When I read that AI is the new plastic—with its promise of shaping computers to the way we want them, just like plastic enabled innovations from Tupperware to AirPods—I can’t help but recall the shadow side of plastic. It became ubiquitous, cheap, and disposable, all while we ignored its environmental impact. Today, knowledge workers have similarly embraced large language models (LLMs), using them for everything from code to conversation. But what are the side-effects of this surge?
LLM inference alone now accounts for up to 20 % of global data-center electricity use, with emissions expected to rival those of entire small countries.
Plastic democratized appliances—but it couldn't replace craftsmanship. Likewise, AI isn’t a substitute for human creativity. It lacks nuance, embodied insight, and moral reasoning. Its affordances are unmatched in rapid iteration, but it’s blind to context and values.
As designers, we should treat AI as a material—a powerful, malleable medium with its own strengths and weaknesses. Use it to draft, provoke, prototype. But stay vigilant: audit the energy, verify biases, preserve your human voice. Like plastic before it, AI must be used with care.
In the long run, responsible AI stewardship could parallel plastic stewardship: recycling, transparency, renewable-powered infrastructure. Otherwise we risk repeating history—igniting a technological boom that grows cheap and wide but leaves hidden scars on our creativity and the planet.