Posted by voxadam 9/4/2025
Previously, on https://sustainability.google/operating-sustainably/:
"We've set a goal to achieve net-zero emissions across all of our operations and value chain by 2030." and, in a table later, "Reduce 50% of our combined Scope 1, 2 (market-based), and 3 absolute emissions by 2030. Invest in nature-based and technology-based carbon removal solutions to neutralize our remaining emissions." and "Run on 24/7 carbon-free energy on every grid where we operate by 2030."
Now, in their 2025 Environmental Report at https://sustainability.google/reports/google-2025-environmen...:
"We aim to run on 24/7 carbon-free energy on every grid where we operate by 2030." and "We aim to reduce absolute, combined scope 1, 2 (market-based), and 3 emissions by 50% from a 2019 base year by 2030." (They also restate the thing about "to neutralize our remaining emissions" a bit later in the document.)
Exactly the same stated goals, just written in a different place. There's more hedging around it now: they call it a "moonshot" and say that they might not manage it, but they never actually promised to do it before.
if it's a serious committment to being carbon-free, i think this sounds like a positive change. if it's a moonshot that they have no real intention of actually meeting, then it's a lot less impressive. i guess we'll see...
Not necessarily. In the UK the market for zero-emission electricity is separated from the actual electricity itself. So you basically sell the "greenness" of the electricity separately from the electricity itself. Customers who want green electricity buy the electricity, and buy some "greenness" (called a REGO).
It does make sense, and a lot of people claim it's some kind of scam when it really isn't... But it does make the relationship between the consumption and production of green energy more complex.
For example, suppose 10% of power comes from green energy, but only 5% of people care about it remotely. The market value of a REGO is basically zero. Even if the number of people who care about green energy doubles to 10%, there's still perfect supply so the market value is still zero and it won't have any effect on green energy production. That isn't quite the case but it doesn't seem far off - I couldn't find up-to-date numbers but it seems like the market price of REGOs is on the order of £5-15/MWh, which is very low. Probably because the UK's energy mix is actually very green these days.
No idea how common this system is in other countries.
If you are connected to the grid, the power you use is effectively generated by the power plant with the highest marginal costs. That often means fossil fuels. It doesn't matter whether you are buying generic power from the grid, or if you have a contract stating that the power you use is generated by renewables or nuclear. If you disconnect from the grid, the overall demand goes down, and some of the most expensive generation capacity is no longer needed.
In the long term, if there are many customers buying explicitly carbon-free power from the grid, it may increase the construction of new carbon-free generation capacity. But that effect is difficult to quantify.
Look at those numbers. 100% still is a very ambitious goal. Currently 0 global customers are buying 24/7 carbon free power from the grid, and even getting from 0 to 1 will require a lot of boots on the ground building real infrastructure. Afterwards, getting from 1 to [everyone] is just a matter of scale.
Even the "lazy" strategy of buying up capacity from existing nuclear plants requires outbidding the existing customers (if the law allows it), and only works in regions where nuclear plants actually exist.
It was a byproduct of the EU habit to create markets and competition where they don't naturally exist. You can buy electricity from any power company, and that company can then generate it itself or buy it from the market. Once you have a market like that, it's easy to add requirements such as carbon-free power. As long as the fraction of the market buying indulgences is lower than the share of the power generation meeting the requirements, fulfilling them is essentially free.
Okay, let's ignore "most of the time" and focus on times when carbon-free power was more expensive. Were these companies actually footing the bill in that case? Because increasing the marginal funding toward clean sources when the grid is dirtiest is the whole point.
In any case, carbon-free power from the grid is mostly a legal/financial construct with limited real-world impact. When the grid is dirtiest, market prices are set by fossil fuels with high marginal costs, and clean sources get the same price without any tricks.
You don't need climate pledges when the thing you're pledging to do saves you substantial amounts of money, and we really don't need quite so much climate moralizing when now the transition to carbon free electricity seems obvious, inevitable, and imminent because of direct immediate cost.
Now the question is just how can we accelerate this a little without creating false economies or distorting the market too much, just a small incentive here and penalty there along with funding for big projects to make them happen earlier.
That's precisely why a 24/7 goal matters. Energy is already cheap and clean for a large portion of the day, but covering the rest of the day will require significant investment.
There are a few residential utilities in Australia which give users free electricity for 3 hours a day to encourage users to divert their usage to that period.
Widespread solar deployment significantly changes the electricity market. Further deployment is going to push fossil plants to charge higher rates when they're needed to the point where supply outstrips demand and it becomes cheaper to shut down a plant which will further increase overnight prices.
If I emit 5g of CO2 but purchase similar in offsets - the offsets have to achieve a reduction of 5g unless something like a scam is going on.
Sometimes, the carbon offsets are from projects that would be built anyway, and there is a lot of sketchiness all around.
It might be better than nothing or it might empower companies to pollute more and then say, "Look, we are offsetting all this!"
the classic example is landowners selling offsets for the mature forests on their land - that's good! and incentivising the preservation of forests is good! but that forest already existed, and has existed for hundreds of years, so does it really make sense that some company can count it against their carbon emissions and claim to be "net zero" just because they paid a landowner to not cut it down?
it makes the whole concept of offsets less of an overall pure good than emitters simply not emitting carbon in the first place, because that is actually a very simple and clear calculation.
If you buy a 1 GWh block of wind/solar energy for a month, you're not doing anything to supply clean energy when it's dark and the wind isn't blowing.
"24/7 carbon free" means increasing the time resolution from 1 month to 1 hour, and funding projects that deliver clean energy when you're actually using it. This includes stuff like batteries, nuclear, and geothermal.
From the top of my head, on the previous definition a Google Street View car running the streets for instance would have been required to be net-zero emissions, while in the new definition I read it as out of scope.
Overall that feels like a huge shift.
If China can pave mountains, a little desert with caliche should be easy
Electrical signal attenuation increases with the square of the distance, so you'll lose ~95% of the power to heat loss in the wires if you try to power Seattle from solar in Nevada -- not very eco-friendly, you'd agree? Also the extreme heat destroys solar panels. Also, dust. Also the permitting of stuff across state lines is so time-consuming it's effectively illegal.
There are a lot of very good reasons why we haven't covered the desert in solar panels.
not true. in standard HV-AC lines, power losses are ~10% per megameter. HVDC gets to 3-5%. So Nevada to Seattle would be at most 20% loss, and in practice 15%, and with HVDC closer to 7%.
https://www.nationalgrid.com/sites/default/files/documents/1...
https://en.wikipedia.org/wiki/Inverse-square_law
This is where you picture an expanding wireless sphere of transmission from a point source and since the surface area of this sphere grows by the square of the distance you get this "power attenuates by the square of the distance" rule.
This of course doesn't apply to power over a 2D cable.
My point of view with Tesla vs Edison is that they were both right and wrong under select circumstances.
Power transmission lines at 60Hz primarily have ohmic losses, which are linear with length of the conductor.
Interesting fact - Power transmission lines are long enough that the capacitive and inductive effects do matter a little bit, even though it's only 60Hz. That's why spacing between conductors is important. 3-phase lines will also rotate the order of conductors every so often to keep the average spacing between all pairs of lines similar.
Strongly agreed.
> Electrical signal attenuation increases with the square of the distance, so you'll lose ~95% of the power to heat loss in the wires if you try to power Seattle from solar in Nevada
What? HVDC lines are usually estimated to have 3.5% power loss per 1000 km. Since power transmission is done using power lines, the inverse square law doesn't really apply here.
> There are a lot of very good reasons why we haven't covered the desert in solar panels.
That does remain true however. Cost concerns, grid access concerns, environmental concerns are all good reasons.
I think you can also reduce heat loss by cranking the voltage up, right? I imagine that's how current interstate/cross-country power deals work
> Also the permitting of stuff across state lines is so time-consuming it's effectively illegal.
This is true in general, but in this specific case, there are a lot of obvious ways to get around the problem, because Nevada is a moth-eaten shirt of federal land reservations — Nevada-the-political-entity only owns/regulates ~15% of the land of Nevada-the-geographic-territory.
With the current state of the US federal government, lobbying to privately use one of those federal reservations would be a walk in the park; and once you're going "California -> federal land" instead of "Calfornia -> Nevada", regulation gets a lot simpler.
Fun fact: there's a National Forest in Nye County (bordering California) that runs right up to the edge of the DoE-reserved area where they did the nuke tests. The feds are fine with running HVDC lines through National Forests (they're not Parks, after all), and "repurposing nuked ground for solar" is actually an easy-to-sell narrative at all levels. You could build solar there and backhaul it to California without ever touching land regulated by Nevada-the-political-entity.
You can't cite efficiency percentages in a vacuum to imply they are a better or worse than alternatives, because those aren't percentages of the same kinds of things, and they don't tell you about the economics, production in absolute terms or EROEI.
Generating solar energy in deserts is often done with a mirror based heating system for this reason.
PV are designed to account for heat and "less efficiency" means they risk performing at 17-18% instead of 20%. And it's actually generating more total energy at 18% because more total sunlight is hitting it, an advantage in desserts.
I was thinking more long term though, deserts see much faster yearly degradation than places with more normal temps. (up to 2-3% compared to the standard 0.5-0.8%)
That's just an economic factor rather than a blocker. PVs are cheap as right now, and could be even cheaper if they weren't tariffed. I wouldn't be surprised if PVs in the desert is nonetheless the right approach right now, and not concentrators.
Coal starts with pulling down entire mountains to get to the coal. The whole process starts with environmental destruction and that's how it ends.
The thermal mass of the panels is no where near significant. Especially compared to a run away greenhouse effect we know coal to cause.
[0] https://www.nature.com/articles/s43247-024-01619-w.pdf
[1] https://acs.figshare.com/articles/journal_contribution/Net_Radiative_Forcing_from_Widespread_Deployment_of_Photovoltaics/2871685With proper site selection and albedo managed via density and bright ground treatment, you can expect net neutral local heat impact.
This is reasonable, it's a nearly impossible goal to achieve.
Meanwhile: "Amazon meets 100% renewable energy goal 7 years early"[1]
[1] https://www.aboutamazon.com/news/sustainability/amazon-renew...
Google already claims to have matched 100% of their energy with renewables since 2017.
Clean energy isn't even some virtuous thing, it's just no-brainer smart energy policy. The only reason for Google to be backpedaling on its pledge is political. It's in a backsliding petrostate and it has to play the political game.
No. That is not true
* Unless you are evil using school yard excuses is off brand. You do the right thing
* Carbon free energy is cheaper to produce these days, so the premis is wrong
This politics. Burn the planet, it is what Trump would do
If it's cheaper to produce, then there's no point in doing anything special, because it'll just take over via market forces. The "problem" with renewables is that the economics are fantastic in the marginal case, but if you need it to take over baseload when the sun is down, wind isn't blowing, etc., it's much more expensive because you need to build energy storage, which fossil fuel generators get for free. Going from 50% of power generation to 100% is going to be many multiples more expensive than 0% to 50%.
Until things like flow batteries become economical, in places without rivers you can dam for generation/pumped storage, it's probably not going to be feasible to go 0% carbon unless you just replace all of your baseload generation with nuclear or something.
It is taking over via market forces. New utility-scale solar+battery capacity is already cheaper than building a new coal plant.
The only reason it's not doing it more quickly is because of government intervention in the markets in the US to nerf it.
Another option for powering data centers is small modular nuclear - but again, China is far ahead on that with the helium-cooled pebble-bed model, although it really seems optimized for high-temp industrial process steam generation over electricity at present - so no, cheap safe nuclear power is not coming to the USA anytime soon.
Certainly, Trump's vitriol on wind power and solar power could mean a loss of government contract opportunities for AI-centric firms that promote such energy source - but still, the tariffs on China's monoocrystalline PV panels go back a decade and are just as supported by both political parties, because the major donors to both parties see renewable energy as a threat to their profit margins, be they on the executive or the shareholders side of the equation.
The kerosene lamp manufacturers don't want to be replaced with electric lightbulbs. This kind of monopolistic investor capitalism always stifles innovation and progress - all they see is the losses from collapsing demand for fossil fuels, and if the raw inputs are solar and wind instead of oil and gas, it's harder to collect rents.
Because the US should be doing it domestically, just like China is.
The incentive to do is partially based on tariffs, but the incentive hasn't spurred the private sector and supply chain maintainers into action, and a more comprehensive strategy would need to be done. There is no consensus for further subsidies from the national government to spur this production either.
However, really the answer to dumping is to exploit the dumper viciously, buy as much below market rate product of theirs as you can while they bleed money (especially a capital good like solar panels).
The US could be driving down the cost of energy nationwide with the dumped power generation equipment. Lower cost energy could then be turbocharging (nearly) every other industry in the country.
The only industry this harms is fossil fuels, and solar is being hamstrung to protect this legacy industry at the cost of all potential future industries.
Pretty much by definition, products that can be dumped are ones that can be manufactured with easily scaled production equipment at low margin and high volume. If you can dump your competitors out of business, so can someone else, to you.
And that's the way it generally happens (c.f. DRAM in the 80's). The "dumping" phase of a technology curve is succeeded not by a cartel with a tight grip on the previously competitive market, but by an extremely competitive market of producers all undercutting each other. Almost everyone wins, except the original high-margin producers. But they (c.f. Intel in the 90's) tend to move on to other products anyway. No one wants to make junk for a living.
Which makes very little sense, because the US solar panel industry is practically nonexistent. There are some companies that do last-stage assembly of solar panels in the US - probably to meet "made in USA" requirements - but the solar cells themselves are almost all made overseas.
In short, there's nothing there for China to dump on - the US is not a competitive threat to China in this sector. All indications are that Chinese solar panels are cheap because they're being manufactured in immense volumes to help meet China's own energy demands. The international sales are a bonus, not the primary goal.
And at the cost of the future, period. Climate change promises to be extremely onerous to future generations, from a purely economical POV.
this is just a massive self-own at this point
And it would be a lot cheaper to run those factories if solar power were cheap.
Tariffs leading to expanded US solar panel production might be one of the few things that would lead Trump to cancel tariffs. Solar is somehow offensive to him.
pointing out that the current efforts don't do that is all that needed to be said. your criticism of that comment, to me, proved that you understood the comment enough that one needs to look further into the motivations.
I'm not suggesting that you're one of those types, but it reminded me of that ludicrous line of reasoning.
Understanding that the mismatch exists, why it exists and what can be done to steer things in the desired direction is something I can get behind though!
What evidence you have to justify your accusation?
From another poster here
Now, in their 2025 Environmental Report at https://sustainability.google/reports/google-2025-environmen...:
"We aim to run on 24/7 carbon-free energy on every grid where we operate by 2030." and "We aim to reduce absolute, combined scope 1, 2 (market-based), and 3 emissions by 50% from a 2019 base year by 2030." (They also restate the thing about "to neutralize our remaining emissions" a bit later in the document.)
We also all saw Sundar being paraded around like a prisoner of war at the inauguration (to be generous, another interpretation being that he is an enthusiastic collaborator in a despotic regime). He's clearly involved politically in this regime, and they have made their stance on clean energy clear. To be willfully blind to that reality is simply a choice I'm not making.
Which is good, because carbon offset are a scam.
Now, if you are a fossil fuel megacorp and want to burn the entire cap and trade system to the ground, creating a 'carbon offsets are a scam' meme and destroying the 'trade' side is a great way to manufacture consent to get rid of that pesky 'cap.'
The problem is that what gets exchanged in the marked is a certificate but the purpose of the market is to create a positive externality. This means the buyers and sellers don't have an inventive to be honest.
The buyer of the carbon credits doesn't actually need the carbon to be captured. They just want a certificate for X credits, so they can emit elsewhere or get some other benefit.
The seller doesn't actually need to capture the carbon. As long as they can make a convincing enough case to the buyer that they did capture the carbon, the buyer is happy to buy.
This is unlike a typical market where the seller does have an incentive to fool the buyers into a buying a subpar product, but the buyer has a lot of economic incentive to actually not be fooled.
https://ww2.arb.ca.gov/resources/documents/mrr-enforcement
The bigger issue, as mentioned above, is that there's often a time lag between when the seller receives the money and when the seller can actually put it to use to reduce carbon emissions. One of the biggest sellers of carbon credits, for instance, is CA high-speed rail, which is decades away from completion. If it doesn't actually complete, it's not going to take any cars off the road or planes out of the sky, and so all the carbon credits it sold would just allow fossil fuel emitters to maintain status-quo emissions.
But as a way of diverting private resources from CO2 emitters to greener alternatives, cap & trade has been pretty effective. Over half the cars in my Bay Area city are now EVs; Tesla was kept afloat for many years by selling carbon credits.
Often data centers need to run at three nines of availability (ie: 99.9) and at that level of availability the price of solar power is many multiples of conventional power due to the cost of batteries and over provisioning.
But your point otherwise stands; solar + battery is cheaper than any fossil fuel form of power.
In the case of geo-politics if the US wanted to install solar to cover its energy load it would essentially be a massive payment to China for the solar arrays since there isn't productive capacity in the US. It would also end up being a long term dependent on future purchases of solar from China when they would need to re-up every 30-40 years at end of asset life.
30-40 years of cheap, CO2 free energy sounds pretty good to me.
Edit: US has no competitive way to manufacture panels, and with cheap imports, it never develops such industry!
We should look to non Chinese suppliers strategically, but they can't remote shutdown panels in an instant like Russia can stop a flow of gas.
Ther rated lifespan for a solar panel is on the order of 20-30 years, but they're still producing 70-80% of their output after that time. They last a very long time - if not destroyed from above.
Unless war compels it. But there's no Nordstream of solar panels that can be shut down in the meanwhile.
Greed, greed, greed. They'd rather see the world burn than not be first in AI whatever that means. They could still achieve net zero, they have just chosen not to. Unimaginable amounts of money but it's never enough.
Can anyone shed any light on the difference between net-zero and what they will actually be able to achieve? Will be hard given that the workloads for AI have sucked up all available energy resources. I have to think at the margin it is getting very expensive to acquire those resources.
That would be more compelling than to simply claim to have a lot of knowledge of history
Today in America it is hard to get more than half the population to agree on anything. There is no unifying catalyst to bring about the consensus, cooperation, and personal sacrifice that would be necessary to implement a centrally planned economy in the United States today.
Really, what is the utility of a rent seeker? By definition they are a leach on the actual productive elements of an economy. A waste. An inefficiency. I mean, you pay someone to do something, but you have to pay them even more than what might be required because they have this nonworking person sustaining themselves on rent seeking you also need to pay stuck on that back of that worker. That is the situation in so many economic sectors. It leads to a huge loss of money and productivity objectively. It leads to businesses outsourcing labor due to the artificial costs brought on by speculative rent seekers in a given local economy.
My landlord has no job. They collect probably 16k in rent a month from all the units in the small complex in which they raise rents yearly. In terms of investment in the property they once hired a handyman to cauk my sink so probably not even $200 a unit a year in overhead on average. They don’t do renovations or any work between tenants beyond sweeping up the floors and they try and keep as much security deposit for themselves as possible for stupid things. I pay a substantial portion of my income not to upkeep the roof over my head, but to ensure my landlord can have an upper class existence in a high cost of living area while not having to work a job and contribute any labor or ideas.
Does that seem like a good system to you? Enabling freeloaders to leech off the people doing the actual work that makes the world spin, to a point where they are paid far more than most workers? Passing that lifestyle to their nonworking children through inheritance? Essentially creating a hidden nonworking feudal kingdom sustained by the functional economy? One that has outsized political and media influence to ensure the boat is not rocked? Really think about what is actually happening in our economy.
When they seize these properties, how should the government determine who gets to live where? Should it be based on your profession? Your family size? Your standing within the Party, Comrade? Perhaps your ability to bribe the government agent who is responsible for assigning housing? What if there’s not enough housing for everyone and the government is behind on building more? How many random strangers are you okay with sharing your apartment with?
Resources are fundamentally finite and unequally distributed, even under communism.
Another example of communism that isn’t some impoverished corrupt nation economically isolated from global trade through US led embargo: on base housing at your local military base.
Military on base housing has been privatized since the 90s. You are provided a housing allowance as part of your pay that is then paid to corporate landlords.
https://en.m.wikipedia.org/wiki/Military_housing_privatizati...
In fact, quite the opposite: most of the poverty reduction in the last 50 years has been in China, for example. Most of the cheap stuff we buy is manufactured there. Being the "factory of the world" doesn't seem like a definitive invalidation of that system.
> has resulted in big changes
The change was allowing market forces to align incentives.
Before capitalism and communism was a thing, we had normal economics. And the worst part of this in some places was the bourgeoisie. They kept the resources in their control. And communism was supposed to be a banging of the piñata that shook the tree. Capitalism was a response to communism, yet installed a basic transition to a socialist education system and for educational wealth to be added.
I guess a good antidote to bad actors is for a wealth tax. Unused or underutilized wealth gets taxed until the net tax is reduced to whatever financial return it has.
Sure, resource allocation can be a instrumental to those goals of human well being, but not the directly optimized metric. So, for example, I care very little gdp/capita when you can just as easily measure life expectancy or food stress levels.
On such counts, capitalism has a mixed record. Today, for example, a lot of countries that are "more capitalistic" than others have poorer metrics of human development than those that are "less capitalistic".
That requires resource allocation as a prerequisite and a functioning democracy that values this metric. Blaming the first one for not having the latter is a choice, but a strange one.
People decide on resource allocation, period. It is as "efficient" as choosing human laws versus natural selection, the latter being exactly the model capitalism is copied from.
This is a run-away positive feedback loop which is kept in check by regular revolt and redistribution of wealth (see eg. the labor movement).
Another example of the latter, is that large companies should always be majority owned by their employees in a 1/N fashion. The idea being that even if such players try to rent-seek, at least the rent is being extracted in favor of normal people rather than a small class of rich people.
Not that I advocate nature red in tooth and claw as the epitome of governance.
And they forced ai on top of search because they wanted to.
This is a thread about them not being able to meet carbon neutrality commitments because of AI, which they feel they need to because competitors are eating their lunch.
Are the competitors not eating their lunch? Are they doing so because they're more green? Otherwise what are they expected to do? Not compete? “Innovate” themselves out of not needing energy consuming AI?
AI is not making anyone money except Nvidia.
As in, fiduciary duty does not prevent companies from being net-zero nor does it mandate AI.
Where there is a layer of abstraction involving money, there is always bullshit.
EX: zero death from Covid, oh, we fck the next generation with isolation and screen.
I singular objective like that make you blind, it become a god you sacrifice everything else to.
I know what response to expect for the above. It will be contemptuously dismissed with the remark that it's just a fantasy. But I hope you're seeing the congress people running out of town halls via the backdoors in disgrace after their constituents heckled them and challenged their demagoguery. Sadly, it's a tad bit late. This should have been the response to much less provocation. Still, many of those politicians are now afraid to blindly endorse the ideas of their overlords. This concept isn't new either. This was widely experimented with during the French revolution. They had provisions to recall and replace the reps in such instances. What people need are political education and a strong allegiance to the constitution.