Damn the AI Torpedos

Andrew Paul writing for Popular Science:

Amid the multiple waves of dramatic layoffs that have roiled Big Tech in recent months, Microsoft first announced plans in January to ax approximately 10,000 jobs from its global workforce by March 2023. The major cutback occurred as a new “AI arms race” kicked off between companies including Microsoft, Google, and Meta. All three and others are rushing to deliver on their lofty promises of chat programs, text generators, and revolutionary online search aids to consumers. Industry observers continue to urge caution against wantonly unleashing hastily tested, frequently problematic generative AI software.

This is just one of several posts this morning reflecting on the ethics of artificial intelligence. Robin has a great one. So does Ethan. As well as Pixel Envy.

The AI climate feels like some sort of dueling storm systems. Blowing in from the West is lightning in a bottle, this shiny new technology that’s both certain and designed to disrupt markets on a size no scale that none of us has seen in tech since “Web 2.0” stickers found their way on laptop cases. Or gosh, this AI lightning could very well be the most disruptive thing Millennials have seen in their lifetime!

Coming in from the East is an icy weather pattern, a chilling warning about the potential implications of the natural language processing powers of ChatGPT and the like.

Now, I’m no weatherman, but my Tech Doppler 5000 Pro Max Plus tells me these opposing storms will have to converge at some point. They have to, right? At least that’s what I hope. The idea that businesses are already waging an “AI arms race”… that one those very companies, Microsoft, can invest $11 billion into OpenAI while laying off the folks responsible for keeping AI, um, responsible… that real (read: not generated) people are already losing jobs to this tooling… and that all of this is happening in an environment that has little-to-no oversight and regulation… my grandma would have said it best: it just ain’t right.

In addition to not being a weatherman, I’m also no bleeding heart liberal who believes everything ought to be subject to regulatory bodies and systems. But I do think there’s got to be a line somewhere where the implications of a new thing are too great to ignore and warrant more, I don’t know, thought?

I remember the day we drove our first daughter home from the hospital and the gnawing feeling that I was totally unsuited to be a dad. I did what I could to prepare myself for the day it happened, but it blew my mind that we were bringing home a living, breathing human that we would be (and still are) responsible to raise into a delightful member of society. There were no classes to take. No manual or instructions. But I sure did get a lot of feedback from folks about the potential challenges in front of us — many of which were true!

Artificial intelligence feels a little like that to me. This baby of a thing has been in the works for some time — and we’ve all handled some form of AI, like voice assistants and smart speakers — but this is different. And now it’s in the backseat of the car, kicking and screaming to get out as we bring it home. But it’s being treated like it’s a mature technology with clear-cut use cases that are designed to make everyone’s lives better, fitter, and happier.

I thought businesses are supposed to be smart and know that a good business plan requires good objectives. Who knows? Maybe there’s something in the water that’s made everyone working with artificial intelligence forget that. Invest! Integrate! Ship! Faster! Sure, but what are you trying to solve and how is that going to happen? I don’t think we’re anywhere near knowing that when it comes to all this.

But, hey, the tech sure is shiny, right? Maybe too shiny. It’s blinding the path forward, but damn the torpedos.