Dupe from just a couple of hours ago, which quickly fell off the frontpage?

https://news.ycombinator.com/item?id=46389444

397 points 9 hours ago | 349 comments

▲ cm2012 11 hours ago | parent | next [-]

Interestingly there was no push back in the prior thread on Rob's environmental claims. This leads me to believe most HNers took them at face value.

▲ worik 6 hours ago | parent [-]

Umm... are they not correct?

The energy demands of existing and planned data centres are quite alarming

The enormous quantity of quickly deprecating hardware is freaking out finance people, the waste aspect of that is alarming too.

What is your "push back"?

▲ cm2012 4 hours ago | parent [-]

Happy to provide. I will say that literally all these sources are already available in this HN thread, but its hard to find and many of the comments are down voted. So here you go:

This link has a great overview of why generative AI is not really a big deal in environmental terms: https://andymasley.substack.com/p/a-cheat-sheet-for-conversa...

GenAI is dramatically lower impact on the environment than, say, streaming video is. But you don't see anywhere near the level of environmental vitriol for streaming video as for AI, which is much less costly.

The European average is 56 grams of CO2 emissions per hour of video streaming. For comparison: 100 meters to drive causes 22 grams of CO2.

https://www.ndc-garbe.com/data-center-how-much-energy-does-a...

80 percent of the electricity consumption on the Internet is caused by streaming services

Telekom needs the equivalent of 91 watts for a gigabyte of data transmission.

An hour of video streaming needs more than three times more energy than a HD stream in 4K quality, according to the Borderstep Institute. On a 65-inch TV, it causes 610 grams of CO2 per hour.

https://www.handelsblatt.com/unternehmen/it-medien/netflix-d...

Here is another helpful link with calculations going over similar things: https://nationalcentreforai.jiscinvolve.org/wp/2025/05/02/ar...

▲ socialcommenter 2 hours ago | parent | next [-]

I won't ask about or speak to the overall message of your first link, I'm interested to digest it further for my own benefit. This is striking to me though:

    Throughout this post I’ll assume the average ChatGPT query uses 0.3 Wh of energy, about the same as a Google search used in 2009.

Obviously that's roughly one kilowatt for one second. I distinctly recall Google proudly proclaiming at the bottom of the page that its search took only x milliseconds. Was I using tens-hundreds of kW every time I searched something? Or did most of the energy usage come during indexing/scraping? Or is there another explanation?

▲ worik 2 hours ago | parent | prev [-]

I think you miss the point, I looked at that Substack, it is way off point

It is the training of models, is it not, that requires huge quantities of electricity. Already driving up prices for consumers.

OpenAI (that name is Orwellian) wants 25GW over five years, if memory serves. That is not for powering ChatGPT queries

Also the huge waste of gazillion of dollars spent on computer gear (in data centres) that will probably depreciate to zero in less than five years.

This is a useful technology, but a whole lot of greed heads are riding it to their doom. I hope they do not take us on their ride

	▲	cm2012 2 hours ago \| parent [-]
		Its still a small portion of the total energy use and already powering a lot of things. And to compare other uses accurately, you'd also have to consider the cost of creating those things (like creating a full tv show, or car manufacturing, etc).

▲ steveklabnik 10 hours ago | parent | prev [-]

> 397 points, 349 comments

Probably hit the flamewar filter.