Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Mistral reports on the environmental impact of LLMs

mistral.ai

Post reply on HN
Print view View original

21–30 of 59 posts

Re: Mistral reports on the environmental impact of LLMs

#21

Post by plants »

j-pb wrote: post #7
preciz wrote: post #3

Earlier quoted context omitted.

And each toilet flush you make should also have a Co2 calculation which should go against your daily carbon allowance.

people downvote your sarcasm, but if you do the calculations you're kinda right. 1Kg of Beef costs: - The energy equivalent of 60.000 ChatGPT queries. - The water equivalent of 50.000.000 ChatGPT queries. Applied to their metric Mistral Large 2 used: - The water equivalent of 18.8 Tons of Beef. - The CO2 equivalent of 204 Tons of Beef. France produces 3836 Tons of Beef per day, and one large LLM per 6 months. So yeah…

Those are incredible stats. As a vegan who uses LLMs at work frequently, I would love to have the source as well :)

Re: Mistral reports on the environmental impact of LLMs

#22

Post by j-pb »

plants wrote: post #21
j-pb wrote: post #7

Earlier quoted context omitted.

people downvote your sarcasm, but if you do the calculations you're kinda right. 1Kg of Beef costs: - The energy equivalent of 60.000 ChatGPT queries. - The water equivalent of 50.000.000 ChatGPT queries. Applied to their metric Mistral Large 2 used: - The water equivalent of 18.8 Tons of Beef. - The CO2 equivalent of 204 Tons of Beef. France produces 3836 Tons of Beef per day, and one large LLM per 6 months. So yeah…

Those are incredible stats. As a vegan who uses LLMs at work frequently, I would love to have the source as well :)

  - https://mistral.ai/news/our-contribution-to-a-global-environmental-standard-for-ai
  - https://www.statista.com/statistics/1201677/greenhouse-gas-emissions-of-major-food-products/
  - https://www.statista.com/chart/9483/how-thirsty-is-our-food/
  - https://blog.samaltman.com/the-gentle-singularity
  - https://x.com/sama/status/1890820962993533232
You got to do the calculations yourself though.

Re: Mistral reports on the environmental impact of LLMs

#23

Post by aziaziazi »

preciz wrote: post #3
greyadept wrote: post #2

I would really like it if an LLM tool would show me the power consumption and environmental impact of each request I’ve submitted.

And each toilet flush you make should also have a Co2 calculation which should go against your daily carbon allowance.

I don’t flush my toilet, I Kildwick [0] but J-pd has a more interesting comparaison

[0] https://www.kildwick.com/

[1] https://news.ycombinator.com/threads?id=j-pb

Re: Mistral reports on the environmental impact of LLMs

#24

Post by ACCount36 »

jeffbee wrote: post #17

These conclusions are broadly compatible with "The Carbon Footprint of Machine Learning Training Will Plateau, Then Shrink" or, as I prefer, the PDF metadata title that they left in there, "Revamped Happy CO2e Paper". Despite the incredible focus by the press on this topic, Mistral's lifecycle emissions in 18 months were less than the typical annual emissions of a single A320neo in commercial service. https://arxiv.o…

The press focus is a mix of the usual "new thing BAD", and the much more insidious PR work by fossil fuel megacorps.

Fossil fuel companies are damn good at PR, and they know well that they simply can't make themselves look good. The next best thing? Make someone else look worse.

If an Average Joe hears "a company that hurts the environment" and thinks OpenAI and not British Petroleum, that's a PR win.

Re: Mistral reports on the environmental impact of LLMs

#25

Post by jeffbee »

ACCount36 wrote: post #24
jeffbee wrote: post #17

These conclusions are broadly compatible with "The Carbon Footprint of Machine Learning Training Will Plateau, Then Shrink" or, as I prefer, the PDF metadata title that they left in there, "Revamped Happy CO2e Paper". Despite the incredible focus by the press on this topic, Mistral's lifecycle emissions in 18 months were less than the typical annual emissions of a single A320neo in commercial service. https://arxiv.o…

The press focus is a mix of the usual "new thing BAD", and the much more insidious PR work by fossil fuel megacorps. Fossil fuel companies are damn good at PR, and they know well that they simply can't make themselves look good. The next best thing? Make someone else look worse. If an Average Joe hears "a company that hurts the environment" and thinks OpenAI and not British Petroleum, that's a PR win.

I suspect the press is also aligned against machine learning because they are still Big Mad® that the internet destroyed their revenue model (charging individuals $50 to advertise used cars, for example).

Re: Mistral reports on the environmental impact of LLMs

#26

Post by mlnj »

j-pb wrote: post #9
bluefirebrand wrote: post #8

Earlier quoted context omitted.

The difference is that food is important and live-giving and LLMs are a very fancy magic 8-ball

You don't need beef, beef is a lifestyle choice. I use LLMs to do all of my coding these days, it's certainly more essential for feeding me than beef.

Nothing stopping us to live the hermit life in the mountains. But here we are trying to get bits and bytes to write that JIRA ticket instead of us.

Re: Mistral reports on the environmental impact of LLMs

#28

Post by motoxpro »

bluefirebrand wrote: post #8
j-pb wrote: post #7

Earlier quoted context omitted.

people downvote your sarcasm, but if you do the calculations you're kinda right. 1Kg of Beef costs: - The energy equivalent of 60.000 ChatGPT queries. - The water equivalent of 50.000.000 ChatGPT queries. Applied to their metric Mistral Large 2 used: - The water equivalent of 18.8 Tons of Beef. - The CO2 equivalent of 204 Tons of Beef. France produces 3836 Tons of Beef per day, and one large LLM per 6 months. So yeah…

The difference is that food is important and live-giving and LLMs are a very fancy magic 8-ball

You could eat a few bites less of beef in one meal and that would be your equivalent AI use for a lifetime.

Re: Mistral reports on the environmental impact of LLMs

#29

Post by JimDabell »

plants wrote: post #21
j-pb wrote: post #7

Earlier quoted context omitted.

people downvote your sarcasm, but if you do the calculations you're kinda right. 1Kg of Beef costs: - The energy equivalent of 60.000 ChatGPT queries. - The water equivalent of 50.000.000 ChatGPT queries. Applied to their metric Mistral Large 2 used: - The water equivalent of 18.8 Tons of Beef. - The CO2 equivalent of 204 Tons of Beef. France produces 3836 Tons of Beef per day, and one large LLM per 6 months. So yeah…

Those are incredible stats. As a vegan who uses LLMs at work frequently, I would love to have the source as well :)

This is a good article on the subject:

> Using ChatGPT is not bad for the environment

— https://andymasley.substack.com/p/individual-ai-use-is-not-b...

He’s done some good followup articles as well:

https://andymasley.substack.com/s/ai-and-the-environment

Re: Mistral reports on the environmental impact of LLMs

#30

Post by kingstnap »

greyadept wrote: post #2

I would really like it if an LLM tool would show me the power consumption and environmental impact of each request I’ve submitted.

You can assume the API price is roughly proportional to electricity usage.

If you buy $10 in tokens, that probably folds into ~$3 to $5 dollars in electricity.

Which would be around 30 to 90 kWhr in electricity.

Depending on the source, it could be anywhere from ~500g/kWhr (for natural gas) and ~24g/kWhr for hydroelectric.

It's a really wide spread, but I'd say for $10 in tokens, you'd probably be in the neighbourhood of 1 kg to 40 kg of emissions.

What's a good thing is that a lot of the spread comes from the electricity source. So if we can get all of these datacenters on clean energy sources it could change emissions by over an order of magnitude compared to gas turbines (like XAi uses).

Post reply on HN