commercial sector of course. As one official said 'how can they expect me to execute their policies if they don't care about me?'
2
Bursztyn et al. (2023) run an experiment to evaluate the welfare effects of social media for college students
As the societal impact of social media becomes so evident, thinking about counter measures grows as well. I've seen group discussions by young teen girls - 12/14 yo - who mostly supported (some form of) bans of smartphone (use), social media use etc. and seemed quite unhappy with how especially girls use social media (when boys/men fight they confront each other and typically its over quickly, while girls/women try to destroy each other a bit more vigorously and with impressive stamina).
Here in the NL schools are introducing 'anti smartphone' policies where they have to be put in lockers as schoolday begins. This has a remarkable effect: since a few months you can see children actually running around (again) during the breaks instead of being individually absorbed by their phone (Dutch teens / young adults are the least physically active of all Europeans, viva el ebike!).
Still i think the internet has structurally changed puberty, especially for girls. But i think Jonathan Haidt has more to say about that than Acemoglu since the matter is entirely political.
The result is quite astonishing, but I would disagree with some of the referenced papers on which it is based. As a software and data engineer, I really would estimate the productivity gains to be much higher than the ~50% stated in the study by Peng et al. - rather between something like 100% and 1000% (yes, that is 10 times). This could be also due to the more powerful models available today and increased experience in knowing which tasks to apply them to, and how. (The study's task was also unusual since no software engineer worth their salt would (a) write a web server from scratch, and (b) do so in JavaScript. ;P)
I think the 50% productivity gain is already at the high end. Be aware that the internal combustion engine and the industrial revolution only brought with them productivity gains of about 20% to 50% across the entire economy. There clearly are areas where there is massive productivity gains, such as software engineering or even music (see: https://www.youtube.com/watch?v=zbo6SdyWGns) but in most other sectors the gains are far more modest. In my own field of finance, I am using AI, but so far the productivity gains are only in the range of about 5% to 10% because generative AI is pretty bad at data analysis - even the latest generation. I know this is going to get better, but I doubt that we will see productivity gains of more than 20% in finance.
Plus, don't forget that the vast majority of our economy is based on human to human interaction and physical actions by humans. Think of construction, retail shops, any sales activity. AI can make these processes more efficient, but there are limited efficiency gains in social care, construction, food preparation, teaching, etc.
Finally, you also have to be aware that if we make people redundant due to increased productivity gains from AI, the unemployed will be a drag on national productivity because they depend on the social welfare system. Thus, what you gain in productivity in one part of the economy, you lose in another.
Indeed, the overall picture for the economy might be very different (and very difficult to model).