As the never-ending march of AI progress continues, with every big tech company embedding it deeply across product ranges, some of the implications of power consumption required to run prompts are becoming clearer.
Google has spent much of the last couple of months rolling Gemini out to everything it makes, and defending its AI Overviews which erroneously suggested users use glue on pizza or eat rocks. However, it seems the company’s big AI push has had big ramifications on its climate footprint targets,
Yesterday, Google confirmed its greenhouse gas emissions have climbed by 48% thanks to energy consumption from its data centers, and that its “extremely ambitious” goal of reaching net zero emissions by 2030 “won’t be easy”.
As reported by The Guardian, the International Energy Agency suggests Google’s electricity consumption could double from 2022’s levels, too, meaning this rise could be the tip of the iceberg.
Why Google’s emissions are rising
Google is investing heavily in tools, resources and infrastructure to minimize its emissions but that is coming up against the heavy energy use of training and running artificial intelligence models in data centers.
A study last year by AI startup Hugging Face found that generating an AI image uses as much energy as fully charging your smartphone, and the cost of initially training the model in terms of emissions is higher still.
Google has signed agreements with the likes of Reddit to help train the LLM (Large Language Model) that its Gemini Chatbot works from, while its DeepMind research arm continues to work on generative AI that can add audio to silent videos or summarize other AI presentations.
It’s not just Google, either. Microsoft continues to push for AI advancements at the risk of its own emissions targets, with OpenAI’s Sam Altman seemingly confirmed in May of this year that the company burned through $520 million in cash in 2023.
So there is no sign of the AI progress slowing down, nor the emissions generated. Let’s just hope that, as Bill Gates told Sky News last week, the payoff in terms of technology to tackle the climate crisis from investment in AI will balance things out in the end.
