Generative AI could generate tons of e-waste by decade's end

https://techxplore.com/news/2024-10-generative-ai-generate-millions-tons.html

by bikenaga • 2 years ago

34 41 2 years ago

techxplore.com

wildrhythms• 2 years ago

But have you considered all the money to be made storing and running ads on that e-waste?

ForOldHack4• 2 years ago

Even in death, old TVs still command attention.

drewbeck• 2 years ago

Fortunately we will just be able to ask AI what to do about it. I’m sure it will come up with something.

pimlottc• 2 years ago

It keeps telling me to allocate more resources to making paper clips

lagrange77• 2 years ago

"Oh, thank you for pointing that out, you're totally right! My previous answer would have killed humankind as a side effect. Here's an updated version of the solution: ..."

louwrentius• 2 years ago

The current AI craze doesn’t feel like it’s really about AI. It’s Silicon Valley being desperate for (growing) growth and shareholder value.

Another obvious aspect it that it may help do work that could be done by people, thus helping to keep the worker class - that’s you! - subdued, fighting amongst yourselves.

Meanwhile the ‘good stuff’ can only be offered by those large companies, who have the power to make them ‘good enough’ thus making the rest more dependent.

It’s been a while since 2007, all the incentives are there to keep the ball rolling, but it will stop at some point.

Then we IT people with six figure salaries will learn that we should have listened more to David Graeber.

kylehotchkiss2• 2 years ago

Don't you think people will get bored of it? I am starting to.

Apple has a few new cool AL based components in iOS 18. But one of the ones they've spent the past year harping on is the Cowboy frog emoji generator.

digging2• 2 years ago

Getting bored of it isn't the same as having no use for it. This stuff is 2 years old; we haven't figured it all out. Even if all development on the actual models stopped now, we'd still be seeing new uses of LLMs emerging for years to come.

Mistletoe• 2 years ago

But one of the ones they've spent the past year harping on is the Cowboy frog emoji generator.

You know it sounds like a line from a David Foster Wallace novel, but nope it’s real life. I get that a lot these days.

gruez2• 2 years ago

The current AI craze doesn’t feel like it’s really about AI. It’s Silicon Valley being desperate for (growing) growth and shareholder value.

Isn't this true statement about literally any other technology? Do you think the railroad craze[1] was really about railroads?

[1] https://en.wikipedia.org/wiki/Railway_Mania

Another obvious aspect it that it may help do work that could be done by people, thus helping to keep the worker class - that’s you! - subdued, fighting amongst yourselves.
Meanwhile the ‘good stuff’ can only be offered by those large companies, who have the power to make them ‘good enough’ thus making the rest more dependent.

If you replaced the preceding paragraphs to be about weaving machines, this could pass for ned ludd's writings.

xanderlewis3• 2 years ago

I think the point is that it isn't really about creating something people want; it's about creating something that inevitably becomes a necessity for life and work whether you like it or not.

Twenty years ago we had the iPod, and 'tech' consisted of exciting consumer products we all wanted in our pockets and services that would enrich our lives by enabling our creativity and connecting us. Of course in business one aims to make a profit, but one could hope to do so simply by making things people wanted. Today, it doesn't feel like that. Sure — recent AI developments are hugely impressive, but what value are companies like OpenAI creating and who, other than AI geeks and researchers, is really hankering for this stuff? It feels more like an arms race towards something completely unknown. It's a collective panic that AGI (read: God) is just around the corner and whoever gets there first wins everything.

There was no such insanity in previous 'tech' eras. There's always been competition and hype, but the prospect of unlocking the technology to end all technologies has never before been credible enough.

xnx• 2 years ago

Dupe: https://news.ycombinator.com/item?id=41974159

kylehotchkiss2• 2 years ago

One Person's Trash is Another Crypto Miners Treasure.

Synaesthesia2• 2 years ago

Funny because "e-waste" is also how I would characterise the content that these generative AI models output.

ForOldHack4• 2 years ago

Saying AI will generate tons of e-waste is redundant. AI is waste.

stavros3• 2 years ago

I really don't understand the utter lack of imagination on HN lately. Here's an amazing technology that nobody could even fathom existing five years ago, and people's take is "AI is waste".

Like, really? If you've used ChatGPT and it didn't do anything useful for you, I'm sorry, but the problem was between the keyboard and the chair.

Spivak2• 2 years ago

How are you on HN with this kind of attitude? I get that AI is overhyped for a lot of applications but I just shipped an LLM powered feature at $dayjob that saves around 4-8 hrs per use. Our previous version based on NLP had been scrapped for years because it performed so terribly. We just paid humans to do it. The difference between the old and the new is scoring <5% correct to consistently hitting 95%. It's that bird xkcd.

jiggawatts2• 2 years ago

These “generates e-waste” articles are political hit pieces. I hope we’re all intelligent enough to recognise the absurdity of the statement.

In case it is unclear, a similar article was going around the Internet a few years ago complaining about the “tons of e-waste created by Apple Airpods”.

Yes, that’s true in the sense that Apple was producing “tons” of e-waste annually… two of them. One ton wouldn’t be “tons”, you see, so the article was technically correct in the dumbest and most nakedly deceptive way possible.

Waste is about inefficiency. It implies no or negative utility. There’s a hint of huge landfills and truck after truck of bulky stuff filling that up.

AirPods and AI chips are the opposite of inefficient! They’re some of the smallest, densest, most concentrated and space-efficient products ever made by man.

To invoke the word “waste” in juxtaposition to these things is so absurd, so contrary to their essential nature that it’s clearly a deliberate falsehood, signalling membership in a group by its absurdity. It’s like the official statements on RT that yet another oligarch “fell out” of a window. Everyone knows what really happened there. The claim of defenestration is a coded signal telling the listeners a very different message to the verbatim text.

Someone who is scared of AI wrote this article or had it written on their behalf. It’s the product of a Luddite, or someone whose job is about to be replaced.

“Stop the efficiencies! My career as a copy editor is at risk! I mean… it’s surely wasteful to optimise my job away with a chip that’s a mere square inch in size and weighs a gram!”

Nullabillity• 2 years ago

AirPods and AI chips are the opposite of inefficient! They’re some of the smallest, densest, most concentrated and space-efficient products ever made by man.

Even if we were to take these claims at face value... none of these things imply that the thing is useful.

jiggawatts2• 2 years ago

They’re more useful than mine tailings or household waste, both of which are generated at a scale many orders of magnitude greater than AI-specific chips.

Even as discarded waste, electronics tend to be more useful to recyclers because they contain copper.

jeffbee• 2 years ago

It sounds ridiculous in the article but without the PDF of the paper it is hard to judge. They are saying that the annual rate of e-wasting will increase 1000x in ten years, which sounds a bit off considering how large the IT industry already is, and how small of a component the AI subindustry is. If I had access to the PDF I would scan it for several common errors, such as assuming the perpetuation of an apparently exponential process, or writing down an unrealistically short economic life for the equipment. Warranties often run out in 3-5 years but cloud operators keep their gear for 10-15 years.

94b45eb4• 2 years ago

It’s a race to create an AI good enough to solve the e-waste problem before the e-waste problem becomes such a large problem that it prevents AI advancement.

soco• 2 years ago

It's a race to ignore all this, calling it "woke snowflake worries lol", and enjoying the short term economical benefits of AI making the shareholders happy. And I'm not even sarcastic.

ForOldHack4• 2 years ago

Until of course, "Demon Seed" comes true.

Yoric2• 2 years ago

So far, I haven't seen any sign that LLM/GenAI could be used to solve e-waste. It's definitely an impressive technology, but seems to be much suited at generating porn than at solving global issues.

meta_x_ai• 2 years ago

oh look another article that brings the HN luddites come out of their woodwork and spew "hurr durr AI hype/bad/waste"

Generative AI could generate tons of e-waste by decade's end

Related Stories