Comment by Schiendelman

13 hours ago

I can do things at least 10 times as fast as coding myself now. I'm pretty sure the environmental impact is, frankly, a reduction. My home computer (and feeding myself) for 10 times the time is more than I'm burning in a data center using an LLM.

Indeed.

I often see comments on the ethics and water/energy consumption of AI, but very few for non-vegan* developers boiling water for their coffees in air conditioned offices that require a commute, which I assume is a common default.

*blue-water usage of a burger patty is worth looking up, I was astonished

  • I assume you mean car commute, right?

    I don't know about the US but in France (or Europe) the environmental impact of air conditioning, car driving and meat consumption is well understood by most educated people.

I'm not sure I understand. You realise that our token consumption in coming on top of our other energy/water consumption right?

Also, the environmental impact might be small if we run a small number of inference queries every day, but that impact will grow linearly with the number of queries and agents we run.

  • If I spend less time working, which has been the big outcome here so far, I get more life. My own energy usage when I'm not working is mine, not part of the work.