First is a slave taking care of "things" that I can't - I use slave in the sense nobody would ask anyone they cared about more than a slave to wipe their butt for them. There are a lot of people who cannot wipe themselves (or operate a bidet), and I want a machine to do that task.
Second is a social help. Machines can do that somewhat, but only partially. I think that real humans still remain the best here, but machines still can help - in the same way that having a cat can help but isn't the same as a human.
Compared to the, sadly too common, current hellscape of some underpaid and over worked person who comes by and does the bare minimum (if that) in a rush before hurrying off to their next client.
This thought is likely to be revolting to many people, but most of them are going to compare it to a reality that may not exist: They want infinite human caregivers with infinite patience and infinite money to pay those caregivers, but sadly all of those things are finite.
AI and automation are often second-best solutions to some idealized version of reality, but when the idealized version isn’t attainable pragmatic people start accepting the second-best solution. There are a lot of unpragmatic people who will get very angry at the second-best solution for existing, though.
Part of the reason why so many governments have banned single room occupancy housing (the other part is a selfish desire to not allow “those types” to live nearby).
A society that is willing to leave the care of people in the "hands" of machines seems so uncaring to the point of not caring for people at all. I'm fully expecting that the sort of society that will leave human care up to automation would just as easily run people through furnaces instead. After all, no one cares enough to look after them so why keep them around at all?
> when the idealized version isn’t attainable pragmatic people start accepting the second-best solution. There are a lot of unpragmatic people who will get very angry at the second-best solution for existing, though
If you wanna talk about unpragmatic versus pragmatic, it's clear what the actual pragmatic choice is, isn't it?
Btw, those who could afford it have always outsourced this. The blessings of the upper class over millenia.
This is true in some cultures
In other cultures it was expected that the elderly, infirm, sickly, or crippled would go walk into the woods so they were no longer a burden for the younger generations. Many cultures used to kill children that were born "wrong" too.
I'm just saying it's entirely possible the value of human life is currently going down, not up. I find that very troubling
I don't think that many people will have a problem with a robot cleaning wounds, changing bandages, preparing meals because you can't, etc... if it does a good job. Care here is just making sure your wound heals properly. A nurse can also provide psychological care in a way machines can't, but does it have to go together with treating a wound? Some people don't need psychological care, they just need a new bandage, others have a body that needs no intervention but may need psychological care.
There's such a massive gap of things in between that are far more sane and likely where things would end up if such robotics actually pan out.
For example, you are talking about gentle robots. But who ensures the robot offers gentle care?
Then you contrast it with underpaid and overworked person. Is the system that underpays and overwork human workers the same that will ensure the robots will be gentle?
If we are talking about changing the system (and we should be talking about this): who will do it?
And even with that money there's a shortage of Brits willing to wipe old folks arses so they bring in foreign labour to do it.
It would even be more efficient! Imagine if a 10000th of the money put in AI would go to caregiving? And it wouldn't even be a bubble that risks bursting, eating the pension you'd need to pay for that robot, and it would be at least somewhat useful!
You'll likely need human contact.
Providing full-time care for seniors is incredibly demanding, both physically and mentally. There is a limited pool of people with the requisite mindset and aptitude to do it well. Better renumeration, while certainly deserved by those people, won't automatically increase the size of this pool, it may simply attract more ill-suited people to the profession.
And even with the most compassionate and skilled carers, needing to rely on other human beings for your physical needs results in a loss of dignity and independence. It also creates a power imbalance that will always make abuse or neglect a possibility.
Done right, automated care offers the possibility to provide high-quality physical care that is more convenient, private and dignified for the recipient than needing everything to be done by a human. It also allows for that pool of skilled human carers to focus on higher-value support without being spending the majority of the time addressing basic physical needs.
How much do you think we spend on both currently? Total global AI spending was estimated to be about $1.5 trillion by Gartner in 2025. The US alone spends $1 trillion on Medicare, and the home care industry by itself in the US is roughly $300B. And those figures don’t capture the millions of hours family members spend caring for relatives. Globally, spending on care is almost certainly larger than spending on the AI buildout (though perhaps not in some regions like China, which notoriously provides little support for the elderly).
When robots become reasonably competent, caregiving should be near the top of the list for the type of jobs they should be used for, both for ethical and economic reasons.
Complex and varied physical skills, understanding and managing human emotions, and broad medical knowledge.
Anything labeled with fully, always, everywhere… is very hard to perform. Meanwhile small steps are a where the tastier fruits hangs.
No takesies-backsies!
A lot of parallels to the debate on tariffs and free trade: People really don’t like when something new disrupts the work they do. They do like when it disrupts other companies’ (and therefore, other people’s) work and makes the things they need cheaper.
There's a group of people whose job is precisely to define this both in abstract and in practice, and many do this full time: Judges, labor attorneys, legislators.
>and which jobs are “worthy” or not of being reserved to humans.
That would be legislators, although to be fair they act as proxies for citizens, so it would be literally up to all of us to define that.
Definitely not dreamy, it's down to earth legislative action.
I'd be a bit less dreamy and ask you directly, do you think automated caregiving should be allowed? Because there's a very concrete proposal that you glossed over in order to criticize the abstract argument. Point blank, should machines take care of the elderly? Much simpler discussion.
Not even sex is off limits, there will be sexbots. There will be robot boyfriends and bodyguards. Caregivers for sure.
In fact, I can see a lot of parents simply leaving their kids with robots, to be raised by robot nannies — who will have a ton of patience and will have seen millions of cases and know exactly how to handle and motivate your kid better than you and your kid combined.
If the data will show that robots get better outcomes than human parents at raising kids, what do you think a lot of people will do?
1.) The basis of the world economy is labor. Assuming we take AI industry claims about potential impact of the technology seriously AI threatens this arrangement without offering anything new to replace it. Letting a handful of extra-national businesses threaten the world economy is stupid.
2.) AI improves through access to high quality training data. Using AI internally funnels your company's institutional knowledge, proprietary data, and processes into the training pipeline for the AI tools you use. You are now paying to improve the tools your competitors use and building an existential threat to your business by setting the stage for your AI provider of choice to vertically integrate into your industry segment once sufficient training has been achieved.
3. Letting extra-national organizations operate the computer security equivalent of WMD with no oversight is a ludicrous national security threat.
4. On a warming planet with aging power distribution infrastructure and significant and growing local and regional water availability issues, massive increases to either power or water use is deeply unethical if not outright negligent. Doing both simultaneously solely to profit a handful of investors beggars belief.
5. The financials are fucking unhinged. If investment figures are to be believed the AI industry would have to capture 100% of the economic output of France, or if you prefer Ireland, Belgium, Sweden, Austria, Norway, and Denmark combined. Annually. Just to break even.
It's also funny that the jobs he's reserving for us involve changing diapers and housekeeping for the elderly, and aren't the jobs people actually want to do - these definitely must go to machines.
Can we reserve mathematics and software engineering jobs for machines and have robots mop the floors instead? I can't think of a way in which that would make the world worse.
It’s not perfect, but still better than anything else we’ve tried.