The climate crisis has bigger problems than AI

We’ve been talking a lot about AI Decoder lately; it’s inevitable. But there’s one piece of feedback we’ve received that I’d really like to spend some time on: how the lightning explosion of AI tools affects climate.

After all, to run AI at scale, we need to build lots of data centers and fill them with power-hungry GPUs. This takes a lot of energy, and is it worth using all that juice? comes up frequently when we talk about AI. It is at the same time a question of practical concern – “can our aging network withstand all this?” — and a moral objection — “we shouldn’t build these systems because they would destroy the planet.”

What’s particularly tricky is that big tech companies like Amazon, Google and Microsoft have spent the last few years working with governments around the world to set ambitious goals around sustainable energy use so that we can slow the rate of climate change to simply “bad”. rather than “catastrophic”. But now, with AI, all of these companies are exceeding their emissions targets and are actually achieving worse over time. That’s not great.

But putting a bunch of computers in a data center and running them at full tilt is basically like all works now. If you have a moral objection to AI based on climate concerns, you might also have a moral objection to TikTok and YouTube, which constantly ingest and encode millions of hours of video. You might have a moral objection to video games, which run on power-hungry GPUs in people’s homes and often require intense data center workloads for online multiplayer. And I’m going to guess, but I’m pretty sure anyone with climate concerns about AI also has a pretty harsh assessment of cryptography.

I mean, think of it this way: the Nvidia H100, which is the gold standard for AI GPUs, is quite similar to the gaming-focused Nvidia RTX 4090 in terms of power consumption. What framework should we use to assess the climate impact of these cards and how feel how they are used?

It’s confusing and complicated, and there are a lot of apparent contradictions along the way. So it’s perfect for Decoder. To help resolve this, I invited Border senior science reporter Justine Calma on the show to see if we can untangle this knot. Let us know how we get on.