4 ms·

> We agree that [...] impact on labor [...] are three of the main risks of AI I'm perpetually puzzled by that argument. It popped up the other day in a FT arti

> We agree that [...] impact on labor [...] are three of the main risks of AI I'm perpetually puzzled by that argument. It popped up the other day in a FT article about how jobs would be impacted, and in the grander scheme of things it's the general one whenever automation optimizes something that used to be done manually before. People equate jobs being removed to something necessarily bad, and I think it's wrong. The total output didn't change, we don't need meatbags to sit in a chair rather than having it done by silicium instead. People don't need to be occupied all day by a job. The problem is not with new technology lifting up something that was traditionally done by a human, but with how the system distributes sustenance (or in this case, doesn't distribute it), and how such transitions are supported by society. Tractors replaced people working in fields, and that allowed modern society. Now we're at the first tangible glimpse of actually having robots serve us, and we want to kill that since it's incompatible with our system of modern human exploitation, rather than changing the system itself, because god forbid this is gonna benefit everyone. I think that sucks.

biophysboy 4y ago
The worry is that AI like GPT will just exacerbate modern human exploitation, because the surplus will not be allocated. The same thing could be said about any disruptive invention, but AI is particularly centralized, expensive and technical
- foogazi 4y ago
  > The worry is that AI like GPT will just exacerbate modern human exploitation The worry should be human exploitation - exacerbated or not
  - dragonwriter 4y ago
    The worry is human exploitation; exacerbating it, however, exacerbates the concern.
  - charles_f 4y ago
    Yeah, my point exactly
falcons_shadow 4y ago
Although I agree with you, I can't help but imagine the disruption on an exponential scale vs the typical linear one that are often the center points of these examples. It took many years for the tractor to become more useful via improvements, more widespread, more easily produced and distributed which of course softened the landing for anybody that was going to be affected by it. So really its not the scale of the disruption that could cause issues but the scale and the timeline combined. I think that this is where the original article failed to demonstrate that this is how those companies will secure their power position.
beepbooptheory 4y ago
It's already been pretty clear that the real stakeholders here will let it all burn down before considering any fundamental changes.
mcguire 4y ago
This is true only if the "output" is of the same quality as the work done by humans. For tractors, that's unarguably true. For writing and artwork? (Unless someone wants to make the argument that automation has allowed and will allow humanity to do things that shouldn't have been done in the first place. Progress comes with consequences.)
fhd2 4y ago
I'd love to live in an AI powered utopia. Where everybody just does what they find valuable, regardless of whether there's someone willing to pay money for it, or whether an AI could do it more effectively or efficiently. By following our passions we could continue to make truly beneficial advancements, and all live enjoyable lifes. But I'm pessimistic about it. If we preserve the status quo where everyone needs to do something somebody else is willing to pay for - which I assume - we're setting up a dystopian tragedy of the commons where only those doing things nobody else wants to do - be it for moral concerns, dangers or tediousness - make a decent living. I don't see how we can change that trajectory with our current economic systems, and I don't see us changing those any time soon. But maybe it just has to get worse before it gets better. I _do_ think swift coordinated global efforts in lawmaking could contain this problem long enough for us to be able to change our systems slowly. It might feel idiotic to not be able to use technology available to us. But I also can't kick my neighbour's door down and take his TV, even though I want it and it'd be technically easy to do.
- ChatGTP 4y ago
  Most likely scenario, people will just do the same jobs, with AI assistance. Can you not already see this happening ?
  - fhd2 4y ago
    LLM assistance tooling is only just starting out from what I see. And since this means fewer people will be needed to do what's being done right now, some of them will need to find new things to do. Whether humanity needs these things, or whether it at least doesn't make the world worse, is a different question. It already happened with any kind of automation in the past. Do we need thousands of people spending their days building SEO pages, flooding the internet with low-value content in order to sell something? Probably not, but it's a way to game the system and make money.
    - ChatGTP 4y ago
      Unless this is the end of innovation, I'm sure there will be something else to do.
wcerfgba 4y ago
We need to focus on how to transition. The ideal you describe aligns witb socialism, Adam Smith's vision of reduced labour, and 'luxury space communism', but the outcomes are more effected by our economic structures than this disruption of any one technology like AI.

biophysboy 4y ago

The worry is that AI like GPT will just exacerbate modern human exploitation, because the surplus will not be allocated. The same thing could be said about any disruptive invention, but AI is particularly centralized, expensive and technical

foogazi 4y ago
> The worry is that AI like GPT will just exacerbate modern human exploitation The worry should be human exploitation - exacerbated or not
- dragonwriter 4y ago
  The worry is human exploitation; exacerbating it, however, exacerbates the concern.
- charles_f 4y ago
  Yeah, my point exactly

falcons_shadow 4y ago

Although I agree with you, I can't help but imagine the disruption on an exponential scale vs the typical linear one that are often the center points of these examples. It took many years for the tractor to become more useful via improvements, more widespread, more easily produced and distributed which of course softened the landing for anybody that was going to be affected by it. So really its not the scale of the disruption that could cause issues but the scale and the timeline combined. I think that this is where the original article failed to demonstrate that this is how those companies will secure their power position.

beepbooptheory 4y ago

It's already been pretty clear that the real stakeholders here will let it all burn down before considering any fundamental changes.

mcguire 4y ago

This is true only if the "output" is of the same quality as the work done by humans. For tractors, that's unarguably true. For writing and artwork? (Unless someone wants to make the argument that automation has allowed and will allow humanity to do things that shouldn't have been done in the first place. Progress comes with consequences.)

fhd2 4y ago

I'd love to live in an AI powered utopia. Where everybody just does what they find valuable, regardless of whether there's someone willing to pay money for it, or whether an AI could do it more effectively or efficiently. By following our passions we could continue to make truly beneficial advancements, and all live enjoyable lifes. But I'm pessimistic about it. If we preserve the status quo where everyone needs to do something somebody else is willing to pay for - which I assume - we're setting up a dystopian tragedy of the commons where only those doing things nobody else wants to do - be it for moral concerns, dangers or tediousness - make a decent living. I don't see how we can change that trajectory with our current economic systems, and I don't see us changing those any time soon. But maybe it just has to get worse before it gets better. I _do_ think swift coordinated global efforts in lawmaking could contain this problem long enough for us to be able to change our systems slowly. It might feel idiotic to not be able to use technology available to us. But I also can't kick my neighbour's door down and take his TV, even though I want it and it'd be technically easy to do.

ChatGTP 4y ago
Most likely scenario, people will just do the same jobs, with AI assistance. Can you not already see this happening ?
- fhd2 4y ago
  LLM assistance tooling is only just starting out from what I see. And since this means fewer people will be needed to do what's being done right now, some of them will need to find new things to do. Whether humanity needs these things, or whether it at least doesn't make the world worse, is a different question. It already happened with any kind of automation in the past. Do we need thousands of people spending their days building SEO pages, flooding the internet with low-value content in order to sell something? Probably not, but it's a way to game the system and make money.
  - ChatGTP 4y ago
    Unless this is the end of innovation, I'm sure there will be something else to do.

wcerfgba 4y ago

We need to focus on how to transition. The ideal you describe aligns witb socialism, Adam Smith's vision of reduced labour, and 'luxury space communism', but the outcomes are more effected by our economic structures than this disruption of any one technology like AI.