As a business owner committed to doing the right thing — particularly when it comes to the environment — I was alarmed to learn that creating a single AI-generated image consumes the equivalent energy of charging four iPhones from scratch. With this in mind, we’ve decided to limit the use of AI tools in our team to exceptional circumstances.
Over the past few weeks, I’ve noticed how pervasive AI-generated images have become. A quick scroll through my LinkedIn feed before writing this revealed 22 posts featuring AI-created visuals. That’s a staggering 1,500 kilowatt-hours – just from one feed. Extrapolate that across platforms and users, and the energy consumption becomes astronomical.
This isn’t just a lot of energy — it’s an alarming amount. A recent TechCrunch report warns that data centres may run out of available power in just two years. Yet, despite AI dominating daily conversations, the regulatory focus seems to overlook this crucial issue. Topics like copyright, misinformation, safety, job protection and operational efficiency dominate the discourse, but the environmental toll rarely gets a mention.
When environmental concerns are raised, they’re often dismissed with the argument that AI’s potential to solve sustainability challenges justifies its energy demands. Does this reasoning really hold up under scrutiny?
At this month’s Web Summit in Lisbon, several speakers highlighted the shift of AI development to the US, where regulations are more lenient compared to the increasingly stringent policies in the EU. What the future holds under the incoming US administration is uncertain, but with influential figures like Elon Musk involved, a strategy of “less regulation, more development” seems plausible.
However, this brings us back to the pressing issue of powering AI. The environmental impact is a challenge the tech industry cannot afford to ignore or defer. Smart businesses creating AI solutions must take collective responsibility and proactive measures to address this. If they don’t, we risk reaching a point where the call to “go nuclear” on AI could become more literal than figurative — after all, if you really want to control any tech, just pull the plug.
AI has the power to transform industries and solve some of the world’s toughest problems, but that transformation cannot come at the planet’s expense. If we’re building the future of AI on borrowed resources, we must ensure we’re not leaving the next generation to pay the bill.
No comment yet, add your voice below!