>The economy grows at twice its normal rate. Wages for knowledge workers don’t rise,
>As a society, we’re far richer than we’ve ever been, but many fewer workers have jobs in knowledge work, and unemployment has risen beyond typical recessionary levels.
Man why do people hate AI so much. smh.
2Lore2Law on
Is this a research paper or a press release from a business with do-or-die need to attract more investment capital?
JesterOfAllTrades on
The modest and substantial scenarios are honestly nicer than the usual spiel that these people say – to be clear these are the same people who have unironically used terms like ‘permanent underclass’ and ‘plebs’. But those two scenarios look reasonable ISH. I expect it to be somewhere between the two maybe.
But let’s keep in mind that the people at anthropic are not economists, and just cos they’re good at building at language models doesn’t mean they are at building economic models. This is not NBER.
That said, these anthropic guys are really good at presentation huh. Seriously every time they make super pretty papers.
Windows_10-Chan on
It’s a shame I can’t link an anthropic piece to most normal people without being laughed out of the room because this actually does have a pretty good breakdown of how automation works even if you take their specific forecasting with a grain of salt.
> This paper presents a framework for assessing the economic consequences of AI between
2026 and 2030. In the model, AI automates a growing share of cognitive work, raising
productivity and displacing workers who must search for jobs in other occupations. The
model maps future paths of AI capabilities into implied paths for GDP, the labor share,
wages, labor reallocation, and unemployment. We illustrate the framework by considering
three scenarios: modest, substantial, and extreme.
> We also surveyed US adults about their expectations for AI.
The paper also had some input from economists like daron acemoglu, David autor etc.
> This paper seeks to place the wide range of views on AI’s economic effects inside a common
framework, as alternative settings of a small number of parameters that can be reasoned about
and, increasingly, measured. The framework is a standard task-based model with two groups
of occupations, an ideas channel, and a frictional labor market. The scenario explorer that
accompanies the paper lets anyone set the parameters and see the paths they imply for GDP,
growth, wages, the return to capital, the labor share, and unemployment through 2030.
WenJie_2 on
> For example, we did not include scenarios where humanity develops hyper-capable robots.
Important to understand for the people who think that the singularity will manifest some sort of matter replication beam into existence.
There is a lot of magical thinking going on in this paper.
The idea that wages for non knowledge labor will rise seems really wishful. If unemployment rises beyond recessionary levels who exactly is hiring these people? Capital doesnt need to spend more than its already spending.
It also admits that despite there being more capital the purchasing power of that capital will decrease which will drive inflation.
Theyre describing being pissed on but calling it a rainstorm with a smile on their faces.
13 Comments
>The economy grows at twice its normal rate. Wages for knowledge workers don’t rise,
>As a society, we’re far richer than we’ve ever been, but many fewer workers have jobs in knowledge work, and unemployment has risen beyond typical recessionary levels.
Man why do people hate AI so much. smh.
Is this a research paper or a press release from a business with do-or-die need to attract more investment capital?
The modest and substantial scenarios are honestly nicer than the usual spiel that these people say – to be clear these are the same people who have unironically used terms like ‘permanent underclass’ and ‘plebs’. But those two scenarios look reasonable ISH. I expect it to be somewhere between the two maybe.
But let’s keep in mind that the people at anthropic are not economists, and just cos they’re good at building at language models doesn’t mean they are at building economic models. This is not NBER.
That said, these anthropic guys are really good at presentation huh. Seriously every time they make super pretty papers.
It’s a shame I can’t link an anthropic piece to most normal people without being laughed out of the room because this actually does have a pretty good breakdown of how automation works even if you take their specific forecasting with a grain of salt.
This is your brain on TESCREAL
https://preview.redd.it/4qmruygk7ioh1.jpeg?width=1070&format=pjpg&auto=webp&s=37352dcc8020ba51907493f39634de501eebc7c8
Whew, for a second I was worried there’d be no meaningful and fulfilling work left for humans
It’s the interactive blog post that goes along with the following paper:
[Economic Scenarios for Transformative A](https://www-cdn.anthropic.com/files/4zrzovbb/website/cf58f84d46a4a76bf5a5b039ac695fba6b80041c.pdf)
> This paper presents a framework for assessing the economic consequences of AI between
2026 and 2030. In the model, AI automates a growing share of cognitive work, raising
productivity and displacing workers who must search for jobs in other occupations. The
model maps future paths of AI capabilities into implied paths for GDP, the labor share,
wages, labor reallocation, and unemployment. We illustrate the framework by considering
three scenarios: modest, substantial, and extreme.
> We also surveyed US adults about their expectations for AI.
The paper also had some input from economists like daron acemoglu, David autor etc.
> This paper seeks to place the wide range of views on AI’s economic effects inside a common
framework, as alternative settings of a small number of parameters that can be reasoned about
and, increasingly, measured. The framework is a standard task-based model with two groups
of occupations, an ideas channel, and a frictional labor market. The scenario explorer that
accompanies the paper lets anyone set the parameters and see the paths they imply for GDP,
growth, wages, the return to capital, the labor share, and unemployment through 2030.
> For example, we did not include scenarios where humanity develops hyper-capable robots.
Important to understand for the people who think that the singularity will manifest some sort of matter replication beam into existence.
https://preview.redd.it/6rvlzpkn8ioh1.png?width=1554&format=png&auto=webp&s=339ceecbde4f0a4bbba637ea2ca641b5b6aa3dd7
The white collar class is not going to adjust well to having to move into blue collar jobs
https://preview.redd.it/bshrl38p9ioh1.png?width=1216&format=png&auto=webp&s=05c7ab63adabc355a20e9697cae276c0357d0283
baumol goooood, job dislocation baaaad
I thought the extreme scenario is we all die
There is a lot of magical thinking going on in this paper.
The idea that wages for non knowledge labor will rise seems really wishful. If unemployment rises beyond recessionary levels who exactly is hiring these people? Capital doesnt need to spend more than its already spending.
It also admits that despite there being more capital the purchasing power of that capital will decrease which will drive inflation.
Theyre describing being pissed on but calling it a rainstorm with a smile on their faces.