What's the Jevons paradox?
"the Jevons paradox occurs when technological advancements make a resource more efficient to use (thereby reducing the amount needed for a single application), however, as the cost of using the resource drops, overall demand increases causing total resource consumption to rise.
The original example posited by Mr. William Stanley Jevons, summarized nicely by Axios, was coal. Progress in steam engines, which enabled them to use less coal, didn’t lead to a drop in coal demand — it led to a huge rise."*
The Jevons paradox now applies to AI.
As AI tokens become less expensive, AI becomes more attractive to businesses.
New businesses will emerge that develop AI products (for consumer and B2B applications), making AI more easily accessible to more people.
Existing businesses will incorporate AI into their systems and processes, exposing more people to AI and its capabilities.
And all of this AI activity will lead to an exponential rise in total AI usage.
The energy, workforce, and overall societal implications of this are significant.
And also somewhat terrifying.
Will AI affect every business? Almost certainly.
But not all businesses will be affected in the same way.
I recently went to get my hair cut and was discussing AI with my barber, Joe.
(I've been going to the same barber shop since I was 14 years old; Joe's father used to cut my hair while Joe was learning his craft alongside him, so Joe and I have known each other for a very long time.)
Is a robot going to replace Joe? No, probably not anytime soon. And I told him as much.
But I also pointed out that while he was cutting my hair, Joe was interrupted three times by clients calling his phone to schedule appointments.
I said, "Joe, you don't want to spend your time on the phone booking appointments, you want to spend your time cutting and styling hair, right?"
He nodded in agreement.
"So what if my AI assistant could call your AI receptionist, find some time for me to come in that worked for the both of us and then put the appointment in our calendars... without you or me ever having to get involved in the process... do you think that would help you?"
That's when Joe realized that not even hair stylists will be unaffected by advances in AI.
For businesses like Joe's, AI isn't a threat, it's an enabler.
And Joe's business isn't unique.
Researchers may use AI to become more efficient in compiling facts and figures. AI still tends to "hallucinate" (read: make stuff up), so logical fact-checking will continue to be needed for the foreseeable future. But will this always be the case? Probably not.
Writers may use AI for inspiration and outlines. Bad writers may try to use it to produce finished works from mere prompts... but as we collectively get better at using AI, we should also get better at recognizing AI-generated works and that won't bode well for "writers" who don't actually want to write.
Programmers can use AI to write computer code. Non-programmers? They'll soon be able to use it to create the code they would otherwise have to pay a programmer to develop.
AI is clearly a threat to programmers, an irony most of us didn't anticipate.
Teachers? Lawyers? Engineers? Doctors?
I can't predict exactly how every profession will start to use AI.
But it's a safe bet that as AI becomes less expensive and more accessible, every profession will begin experimenting with it in some way to see how it can make things easier, better, faster, and cheaper. This will be a very good thing, right up until the point where it isn't.
The Jevon paradox as applied to AI will mean we are all using AI more.
But how will we sustainably produce the enormous energy that AI requires?
What will we do with the workers that AI will inevitably make redundant?
And how do we train AI models to act ethically, responsibly, and in our collective best interests to avoid creating a real-life version of SkyNet?
Perhaps these are the questions we can use AI to help us answer.
Ideally, before it's too late.
* Source (and the inspiration for this post): David Crowther, Sherwood.news
Comments