The Invisible Cloud: 44 AI images cost one phone charge

The Invisible Cloud comes at a moment when AI's resource use is moving from a niche topic to a mainstream one. It is built around a specific gap: training runs like GPT-4's, estimated at 50 to 70 GWh of electricity and around 25,000 tonnes of CO2e, capture headlines, but the UNU-INWEH report it cites says the inference phase that answers everyday prompts is an estimated 80 to 90% of AI's total energy use. ChatGPT alone is estimated to process around 2.5 billion prompts a day, and a typical AI text query is roughly 200 times more energy-intensive than a simple spam filter.

The site's framing leans on comparisons rather than raw numbers: 44 generated images equal a phone charge, a minute of AI film can match the emissions of driving to the local shop, and a long reasoning session evaporates a shot glass of water. It extends that logic to infrastructure, noting that global data centers consumed an estimated 448 TWh in 2025, enough to rank 11th if data centers were their own country, with a carbon footprint of around 189 million tonnes of CO2e, and that e-waste from AI hardware could reach up to 2.5 million metric tons a year by 2030.

Whether the calculator changes behaviour is likely to hinge on how many people find the personal numbers persuasive enough to act on. The site explicitly says it does not ask anyone to stop using AI, only to make the cost visible — a smaller ask, but one whose payoff depends on whether visibility alone shifts habits like editing instead of restarting, or matching the tool to the job.