I Will Not Pay for AI (Premium)

Close your eyes and imagine it’s the 1990s. Microsoft has successfully transitioned its growing family of productivity applications–Word, Excel, PowerPoint, and so on–into a bundle it calls a suite that it sells for less than the cost of buying each separately. That’s an obvious strategy, but it only provides value if you need two or more of those products and if the price is right. But that’s OK because Microsoft is also improving each at a rapid clip, adding new features–many new features, the Microsoft of this era is obsessed with adding new features–and, crucially, making each app feel less like a standalone thing and more like part of an integrated whole. These, too, represent value, and as the competition mounts, Microsoft pushes down on the pedal, iterating faster and faster, integrating more and more, redesigning these apps not only so they look similar, with nearly identical toolbars and menus, but also work similarly at every level. We have reached peak Office.

OK, you don’t need to imagine that, it’s what happened. But now imagine an alternate history in which this didn’t just continue and lead to Office overtaking Windows as Microsoft’s most successful and lucrative product. Imagine instead that Microsoft one day announces something new, a premium add-on to Office that one must buy separately, and at great cost. This add-on, we’ll call it Office Sidekick, brings additional features to the apps in the suite, a few here and a few there, features like spell-checking, grammar-checking, and even extensibility through scripting and macros.

Office Sidekick costs three times as much as Office, and while Office had become a no-brainer by this point, an obvious choice, superior to its competition, and a considerable value, Office Sidekick is … different. Off. It feels like a money grab, a set of features that might have otherwise just been included in Office. Office, a product whose price has been coming down over the years to better compete with the alternatives that have since disappeared, having been destroyed by the successful monster from Redmond.

Office Sidekick isn’t a no-brainer. It’s anything but. Microsoft’s customers see it for what it is and while some corporations pilot it briefly to make sure they’re not missing something, most just give up on it. The value isn’t there.

This is what Microsoft is doing today, of course, with the modern Office, called Microsoft 365, and with Copilot for Microsoft 365 (and, for individuals, Microsoft Copilot Pro). It’s adding features outside the thing you’re already paying for, even though the reason you keep paying for it is to get new features over time as part of that monthly cost. It’s charging exorbitant prices for those things. And hoping we’ll collectively ignore the fact that this type of functionality used to be provided to us as part of the fee we already pay, every single month. Because today’s Office isn’t a one-and-done purchase–a perpetual license, as Microsoft calls it–today’s Office, Microsoft 365, is a thing we never stop paying for. And that’s not enough for Microsoft. It would like more from its customers. Every single month.

I’m not doing that. It’s not the 1990s anymore. Not only do we as consumers of technology have more and better choices, but we can also mix and match as we see fit. Those who pay for Microsoft 365 for the value they receive, as I do, can continue using its constituent parts as they wish, but they can get the functionality provided by Copilot from elsewhere. There are alternatives, too many to count, many of which are as good or better. And less expensive or even free. It’s a different world.

I focus largely on Microsoft, but this isn’t really about Microsoft. Microsoft is just an obvious example of this overreach, this enshittification. No, this is about all of Big Tech. And it’s time to put Big Tech on notice.

I will not pay for AI.

What I will pay for is products and services that use AI to provide real value. Good products and services.

That may seem like a subtle distinction. But it’s not: Microsoft 365 grew from a modern version of the classic Office bundle into an offering of such value that it became a no-brainer. But Copilot for Microsoft is a money grab, an overly expensive set of functional add-ons that should simply be included, at no additional cost, with the service I’m already paying for every single month. Instead, Microsoft charges a 200 or 300 percent markup for Copilot per month to consumers and business users, respectively, while providing only minor advantages spread too thin across the service’s offerings like too little butter across too much toast.

The problem for Microsoft and the advantage to its customers, who are already teetering on the edge of madness thanks to the incessant inroads of software we never stop paying for, is that there are alternatives. Those competitors it vanquished in the 1990s are long gone, and they’re not coming back. But they’ve been replaced by a legion of alternatives, most tiny, that Microsoft isn’t equipped to fight. The traditional wars of the past have been replaced by a new era of guerilla warfare. There are free alternatives and less expensive alternatives. Choices.

These choices, this “magic of software,” as Microsoft co-founder Bill Gates was so fond of saying, are what will set us free. As Microsoft, OpenAI, Google, and other Big Tech firms all tighten their grips on AI to continue and expand their dominance into this new era, more and more alternatives slip between their fingers eager for their customer base. Big Tech unleashed the AI genie, and not only can they not put it back in the bottle, but they can’t keep it for themselves either.

OK, I’m mixing idioms all over the place here. But that’s how you know it’s real. I didn’t use AI to write this, I used my tortured brain and my clumsy fingers, hammering away on a not-necessarily-ideal laptop keyboard to put these words into a plain text editor not made by Big Tech that will be published to a website not powered by a Big Tech platform. And it feels good. Because it’s organic and human, and it’s in my control. As are my buying decisions.

And I will not pay for AI.

Don’t get me wrong, I will use AI, we all will. I will use it when and where it makes sense, and not use it when and where it does not. I will use AI for what it is, an MSG-like enhancer that makes the tools I already use–or new tools I’ve not yet stumbled across that are better–to be more efficient. A better writer, perhaps, in my daily work. A better software coder. A better … human being? Maybe someday: Today’s AI is small and uncertain, it needs time to mature.

And I have, of course, paid for AI leading up to this: I paid for Copilot, Gemini, and ChatGPT, and for an Adobe Photoshop CC thing, whatever it is–a plan, I guess, a subscription–because it was important to understand this unfamiliar new technology obsession. And I have seen some value in those efforts, though less than I had hoped. My most obvious success came from image generation because obtaining unique and high-quality visual assets is otherwise difficult, time-consuming, expensive, or all three, and this is one area where AI shows promise. Images are not central to what I do–writing is–but it’s still important, still part of the workflow. It still needs to happen. And anything that can save me time and money while giving me what I need, is valuable. It’s just that I can get this anywhere now. Which is the point.

This is not about being cheap, stealing, or not paying for something I need and use every day. I feel strongly that we should support the makers of the software products and services we use and rely on, though I am divided on perpetual licenses and subscription services, use some of each, and find it uncomfortable. What I’m not divided on is AI: While some AI functionality is useful and even crucial, none of it, no one feature or set of features is what I call a licensable moment. That is, where I understand paying for Microsoft 365 and see the value in doing so, I can’t personally justify paying for Copilot for Microsoft 365, a thing that adds a few features, most I don’t need or want, to a thing I’m already paying for.

I will not pay for AI. And while it’s up to each of us to make these types of decisions for ourselves, we all see different value in different things, after all, have different needs, wants, and priorities, I do think we can come together on one point that’s central to this discussion. Interestingly, this notion came to me early in the AI shift as I struggled to cope with this incredible shift that arrived and evolved with such alacrity, and I referenced it above. AI is not a feature. It’s not a product or service. It’s not a line item on my monthly bill, a licensable moment. It’s MSG for software. It’s an enhancer. When used properly, it will make the features, products, and services we rely on, those things we do pay for, better.

I will pay for a word processor or editor, but I will not pay for spelling or grammar checking. I will pay for photo management and a related cloud storage service, but I will not pay for image editing or generation. I will pay for a video editor, but I will not pay for background or noise removal. AI isn’t a feature or a tier. It’s just something I get as part of the value of the software I choose and pay for. The best of these products will employ AI properly. But none of them are AI.

I will not pay for AI.