One being built near me is being built on public lands that are being sold off for datacenters. That's an additional pretty big negative (especially since the specific land was set apart as a wildlife habitat.)

I do really dislike that AFAIK almost all of these are being powered via burning fossil fuels. Any way you slice it, it's going to mean that locals end up paying higher energy bills.

I guess we can assume that the executives at the big AI companies don't believe in man made climate change.

Climate change is a lot less important when you are rich enough that it doesn't impact you personally, and when you don't care about what the world looks like after you die.

Sure, millions of people are going to die due to climate change, and that's going to impact the company's profitability, but by then your stock options will have already vested - so that's a problem for someone else!

They could be doing more to combat climate change, but that's also true for everyone in this thread.

Personally I will forgive them the Datacenter carbon if they are willing to stop flying private jets.

A(G)I is supposed to fix everything so nothing matters /s

Why is this sarcastic? Isn't this a plausible possibility?

We've been hearing some variant of "engineering will solve climate change when it becomes important enough, so we don't need to worry now" since the late 90s.

> We've been hearing some variant of "engineering will solve climate change when it becomes important enough, so we don't need to worry now" since the late 90s.

Well, it was true in the 90s, and it's still true today. Once climate change becomes important enough (that is, causing actual significant damage to important, wealthy countries), it takes just one such country to start a stratospheric aerosol injection project, or, if space launch costs become low enough, put a sunshade between Earth and the Sun. Such solutions have their own risks, of course. But if climate change really becomes bad, they will be tried. And from what I have read, they have a good change of succeeding.

It will still be true another 30 years from now perhaps, but nobody who has the actual funds to spend seems to care. If the world was actually taking it seriously, it would be dealt with like the ozone hole was. Instead, we're still debating whether it's real, and if it's real whether it's important, and if it's important whether it is more important than - I don't know - pretty much anything else.

What qualifies as a rich nation anyway? Europe seems pretty rich, and they're melting. One would think the time is now.

> What qualifies as a rich nation anyway? Europe seems pretty rich, and they're melting. One would think the time is now.

They are still not melting enough. Most of the negative consequences of current summer heatwaves can be pretty much solved by installing US-style ubiquitous air conditioning.

Yup, as I said. Same as rich people do everywhere, they don't have to worry, they won't run out of water or power, if things get too hot where they are they'll just move somewhere nicer. They have the money and zero incentive to solve the actual problem.

An extremely wealthy acquaintance in my local circle is happy to endlessly debate whether climate change is ACTUALLY real or just scientists carving out job security, while flying out for monthly ski holidays wherever there's nice snow in the world when it gets a bit muggy at home.

It is already causing actual significant damage to important wealthy countries. Just not in a way that directly impacts the people with the means to make a significant impact.

It worked out well for every problem so far.

Unless they've somehow restored Norway's glaciers with a refrigerator of some sort, I think you might be mistaking a local maximum for a global one.

Unless it’s social media, advertising, sex trafficking, or fraud.

It’s silly to pretend that technology is anything other than a double edged sword.

I did not say it didn't create new problems, or addressed everything already.

I'll grant you social media and advertising, but the other two are not new, and were way worse in the past.

There's not.

The first problem, in order to reverse a chemical reaction, you need to provide at least the offsetting released energy of the initial reaction. That means for CO2 you need to provide at least as much energy as was released when the carbon oxidized in the first place.

The second problem is that the CO2 is diluted in the atmosphere. Consider the amount of energy needed to remove salt from ocean water, it's not simple.

It just takes a lot of energy to that what was done by burning fossils in the first place.

Maybe not. You don't need to chemically turn the CO2 back into hydrocarbons and free oxygen. You just need to remove it from the atmosphere.

Sounds significantly more expensive than not producing it in the first place though, doesn’t it?

Except for the politics

> The first problem, in order to reverse a chemical reaction, you need to provide at least the offsetting released energy of the initial reaction. That means for CO2 you need to provide at least as much energy as was released when the carbon oxidized in the first place.

This is only a problem if AI won't develop cheap carbon neutral energy sources, like fusion or advanced fission, or more efficient solar with cheap grid-scale batteries.

There are many practical substances that can absorb CO2, and "regenerating" them may not be necessary (the resulting carbonate compounds can be useful without releasing the carbon) or as energy-intensive as the original reaction. This has been an active field of research for over a quarter of a century.

there needs to be some form of energy input to analyse the co2 to another molecule. the logical input would be the sun. but why not use the sun directly via solar instead of going the long way over fossils?!

Nobody in the thread is disagreeing with using solar power. The context is an argument about dealing with the CO2 that's already there, and how feasible those approaches are.

There are non-fossil sources of energy, though. Even without a major discovery like fusion power, we could in theory stop the progress of global warming at any time by building a lot of fission powerplants powering direct carbon capture.

Given that everything important has trade-offs, no, AI, even if it were very very smart wouldn’t be able to “fix everything”

If we compare 2025 to 0025, a truly remarkable number of problems have been solved by, effectively, intelligent people doing research. While it is impossible that AI will fix everything it is plausible it will fix most things since it looks like it'll be able to do research faster, better and with a higher quality than humans as well as more effectively communicate the results to the masses.

Yet Altman and Amadeo are talking how AI can kill us. So either it solves some problems it creates or kills off humanity... This is a gambling addict level rationalisation of betting your house to win it all at once.

The path they are talking about solves also a lot of problems, tho. They don’t care, because they’ll already live on Moon, Mars or will be dead.

Do you think an LLM can change the laws of physics?

Which societal problems require changing the laws of physics to fix?

Entropy means the costs of cleaning carbon out of the air are only going up.

The rule of thumb is that technology creates problems that cannot be solved with the same level of technology.

It’s sarcastic in context. My comment doesn’t make sense without GP.

Or to spell it out more clearly, the sarcasm has nothing to do with the truth value of the claim (I’m not making an assertion in either direction).

Google, at least, purchases 100% of its electricity from renewable sources. This doesn't mean that they have renewables feeding directly into all their datacenters at all times -- but for every kWh they consume, they purchase a kWh of renewable energy somewhere and add it to the grid.

Do they shutdown when there is no renewable power available? I doubt they have enough battery capacity... The whole per unit buying is misleading. In the moment is what really matters. Not to forget them paying for renewable power can move others to non-renewable sources...

What it means is that they only pay for generation that's renewable and the energy company assures them by contract that they'll only be paying for renewables.

The rest of us, who might be on green tariffs usually get a commitment that on average almost all our energy comes from renewables. When supply is short then I'm the one who's paying for gas generation rather than the company who has committed to greenwashing. But that also drives investment in renewable generation, in the UK renewables have gone from 2% of power generation in 1991 to around 50% now.

No, but they cover their fossil fuel energy usage by entering into PPAs [1] for new clean energy projects that get added to the grid. This allows renewable projects that would otherwise not have been economical to finance to get built. This is additional renewable capacity that that goes into the energy mix. It moves others onto renewable sources, often without them even knowing it.

From a climate change perspective, what happens at one particular moment doesn't matter -- it's CO2 emissions globally over time that do.

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

By me they are converting cornfields to data center campuses. People hate it and are against it.

They are also building solar farms independent of the data centers. People hate it and are against it.

They just built a Costco which turned some cornfields into a giant warehouse. No signs, no protests, no complaints at the city council meetings.

What's so great about cornfields? I can imagine hating it if they were deforesting the Amazon or something.

The argument they make is, "it is food and we eat it". Nevermind most of the corn grown is not eaten by people.

Yeah, but our food eats it too.

Mostly our cars drink it

45% and rising.

I have bad news about what they are deforesting the Amazon for (hint: cornfields)

Chocolate and coffee I believe

By area it's almost entirely corn and soy (mostly for animal feed)

In the US, there's currently about 300x more land used by cornfields than by data centers.

Does that seems unreasonable? We already grow more than enough food in this country and surely we're not getting 300x more value from the corn.

You do get less efficiency from the gas you put in your car though.

I wonder what causes more harm? High fructose corn syrup or data centres?

It's almost like people want things built that they directly benefit from.

A warehouse provides hundreds of lower education jobs. Well at least until they go completely robot.

Not everyone is a software developer.

Aren't most people directly benefiting from AI?

No? I can think of so many things in life that were better just a couple years ago.

And you think they wouldn't directly benefit from solar power?

What if the cornfield owners got paid market rates for the power generated when that land is given over to solar? This should be (and ChatGPT agrees) much more profitable given that it is vastly more energy efficient.

Directly? No. Indirectly? Of course they do. Everyone does.

Realistically the solution to NIMBYism is to pay the neighbours, but that's not really practical.

NIMBYism isn't rational, it's a manifestation of fear of change, even if they past state of affairs was worse than the future one. I live in a city where the anti-datacentre campaign isn't about some pleasant locale, but a disused toxic industrial land in a steel-mill complex over half a km away from any actual homes.

The corn field is a known quantity.

> Realistically the solution to NIMBYism is to pay the neighbours, but that's not really practical.

If we pay each farmer to generate solar (and maybe also stop paying them to grow corn beyond what's actually economically viable), why would they care what the next farm over is doing? I thought we were talking about places where people live too far apart for "backyards".

They can make all objections or decisions they want on their own private property. It seems people have too much levers to pull to control what others can do with their property.

A lot of people have dumb opinions grounded in ill-thought-out ideas like "natural cornfields are morally superior to human-built infrastructure buildings" or "solar power is a Democratic party conspiracy". In general it should be easier for property owners with skin in the game to make good land use decisions to make them without having to need small-d-democratic opposition from locals with bad incentices and values.

On Nextdoor, I have met several people who believed solar and wind polluted more than coal plants.

I see that from time to time on HN. It’s hard to believe they’ve never encountered mercury or arsenic in their news.

Anyone with kids probably should be aware of Mercury - it's an important guideline during pregnancy, all the mercury in fish. I don't have kids, so I am not sure if the doctors explain how all the mercury got into the fish...