site banner

Friday Fun Thread for April 12, 2024

Be advised: this thread is not for serious in-depth discussion of weighty topics (we have a link for that), this thread is not for anything Culture War related. This thread is for Fun. You got jokes? Share 'em. You got silly questions? Ask 'em.

2
Jump in the discussion.

No email address required.

To some extent, it's understandable - it must be a pretty big blow to one's ego for the art one prides themselves on to be so easily recreated and automated by the equivalent of a Chinese Room, especially when the field is still in its infancy - but I can't help but see many of the naysayers about the ability of AI to achieve supposedly uniquely "human" tasks as being clearly myopic and wrong.

Ego is one thing, basic self-interest is another, but more importantly, I'm not sure who's being myopic here. Technology was supposed to automate away the drudgery, so we can devote ourselves to higher pursuits like art, philosophy, and science. The way things are going, it looks like we're going to automate away the higher pursuits so we can send more people to the cobalt mines, because handing someone a pickaxe, and feeding them insects might be easier than figuring out and maintaining bipedal robots.

Technology was supposed to automate away the drudgery, so we can devote ourselves to higher pursuits like art, philosophy, and science.

I think technology has done a lot of that, though, with things like dishwashers, washing machines, running water, elevators, cranes, cars, and such. These are all "dumb" tech, though, and they hit diminishing returns; we still have to load our dishwashers and steer our cars (for the most part) manually. I think this just speaks to how difficult the precise and fine manipulation of objects in the physical world really is. From what I heard, image generation AI was actually a consequence of trying to solve this problem; we needed AI to be able to perceive the world similar to humans, which meant identifying objects in images, which was able to be reversed in some way to create new images. And this happened much more quickly than the robotics controls, because manipulating stuff in the digital world is much easier than in the real world.

It's still way too early to tell, but I could also see the argument that AI art does automate away the drudgery so we can devote ourselves to higher pursuits, since it's really primarily good at creating high fidelity illustrations while lacking the good taste or artistic vision required to convey some emotion in a pleasing or provocative way (this is arguable). This allows people to work on the more high level vision of what they want their illustration to look like instead of devoting the time required to develop their manual muscle control.

Whether this is good or not is a question of values and not really related to the point, the topic of discussion is more about whether it's possible.

I think your scenario is unrealistic in any case - automation of manual labour tasks is certainly feasible (and has been achieved in many cases) and more such jobs in these domains will eventually become obsolete once technological advances make the cost of doing so lower than employing human labour, but that's besides the point. You can be an AI doomer and still realise that AI has immense potential. Plenty of the people discussed on this forum certainly believe so (Yudkowsky, Bostrom, etc). But there are still a lot of people basically treating AI as a hype-fad pushed by techbro caricatures, who regard automation of all these oh-so-human pursuits as practically impossible and scoff at the mention of AGI, and pretty much every two years their predictions get overturned.

automation of manual labour tasks is certainly feasible (and has been achieved in many cases) and more such jobs in these domains will eventually become obsolete once technological advances make the cost of doing so lower than employing human labour

Way back when, before I joined this place but was already keeping an eye on the rat-sphere, one of the AI Apocalypse scenarios thrown around was "technological unemployment", and one of the counter-arguments to it, other than "it's just a hype-fad" was "comparative advantage". Already at that time the Rats have discredited themselves as careful thinkers in my eyes, because they waved it away with the magic of recursive improvement. A careful thinker would look at where the logic of comparative advantage leads, and it's the scenario I outlined. You see it's not enough that we eventually figure out robotic/AI alternatives to manual labor, and it's not even enough that they are strictly superior to human labor in every way, they have to be superior to devoting the same AI resources to something else while having humans dig the fucking hole.

But back to your point, yeah the progress in AI is pretty wild, and people predicting that nothing will come out of it were clearly wrong. At the same time I'm having trouble saying anything meaningful about it, or where it will end up going, which is why I kind of went off topic relative to your original post.

You see it's not enough that we eventually figure out robotic/AI alternatives to manual labor, and it's not even enough that they are strictly superior to human labor in every way, they have to be superior to devoting the same AI resources to something else while having humans dig the fucking hole.

And yet that's not how mining works in any country but the shittiest on earth, so it's not clear to me why the practice would suddenly spread.

Yes it does. People who have talents in more lucrative fields than mining, tend to go those fields, even if they are better than the average miner.

Almost all mining is already "automated" to a certain extent by either gigantic machines that can do the work of 100,000 kids with pickaxes or other force multiplying tech. He was asking why we would ever go back to that when we have better cheaper ways?

In developed countries, yes., in others not necessarily.

Secondly, the reason we would go back to that is that if we come up with a technology to automate away valuable intellectual labor, you'll get more bang for the buck by investing in that, than investing automating manual labor. Trying to refute the idea with the current economics of mining makes no sense.

There is money/utility in automating all labor, they can't do it in the middle of the congo because they don't have the capacity, infrastructure, expertise, and the situation is too volatile for western mining companies to set up in the worst parts. It isn't a value prop. A single mining wheel can mine more in 240,000 tons of material a day, with the largest mines extracting about a million tons a day. The "big hole" the largest hand dug open pit mine took some 30k miners and 30k support townspeople 43 years to excavate 22 million tons of rock by whatever semi manual means were available to them at the time.

So an almost fully automated mine can do in 22 days what it took 30-60 thousand people 43 years to do in the past. I bet no one even died in those 22 days, or even in 2 years, most open pit mines are pretty safe these days!

"Between 1897 and 1899, a total of 7,853 patients were admitted into Kimberley Hospital. 5,368 of these patients were black and admitted into special designated wards, i.e. a "Native surgical ward" for black miners and a special ward for black women and children. Of these black patients, 1,144 died."

Everything will mechanised and automated. It just makes too much sense, which is why everyone who can manage to do so, does it. Humans aren't the new cogs, they are going to be completely irrelevant.

So an almost fully automated mine can do in 22 days what it took 30-60 thousand people 43 years to do in the past. I bet no one even died in those 22 days, or even in 2 years, most open pit mines are pretty safe these days!

Everything will mechanised and automated. It just makes too much sense, which is why everyone who can manage to do so, does it.

What you're missing is that the question isn't just what a fully automated mine can do compared to 30-60 thousand people, it's also what those resources could have done if they didn't go to automating that mine. There's a reason why rich, automated, countries are buying resources mined by manual labor.

This is all pretty basic economics

More comments

"not how it works" applies to your claims that mining will consist of "handing someone a pickaxe, and feeding them insects".

You haven't heard of "artisenal mining"? Or are you objecting to the "feeding the bugs" part?

I'm saying that there's a reason artisanal mining is limited to the worst places in the world, and it's not because the Congolese have a comparative advantage in digging holes.

and it's not because the Congolese have a comparative advantage in digging holes.

The hell it's not! If they didn't have a comparative advantage there, they wouldn't be able to sell their cobalt.

More comments

I took it as a metaphor.