Artificial Intelligence Has a Climate Problem, But So Does All Technology

We talked a lot about artificial intelligence Decoder lately; it’s inevitable. But there’s one piece of feedback we’ve gotten that I really wanted to spend some time on: how the meteoric rise of AI tools is affecting the climate.

After all, to run AI at scale, we need to build a lot of data centers and fill them to the brim with power-hungry GPUs. That takes a lot of energy, and is it worth using up all that juice? appears often when we talk about AI. It’s both a matter of practical concern — “can our aging grid support all this?” — and a moral objection — “we shouldn’t build these systems because they’re going to destroy the planet.”

What’s particularly complicated is that gigantic tech companies like Amazon, Google, and Microsoft have spent the last few years working with governments around the world to set ambitious sustainable energy goals so we can snail-paced the rate of climate change to just “bad” instead of “catastrophic.” But now, thanks to AI, all of these companies are exceeding their emissions goals and actually achieving worse over time. This is not great.

But putting a lot of computers in a data center and running them at full capacity is basically All works now. If you have moral objections to AI based on climate concerns, you might have moral objections to TikTok and YouTube, which are constantly ingesting and encoding millions of hours of video. You might have moral objections to video games, which both run on power-hungry graphics processors in people’s homes and often require intensive data center workloads for online multiplayer. And I’m going to guess, but I’m pretty sure that anyone who has climate concerns about AI also has a pretty harsh assessment of crypto.

Let’s think about it this way: The Nvidia H100, which is the gold standard for AI GPUs, is quite similar to the gaming-oriented Nvidia RTX 4090 in terms of power consumption. What framework should we apply to assess the impact of these cards on the climate, and how feel how they are used?

It’s messy and complicated, and there are a lot of apparent contradictions along the way. So it’s perfect for DecoderTo facilitate sort this out, I invited Edge Senior Science Reporter Justine Calma on the show to see if we can untangle this knot. Let us know how we get on.