I am extremely skeptical about this sort of thing - I suspect that it's extremely challenging to reliably make assertions about the carbon cost of running a given website given that there are so many unaccountable-for variables; at best the very "edge" of any carbon usage is being guessed at.
Added to that the opportunities for charging the various parties for "certification" are simply too good to pass up it seems.
Yes, but one does not have to reliably make assertions to feed a meme.
People want to be good. The current agenda God says you are good if you sacrifice to the CO2 god. No need to make reliable assertions to feel good about it. Anyone who challenges the act of sacrifice is someone to burn. No challenge means no problems.
Anyone who wants to see both sides of the Co2 discussion is already sceptical at minimum. But it does not make you feel good. Therefore its rare. It feels good to sacrifice to the Co2 god. “Lemme sacrifice you fool.”
This project doesn't seem to care whether I'm sending data to someone else right near us-east-1 or whether I'm sending data to Australia. The energy costs will be very very different.
Not only is it measuring bytes blindly, it's counting rating your energy use as a % of bytes sent, versus total data-center consumption period. As though every byte is equally responsible for the total computing done. Absurdly wrongheaded model.
I would be happy to kiss crypto goodbye and though AI might be stifled it would still be developed further just with a mind to efficiency. This is because it has actual real world uses unlike crypto.
I think CodeCarbon (the python library) might have a better method for estimating carbon usage. They seem to be digging it up on a per-power station basis. Would love to see the join forces.
Iron fertilization of ocean technique already had been proven to work to solve CO2 problem. It qas so successful Canadian government had to arrest the lead researchers at their office to discredit it. The whole idea of taxing CO2 is what gets us to today where CO2 continue to spike. The taxong only brings in money for the rich and governments to spend. It doesnt do anything to reduce CO2. Go check it out CO2 taxing started dwcades ago. And what is current worldwide government position on iron fertilization. You be shocked.
Idk why Wikipedia don't consider making dark mode the default and giving users an option of switching classic light mode manually. Figure that should make a significant dent for the effort involved. I remember putting it forward as a suggestion a few years ago.
It hardly makes a difference on LCD monitors. See also blackle
Edit (25 minutes later): a phone screen seems to use about 1W more on a white background than a very dark one from what I could quickly find on mobile (https://www.xda-developers.com/amoled-black-vs-gray-dark-mod...). 5 billion people seem to use phones for some 4.4h/day (https://backlinko.com/smartphone-usage-statistics). Assuming 70% uses OLED worldwide (probably much less, but I can't quickly find a source for this) and 50% of the time is spent on reading text (not tiktok videos or whatever, where you can't use dark theme), that's 2.8 TWh per year. About 178'000 TWh of energy are used by the world per year (https://ourworldindata.org/energy-production-consumption).
So this should save 0.002% in this best-case scenario where every text-based application switches to nearly black backgrounds and most people have OLED screens and people actually use their device for text at least as much as for media consumption.
Is that worth it? Maybe, but perhaps not as effective as not loading megabytes of JavaScript on every page, not downloading ads and tracking on every page, and funding low-carbon electricity generation needed for your systems. Wikipedia already is pretty good on the first two afaik, the latter I can't easily tell beyond saying in 2019 they're doing a "best effort" in buying green-labeled electricity and making an inventory of other emissions like plane travel (https://wikimediafoundation.org/news/2019/09/19/how-the-wiki...)
Added to that the opportunities for charging the various parties for "certification" are simply too good to pass up it seems.