My thoughts on AI
Generative artificial intelligence is being shoved into more and more workflows, products and services, despite constant proven failures at every turn.
The majority of job descriptions I come across these days at least mention AI, and I typically don't bother with companies that are all in on the fad. But plenty of companies that aren't AI-first still approach some tasks with AI assistance, and I guess I can't fault them for that.
Still, I need to get this off my chest. So here's my beef with generative AI, and why I do everything I can to keep my work human-centered.
It's terrible for the environment.
There may be a day when we can run AI prompts on local machines without offloading the cycles onto giant overheating data farms, but until that day, my main issue with AI will always be the environmental costs.
It relies on stolen work.
This is another one that might not always be true, and you can certainly train data on your own labor, but that isn't what's happening 99% of the time, especially if you're relying on a company to handle the output for you. This is especially true for visual art and writing.
It's making us dumber.
This point's pretty hard to argue: relying on AI is making users demonstrably less able to think creatively and critically. Without my creativity, empathy and ability to transform observations and data into stories, I wouldn't be the person I am today. There's no way I'm handing over those abilities to anyone, much less a faceless robot.
It's wrong most of the time.
There are plenty of statistics proving that AI gets things wrong more than half the time, and the handful of times I've been forced to ask a robot for information, it's always wrong. Turns out my web host is using stolen instructions from a different web host, and confidently serves that information to me whenever I ask it a question. It's hard to trust anything with a success rate that low. And because it's incorrect so often…
It's dangerous.
We've got people letting AI control their vehicles, determine whether a mushroom in the forest is safe to eat, and decide whose medical insurance claims to honor. None of this is a good idea. I got into UX research and design to understand and help people, and handing the reins over to a thieving, hallucinating, uncaring robot isn't the way we accomplish any of that.
Those are my main concerns with the technology, and I sincerely hope we at least address the first two as soon as possible, because if we're going to let our apathy destroy ourselves, we should at least stop hurting others and the planet in the process.
It isn't profitable.
As someone who cares about the needs, goals, comfort and wellbeing of human users, I put a bit less stock in the profitability of a business than whose whose job is literally making sure a company stays in the black. But AI hasn't been making anyone money except the companies that sell the hardware behind it. It's a wildly speculative technology that has yet to see positive results, as much as they desperately try to sell us on the benefits.