With all this talk of AI, we are entering a time of transition and change as our society becomes increasingly dependent on it.
It’s hard to ignore the practical uses of AI. We are seeing AI take redundant tasks and spit out results in a fraction of time. Simple, streamlined and quick. Results that sometimes required years of research, now available at our fingertips with a few prompts and queries. It increases productivity and is an enticing tool. However, as the digital climate changes, it threatens to eliminate or alter a vast number of jobs in the market, and editing is no exception.
As more people turn to AI to offset increased performance demands, we are seeing AI make decisions in spaces we never thought possible. Scraping resumes for keywords on who to hire, suggesting edits and writing AI-compiled content. The more we use it, the more it needs to expand. Hungry expansion.
The limitations of AI
Before I became an editor, I worked in the IT industry, sitting in super-cooled server rooms until wee hours of the morning. I got to see into a world that most people don’t even think about. I got to think about the energy usage of the server fans spinning all around me as I went about my tasks. One thing I learned is that computers have limits.
They are limited by programming biases and the types of information the developers ingest. They are limited by technology and by server cooling requirements. Computers also don’t account for social ethics and transparency, unless explicitly coded to do so. We are seeing these limitations at the forefront of discussion on the ethics of AI, and they are not solved by simply throwing more resources at them.
A power-hungry beast
AI prioritizes productivity. But productivity doesn’t take sustainability or humanity into account. It doesn’t factor in things like cooling requirements and power draw. It either has the required power and cooling, or it doesn’t.
For example, consider that one gigawatt can power approximately 750,000 homes for a year; global AI datacenters are expected to require 220 gigawatts by 2030. That’s equivalent to the energy required to power 165,000,000 homes a year. 220 gigawatts! Great Scott! It’s hard to ignore this power-hungry beast.
We need to challenge the line between productivity and sustainability. We must challenge ourselves to look not only at the results of AI, but also the consequences. AI is stripping the humanity out of our stories. We are also seeing increased demand on editors to add humanity back into AI- generated content, with jobs editing AI-generated content popping up on every platform.
It’s never satisfied
How much are we willing to feed AI and at what cost? In nêhiyâw (Cree) culture there is a name for an endlessly hungry winter spirit. It takes and takes and takes. It’s never satisfied, no matter how much it consumes. It causes starvation and famine. Job loss too.
I won’t say its name. I was taught that we don’t want to invite that kind of spirit into our lives by using its name. It could summon the spirit by drawing attention to its energy. It’s becoming increasingly apparent that we should be cautious in our relationship with AI. We must ask, do we need to replace all the tools in our editor toolbelt with AI, just because we can? When I consider nêhiyâw stories and I see the requirements of future AI, the parallels force me to ask, “Could AI be the hungry spirit my ancestors warned about?”
___
The Editors’ Weekly is the official blog of Editors Canada. Contact us.
Discover more from The Editors' Weekly
Subscribe to get the latest posts sent to your email.
Don Mailhot
Don Mailhot (he/him) is a Cree Mêtis editor, writer, storyteller and musician living in Treaty Six territory. He enjoys writing poetry and short stories. Don is passionate to explore his nêhiyaw (Cree) heritage through language, culture and writing and is currently penning his first memoir. Don is also an active member of the Indigenous Editors Association.