Imminent disruption to the social contract
Discussion
I keep coming back to the idea that AI and automation is a social contract shift as well as a technological shift.
For decades the deal has been:
Work --> Wage --> Spend--> Stable (ish) Economy
But if office roles (and later physical roles) are hollowed out faster than new forms of work appear, that link inevitably weakens. Companies save on labour, but at the same time shrink their own customer base if fewer people have disposable income. What happens when an economy automates labour faster than it replaces purchasing power?
I’m curious as to where people think this realistically heads over the next 5–10 years. Are we headed for a future where the middle class is shrunk massively and we are all scrabbling for lower paid menial work while also being heavily reliant on the state and universal income? surely the good jobs that remain will also have a such a huge number of applicants that salary decreases for those as well.
It is all looking rather dystopian and bleak to me.
For decades the deal has been:
Work --> Wage --> Spend--> Stable (ish) Economy
But if office roles (and later physical roles) are hollowed out faster than new forms of work appear, that link inevitably weakens. Companies save on labour, but at the same time shrink their own customer base if fewer people have disposable income. What happens when an economy automates labour faster than it replaces purchasing power?
I’m curious as to where people think this realistically heads over the next 5–10 years. Are we headed for a future where the middle class is shrunk massively and we are all scrabbling for lower paid menial work while also being heavily reliant on the state and universal income? surely the good jobs that remain will also have a such a huge number of applicants that salary decreases for those as well.
It is all looking rather dystopian and bleak to me.
BikeBikeBIke said:
People have always had exactly these fears about new tools or machines and they've always been unfounded.
Let's hope this time it's just the same as all the other times.
I think that what is coming down the line this time is much more impactful and will also happen much faster.Let's hope this time it's just the same as all the other times.
UBI is inevitable. The objections and obstacles such as they exist today will be overcome, just as we are doing with the transition to EVs.
We might as well crack on with it.
What's absolutely imperative is that we don't elect a reactionary, nostalgia and false 'common sense' driven political party to government, which will be implacably opposed to any form of progressive, er, reform in this matter.
We might as well crack on with it.
What's absolutely imperative is that we don't elect a reactionary, nostalgia and false 'common sense' driven political party to government, which will be implacably opposed to any form of progressive, er, reform in this matter.
272BHP said:
I keep coming back to the idea that AI and automation is a social contract shift as well as a technological shift.
For decades the deal has been:
Work --> Wage --> Spend--> Stable (ish) Economy
But if office roles (and later physical roles) are hollowed out faster than new forms of work appear, that link inevitably weakens. Companies save on labour, but at the same time shrink their own customer base if fewer people have disposable income. What happens when an economy automates labour faster than it replaces purchasing power?
I m curious as to where people think this realistically heads over the next 5 10 years. Are we headed for a future where the middle class is shrunk massively and we are all scrabbling for lower paid menial work while also being heavily reliant on the state and universal income? surely the good jobs that remain will also have a such a huge number of applicants that salary decreases for those as well.
It is all looking rather dystopian and bleak to me.
People talk of a UBI with automation but the British state is nearly bankrupt already and the high earners are starting to exit due to high tax levels. So there isn't going to be the money for it. For decades the deal has been:
Work --> Wage --> Spend--> Stable (ish) Economy
But if office roles (and later physical roles) are hollowed out faster than new forms of work appear, that link inevitably weakens. Companies save on labour, but at the same time shrink their own customer base if fewer people have disposable income. What happens when an economy automates labour faster than it replaces purchasing power?
I m curious as to where people think this realistically heads over the next 5 10 years. Are we headed for a future where the middle class is shrunk massively and we are all scrabbling for lower paid menial work while also being heavily reliant on the state and universal income? surely the good jobs that remain will also have a such a huge number of applicants that salary decreases for those as well.
It is all looking rather dystopian and bleak to me.
As for the impact, yes lots of labour shedding in administrative jobs, many which paid close enough to the average wage to be considered lower middle class occupations.
The problem is not really AI/Automation. In fact it should be the answer to the productivity crisis since the switch to a more service based economy. The problem is that it comes after a sustained period of a mismanaged economy, energy policy and immigration system. That is what creates your dystopian future.
272BHP said:
BikeBikeBIke said:
People have always had exactly these fears about new tools or machines and they've always been unfounded.
Let's hope this time it's just the same as all the other times.
I think that what is coming down the line this time is much more impactful and will also happen much faster.Let's hope this time it's just the same as all the other times.
In my own case AI (as it stands) can't replace me but has made me hyper productive in a way that will shut cheaper foreign competitors out of the market. (If I'm producing 4x as much a cheap engineers in Vietnam doesn't look so cheap, even if they are also 4x as productive due to AI.)
If Artifical General Intelligence reaches the point that it can perform as well a human on any task you could throw at it, it will not be very long at all until it it utterly outperforms even the smartest human. Think hours or days, rather than weeks or years.
And it will continue to improve until we are to that intelligence, what bees are to us... and then carry on. It's limiting factors is the raw compute power. Datacenters. (Which companies happen to be building a lot of right now....)
And after that happens... humans either become pets being looked after or we are a pointless waste of carbon, hydrogen, oxygen, nitrogen, calcium and other trace elements.
People like Elon Musk think the best mitigation for that is to merge humans with AI, and enter a symbiotic relationship. That opens a whole can of worms.
On the less dystopian/utopian potential, is that AGI (and the Artificial Super Intelligence that swiftly would follow) is much further away,and that people adapt to using AI as an output multiplier for thinking tasks. Some companies will obviously think "hey we can get the same output for half the people"... but the successful companies will be thinking "we can get double the output with the same amount of people".
And it will continue to improve until we are to that intelligence, what bees are to us... and then carry on. It's limiting factors is the raw compute power. Datacenters. (Which companies happen to be building a lot of right now....)
And after that happens... humans either become pets being looked after or we are a pointless waste of carbon, hydrogen, oxygen, nitrogen, calcium and other trace elements.
People like Elon Musk think the best mitigation for that is to merge humans with AI, and enter a symbiotic relationship. That opens a whole can of worms.
On the less dystopian/utopian potential, is that AGI (and the Artificial Super Intelligence that swiftly would follow) is much further away,and that people adapt to using AI as an output multiplier for thinking tasks. Some companies will obviously think "hey we can get the same output for half the people"... but the successful companies will be thinking "we can get double the output with the same amount of people".
When you look at future AI, such as AGI or ASI you need to be very careful as to your definitions.
Take AGI. There are two extremes of the same thing. The first is more or less what we have now, but is get the answer right more often, even with really silly complex problems. That could be considered AGI. But it complete lacks the dangerous bit, which is the other end of AGI... where that AI instance has agency, where it can proactively do what it thinks is best. It's not a response to a prompt, it's sitting watching the world doing what it wants when it wants.
That later sort of AGI is much less likely, you can't just sale an LLM to make it bigger and better, it needs all sorts of other moving parts, worldview, memory, prioritisations, task planning etc. While we have ideas of how to make all those parts, we still haven't got to the point of a "spark"... ie. what we call consciousness. And so everything is still deterministic, and can't really do what is described in that extreme example.
So yes, having clever, more correct LLM chatbots will make some jobs a lot simpler, we're a long way off the extreme idea of AGI doing things for us entirely automated and on it's own whim. That's when things start to get a bit more dangerous. It's not impossible, it's just not a matter of throwing data centres at it, it needs real discovery first.
Take AGI. There are two extremes of the same thing. The first is more or less what we have now, but is get the answer right more often, even with really silly complex problems. That could be considered AGI. But it complete lacks the dangerous bit, which is the other end of AGI... where that AI instance has agency, where it can proactively do what it thinks is best. It's not a response to a prompt, it's sitting watching the world doing what it wants when it wants.
That later sort of AGI is much less likely, you can't just sale an LLM to make it bigger and better, it needs all sorts of other moving parts, worldview, memory, prioritisations, task planning etc. While we have ideas of how to make all those parts, we still haven't got to the point of a "spark"... ie. what we call consciousness. And so everything is still deterministic, and can't really do what is described in that extreme example.
So yes, having clever, more correct LLM chatbots will make some jobs a lot simpler, we're a long way off the extreme idea of AGI doing things for us entirely automated and on it's own whim. That's when things start to get a bit more dangerous. It's not impossible, it's just not a matter of throwing data centres at it, it needs real discovery first.
Rivenink said:
he successful companies will be thinking "we can get double the output with the same amount of people".
For the moment in developed economies this is certainly true. Our productivity is low and AI is the equalizer. Most western companies need more work done more than substantially lower employment costs.Of course that could change.
Skodillac said:
UBI is inevitable. The objections and obstacles such as they exist today will be overcome, just as we are doing with the transition to EVs.
We might as well crack on with it.
What's absolutely imperative is that we don't elect a reactionary, nostalgia and false 'common sense' driven political party to government, which will be implacably opposed to any form of progressive, er, reform in this matter.
I agree. Sadly it's inevitable. We might as well crack on with it.
What's absolutely imperative is that we don't elect a reactionary, nostalgia and false 'common sense' driven political party to government, which will be implacably opposed to any form of progressive, er, reform in this matter.
What's coming is a ghastly world of surveillance, micromanagement, compliance and blackmail balanced with the offer of a basic array of income and services (the bare minimum) wrapped in the fluffy rainbow comfort of convenience.
700k graduates on benefits. they have no hope of having anywhere near as nice a life as their boomer parents. they might as well stay home, play video games and masturbate.
So they will be placated with crumbs off the table and blackmailed into quick release/instant gratification and dopamine addiction. No wonder fertility is below replacement.
Lovely.
BikeBikeBIke said:
People have always had exactly these fears about new tools or machines and they've always been unfounded.
Let's hope this time it's just the same as all the other times.
This was the fear when I left school. All the jobs would be taken by robots. I decided to be the bloke that builds the robots. I left school and started working, making electro mechanical prototypes. Let's hope this time it's just the same as all the other times.
I am now retired and there are just as many jobs as there ever were.
Jobs will change, but they always have and always will. Try earning a living as a coal miner or knocker upper these days.
My prediction is there will be loads employed in trying to solve the mess AI has gotten us into
allegro said:
OP I would point you to the industrial revolution, ludites etc and some reading of history.
While that is true it was a) not much fun for many of the workers at the time. Hence why they were smashing the machines
b) the gains were mostly kept within the country and so expanded the economy. While British companies will likely benefit from the improved productivity the AI being used will predominantly be based abroad, so most of the gains will flow to American companies, no doubt funnelling their European profits through tax havens as they do now.
Result increased UK unemployment and no means to cover the benefits to them on top of an already overloaded welfare state, let alone some UBI for all which is offered up as some sort of magic cure.
BikeBikeBIke said:
Skodillac said:
UBI is inevitable.
Marx suggested if machines were doing all the work, humans were redundant and could just kick back on a state funded holiday and only the ones who wanted to needed to work.Remind me how that panned out?
Skodillac said:
BikeBikeBIke said:
Skodillac said:
UBI is inevitable.
Marx suggested if machines were doing all the work, humans were redundant and could just kick back on a state funded holiday and only the ones who wanted to needed to work.Remind me how that panned out?
768 said:
Skodillac said:
UBI is inevitable.
I have a feeling it might, and that it won't be enough to live off. But there won't be any other choice for most.For those who need more, then there's salaried work and entrepreneurial activity to supplement.
To JagLover's inevitable "we can't afford it" knee jerk, we will be making significant savings by abandoning every other form of welfare and pensions. Then we work out how to tax the populace to support the shortfall. It's challenging, but not impossible. "The wealthy will just leave" Cassandrism can also be addressed. There will be change, disruption and there will be winners and losers. It will all come out in the wash and a new normal will emerge. I wouldn't mind betting that the voices least keen on UBI because of "OMG disruption" will be the same voices who were so keen on disrupting the status quo with Brexit. So I take their doom-mongering with a pinch of salt - it's always rooted in their own personal prejudices and fears.
BikeBikeBIke said:
Skodillac said:
UBI is inevitable.
Marx suggested if machines were doing all the work, humans were redundant and could just kick back on a state funded holiday and only the ones who wanted to needed to work.Remind me how that panned out?
Worked out pretty damned well in that context, but then they'd not just got machines to 'do the work', but also to make sure there's so much excess resource that there is no reason for it to fail due to human nature.
Skodillac said:
It will be enough to live off for some (e.g. the retired - and then most of those will supplement with private/workplace pensions, investments, home equity etc).
For those who need more, then there's salaried work and entrepreneurial activity to supplement.
To JagLover's inevitable "we can't afford it" knee jerk, we will be making significant savings by abandoning every other form of welfare and pensions.
So it's enough to live on? If so most people will quit. Or it's not enough to live on? In which case it doesn't replace benefits and pensions.For those who need more, then there's salaried work and entrepreneurial activity to supplement.
To JagLover's inevitable "we can't afford it" knee jerk, we will be making significant savings by abandoning every other form of welfare and pensions.
Gassing Station | News, Politics & Economics | Top of Page | What's New | My Stuff