Words by Pierre Ortalo, Yuzu Pulse

As creatives exploring AI’s possibilities, we’ve been considering its environmental impact lately. Those of us in creative fields tend to be conscious about environmental footprints – and AI has quite the story to tell on that front.

The Hidden Cost of Our Digital Tools

Here’s something that’s pretty mind-blowing: GPT, that AI model everyone’s talking about? Training it just once produces as much carbon as 123 cars running for a year – that’s 552 metric tons of CO2 (1)! For anyone who cares about reducing carbon footprints, this number is significant.

And it gets more intense. By 2027 (that’s just around the corner), AI could be gobbling up between 85 and 134 terawatt-hours of energy annually. To put that in perspective, that is as much as the yearly energy consumption of the Netherlands (2).

The Thirsty Giants Behind the Scenes

Here’s another startling fact: those massive data centres powering our AI tools? In 2022, just Google and Meta used up 2.2 billion cubic meters of water to keep their systems cool (3). That’s as much as all of Denmark uses in a year! For those of us who advocate for responsible resource usage, this raises some serious questions.

And then there’s e-waste. We’re looking at potentially 120 million tons by 2050. It’s a scale that’s hard to even comprehend (4).

The Creative Conundrum

This directly affects the creative community. Every AI-generated image contributes to a carbon footprint. According to a study by researchers at the AI startup Hugging Face and Carnegie Mellon University (5), generating a single image using AI requires about the same energy as fully charging a smartphone. Text generation is a bit lighter on the environment.

Finding the Balance

No one’s suggesting we should abandon AI – it’s revolutionising creative fields in remarkable ways. But perhaps we need to be more mindful? The AI Act is trying to help with voluntary codes of conduct and standardisation, though these voluntary measures might not be enough.

Where Do We Go From Here?

As members of the creative community in this digital age, we’re in a unique position. We can advocate for more sustainable AI practices while still embracing its creative potential. Maybe it’s about being more selective with AI usage, supporting companies that use renewable energy, or raising awareness about these issues.

Also, some argue that the process of making an image by a human is more carbon emitting than by an ai. A claim supported by an article on Nature.com (6). However, in science, one article isn’t truth, it’s the consensus built over many researches that takes us closer to truth.

The numbers speak for themselves – they’re pretty stark. But surely we can find a way to harness AI’s power while keeping our environmental conscience clear. After all, isn’t creative problem-solving what the artistic community does best?