i prompt, therefore i am: how tech forgot about human agency
Tone deaf tech bros no longer seem to understand that their pitch for AI is fundamentally dystopian and dismissive.
Culture writer Sam Kriss visited Silicon Valley to check on its AI startups and wrote a lot of words about it for Harper's Magazine. In those words, I think he captured what seems to be the main objection to how modern AI tools are being pitched to casual users who aren't indoctrinated into the Temple of the Singularity. He glossed over it so quickly that it's all too easy to miss, but it's there.
It's not that it can generate a lot of content. Spam is really annoying, but people are used to spam at this point, so much so that the Dead Internet Theory has become a joke that's not funny anymore.
It's not that it can corrupt people's mental health, because let's be honest here, the healthcare system was already failing them and they're just hitting rock button with automated help.
It's not even that it could automate some people's jobs, although that's a big part of it. People already know about AI-washing and are frustrated that AI keeps trying to take over the wrong things roughly half the time it's deployed. They want to see a proper, comprehensive plan for what a post-industrial society looks like and secure a decent place in it, but that's a political issue more than a technical one.
No, the biggest problem is that so many tone-deaf Silicon Valley ads are asking us to outsource our thinking, our experiences, and our creativity to machines. Forget I think therefore I am. Prompting and waiting for instructions is the new thinking. And there's something so profoundly unnerving and dystopian about it, that many of us intuitively reject the very premise of this.
Need to check on an email exchange or a document? No, you don't! Gemini will give you all the important points so you don't have to waste time actually reading. It'll also write a reply for you, no need to do anything but click tab and send.
Need to write a letter? ChatGPT has you covered. Just give it a quick description of what you want. Anything from a letter to a book, just prompt, copy, paste, and don't worry your little human head about it.
Want to paint something or make a video? Why even bother when you have Sora and Stable Diffusion? You're really going to waste days or even weeks on something that can now be done in seconds from nothing more than a vague description? And if you want to post it on social media later, most people won't even care that it's AI.
Need to write some code or automate something? CoPilot to the rescue. Why even bother trying to solve any problems on your own if it's already seen every technical implementation of pretty much everything there is to see, even code it wasn't even supposed to know about? How could you ever beat it in technical competence? All you need to do is enter the prompt in agent mode, press apply, and you're done.
Out on a date and don't know what to say? In an interview and nervous? Pfft. You silly organ sac. Turn your sad, inefficient organic brain off, turn Cluely on, and follow what it says on the screen. Why trust in yourself when you could trust the power of tens of millions of servers running across the world?
A world where you merely need to exist and follow the outputs of whatever screen is closest to you at any given moment, and if you get stuck, you can just ask it what to do. Doesn't that sound amazing?
Well, no. It doesn't. Becoming little more than just a meat puppet with a credit card to charge for every token processed by a chatbot isn't a fate most of us look forward to. If anything, it's something you find in a dystopian cyberpunk novel, the kind tech bros love to read but can't seem to comprehend. Sorry, grok. At least the way the term was used before that's how the man who gave up on Mars because "AI is going to be woke now," named a Nazi chatbot and ruined it for the foreseeable future.
And yet, while yelling about making everything "agentic" and eliminating everyone's jobs with the power of super-AI any day now, Silicon Valley forgot to explain what's in this for the average person, and refuses to realize why so many potential customers and users would be unhappy with the currently implied arrangement.
Ultimately, introducing new technology, especially powerful new technology, is about respect. It's about figuring out the best way it fits into a user's life to enables the kind of future they actually want. Unfortunately, today's Silicon Valley and its AI bros simply do not respect their users, and their culture has drifted so far off normal that I'm not sure that they can figure this out anytime soon, even when it's spelled out to them…
