I tried get out of this task so many times, in so many different ways.
Not every package requires a letter from the editor, I told them. I’ve been super busy recording a up-to-date podcast, preparing to speak at a tech conference, eating and sleeping, raising kids, doodling, going through my to-do list, and re-tying my shoelaces. I did my best, I tried to convey it to my editor. To be fair, my method of communication was subtle; I simply stopped responding to his emails, imagining that I could exist on a spectral plane where artificial intelligence wasn’t being shoved down my throat every minute of every day.
Like many men before him, he didn’t understand it. “Don’t think of it as a letter from the editor,” he wrote. “Think of it as a short article on the AI package! I know you’re not interested in WIRED having an AI manifesto.”
He’s right. I’m not. Here’s why: In July 2023, when I interviewed for this job, every person involved in the process asked me how I would get into AI. Since then, I have been bombarded – I was drowning, water rushing into my nostrils and flooding my lungs – with questions about my perspective on this technology. Communications and PR professionals would like to know this. The organizer of every major conference taking place anywhere on Earth is curious. My dad has questions. Likewise David Remnick. Every other journalist asked about it too, which is another reason on the very long list of reasons why I can’t go to any industry events, now or ever. My neighbor? He wonders about the AI apocalypse. My dermatologist? He has a up-to-date AI diagnostic tool that he’d be joyful to put my two cents into. My dogs? GPT-interesting.
Fortunately, there are living creatures who don’t want my opinion on artificial intelligence. My husband is too busy creating AI-generated videos on his phone to care what his wife, who constantly yells at him to put the phone down, thinks about AI in the context of her job. And the folks at WIRED, thank God, implicitly understand what I do about artificial intelligence: the technology has existed, been developed, and has been evolving for decades; this particular moment of AI fanaticism refers specifically to the deep learning approach of training AI models on massive amounts of data. Yes, it may seem technically intimidating. But it’s actually not that deep. If you’re not focusing on the intoxicating stream of hype being spread by visionary tech marketers like Sam Altman and Dario Amodei, then why are you even reading this? To the bunker!
The fact is that artificial intelligence is not a up-to-date invention. It is not the mythical solution to all problems, nor is it the great destroyer of worlds (and jobs) that we have been warned about. Training, implementing, and commercializing, and then lathering and rinsing with more data, up-to-date models, more promises, up-to-date warnings, and so on, is incredibly high-priced and resource-intensive. Amen, until we all stop thinking about quantum in a few years. Generative AI is really useful in some contexts, completely useless in others, and definitely unproven in most. This depends on the situation. It’s not a monolith. More and more signs indicate that we are dealing with an economic bubble. The bursting of a bubble can be temporarily disastrous; fascinating and vital aspects of AI technology will exist and will continue to shape parts of our lives. It is “there”, there. But let’s stop trying to do it via email. We should all just write our own emails. Yep, this is the “take” you’ve all been waiting for.
And journalism? Is generative AI poised to disrupt our industry by desecrating the craft of human-led news gathering and storytelling? It depends on who you ask. If you ask Google, they will assure you that sending traffic and revenue to publishers to spread correct information has never been more vital. They will then introduce AI-powered features to the detriment of their search engine and the publications they have embroiled in a decades-long abusive relationship. At this point, for publishers, getting away from Google means either cutting your hair dramatically or shutting down completely. Meta and Mark Zuckerberg have a different opinion. After pulling rug after rug out from under the news business in recent years, financially beating publishers around the world, and turning Facebook into a Shrimp Jesus convention for your Republican aunts and Threads into an apolitical social media network I haven’t thought about in a year, the company has created “a channel of expressive AI-generated videos from artists and creators.” Basically loose. Total shit, totally. And he can’t do it half as well as OpenAI’s Sora.
At WIRED, we will sometimes operate artificial intelligence. In the coming years, artificial intelligence may aid us in text editing. We already operate it for (carefully fact-checked) research and brainstorming with the automated equivalent of a rather stupid intern. Of course, machine learning technology that predates this point has always been useful in investigative reporting. But discovering and communicating up-to-date, newsworthy information? You come across a phrase that says exactly what you want, so damn beautifully? An illustration that makes you want to ~*kiss the chef*~ at a team meeting because you can’t wait for the issue to go to print and the world (at least the world of people who still subscribe to print) to see what you’re doing? It is human work, done by people and for them. Using technology to improve this work where it makes sense and avoiding it where it doesn’t. That’s what we do and will continue to do.
My advice to WIRED readers is to keep an open mind. You should be learning about artificial intelligence now and how to operate some of the tools you hear about all the time. Get the basics and expand on them as you find it intriguing or useful. Whatever you do, don’t make friends with an AI or have cybersex with it. For God’s sake, we all have enough problems already. Make sure your children still have human teachers. And then please relax. The world is constantly changing; the technology didn’t start with ChatGPT. The worst thing about AI may be that we can’t stop talking about it.
I could go on, but I really have to go. As I said: not every package requires a letter from the editor, and I have a mole on my back and I need a human opinion.