Unsubscribe & Resist with AI!

One thing I’m doing with large language models is using them to get rid of unnecessarily nosy corporations, their noisy platforms, and their disappointing politics. In the same vein as Resist and Unsubscribe, I’m taking those services that track the music I listen to (and keep the files for themselves), or record where I am and what book I’m reading (while they charge double what a digital book should cost), and switching them over to AI agents that will serve me and only me and only what I want with NO ADS.

Follow along here and I’m going to pick on a few services and tell you the easiest way to do this: flat files, baby.

Remember When You Actually Owned Your Stuff?

I do. I remember ripping CDs and putting them on what was essentially a glorified USB stick with headphones. My music. My device. My control. Done.

Apple started with this promise. Steve Jobs said “You can have songs for 99 cents” and it worked. You bought the file, you owned it, you could put it anywhere. It was DRM-lite that people accepted because it felt fair and it was convenient.

But then they looked at Spotify’s numbers and thought “wait, why let people own things when we could charge them $11/month forever?” Now the iTunes Store is buried in the UI, search defaults to Apple Music streaming, and if you actually want to manage a local library, you’re treated like some kind of digital hoarder. Meanwhile, if Apple Music decides your favorite album isn’t worth licensing anymore—poof—it’s just gone.

Congratulations, tech bros. You missed the point and ruined music. You have to be born before 1980 but there was a vinyl album that you hugged and those days don’t have to be gone.

It’s Not Just Music

Goodreads tracks what you’re reading and sells that data while Amazon charges you twice what a digital book should cost. Strava knows everywhere you run. Spotify knows your workout playlists. Patreon and Substack take cuts of money flowing between creators and readers who could just… have that relationship directly.

These platforms aren’t symbiotic. They’re leeches.

And the worst part? They convinced us we needed them. That managing our own music library was too hard. That keeping track of our own reading was too complicated. That we couldn’t possibly organize our own information without their beautiful, algorithm-driven, ad-supported interfaces.

That was always bullshit.

Scale Is The Problem

You know what else these platforms did? They turned everything into a stadium when what we wanted was a salon.

Pinboard—the bookmarking service that charges you a one-time fee and then just… works—has been running for over a decade without turning into a nightmare. Why? Because it stayed small. Because the creator, Maciej Cegłowski, resisted every urge toward feature creep and growth-hacking and whatever other bullshit VCs demand.

Meanwhile, social networks got huge and became unbearable. Reddit has millions of users doing unpaid labor to keep communities running—shouldn’t they be on payroll?—indentured servitude much? Or at least able to charge Patreon-style fees for their work? Doesn’t a co-op model make way more sense than this extractive nonsense?

You can’t have an interesting conversation in a sports stadium. Scale killed these platforms. They optimized for engagement metrics instead of human needs, and now they’re just… exhausting.

The Guy Who Invented The Web Agrees With Me

You know what’s validating? Tim Berners-Lee—the person who literally invented the World Wide Web—looked at what it became and said “we need to fix this.”

His Solid project is literally the inventor of the web saying “yeah, we fucked this up, here’s how to fix it.” Solid is built around personal data pods: containers YOU control, not corporations. Your data lives in your pod. You decide who gets access to what. No intermediaries harvesting everything you do.

It’s the same philosophy I’m talking about here, but with serious technical infrastructure behind it.

Here’s the thing though: Solid has struggled with adoption because it requires technical knowledge most people don’t have. Setting up and managing your own data pod? That’s a barrier.

But this is exactly where AI comes in.

Identity, Privacy, and Why This Actually Matters

If you’re going to have AI agents managing your personal infrastructure—your music, your reading lists, your writing, your subscriptions—you need to establish identity. Secure identity. Because these agents need to know it’s you and they need to keep that information locked down.

You can’t have AI managing your stuff if there’s no secure way to say “this is MY data and only I (and my agents) can access it.” Without that, you’re not building personal sovereignty—you’re just replacing corporate surveillance with AI surveillance. That’s not liberation. That’s just a different landlord.

This is where Solid’s architecture becomes crucial. The pods are designed exactly for this: your identity, your data, your control over access. And AI can make them practical for people who aren’t developers.

Imagine: your AI agent lives in your Solid pod. It manages your library, parses your feeds, organizes your files—all within infrastructure YOU own. No corporate platform in the middle. No data leaking out to advertisers. No algorithm optimizing for engagement.

Just your stuff, managed the way you want it, with privacy baked in from the ground up.

Solid is the foundation. AI is what makes it accessible.

AI Can Fix This (No, Really)

Here’s the thing everyone’s missing: AI is finally powerful enough to make personal digital sovereignty practical again.

Remember RSS? That “dead” technology that never actually died? It’s an open standard. Decentralized. Privacy-respecting. User-controlled. No algorithms deciding what you see. No company that can shut it down.

We stopped using it because we got convinced Facebook and Twitter would be better feed readers. What a mistake that was.

But now? You can tell an AI agent: “Check my RSS feeds, summarize anything interesting about X topic, ignore the rest.” Or “Find me feeds about Y and add them to my reader.” The organizational tedium that made DIY solutions feel like work? AI handles it.

Same with music libraries. Same with reading lists. Same with workout logs. Same with everything these platforms claimed you needed them for.

The infrastructure is cheap now. Storage costs nothing. The models are good enough. You don’t need to be a sysadmin. You just need to decide you’re done being farmed for data.

Going Back to 2004 (But With Better Tools)

I’m not being nostalgic here—okay, maybe a little—but we had this right in the early 2000s. Everyone had their own website. You owned your MP3s. Software came in a box and once you bought it, it was yours. RSS feeds connected everything without any corporation in the middle.

Then venture capital strip-mined the internet and turned everything into a subscription service. Everything became a rental. Nothing is yours. We’re just leasing our digital lives from five companies who can change the terms whenever they want.

But we can build it back. Not the exact same way—we’ve got better tools now. AI can manage the complexity. We can have our own stuff, make it work the way we want, and tell Apple and Amazon and Spotify to eat it.

This is what AI is for. Not replacing workers or generating slop or whatever dystopian thing you’re worried about. It’s for rebuilding personal digital infrastructure. It’s for democracy. Intelligence belongs to everyone, and we have to resist these technocrats who aim to own it.

What I’m Doing About It

So here’s my plan, and you can follow along:

I’m ripping my CDs again. I’m putting them on a device I control. I’m using AI to manage the library, build playlists, and sync across whatever devices I want—no subscription required, no licensing deals that can expire.

I’m tracking my reading in plain text files. AI can parse my notes, find patterns, make recommendations from my actual library without feeding data to Amazon.

I’m blocking ads again. Hell, I’m building a time machine back to 2004, but with the advantage of having AI agents to handle the boring parts.

And I’m writing this on my own site, which you’re reading via RSS or direct URL or however you found it—but definitely not through some algorithm that decided you should see it because it would keep you engaged for 0.3 seconds longer.

The Political Part (Because Of Course There Is One)

I can’t get technical in these posts, but I need to be clear about this: the AI future looks bright to me, but American democracy does not.

AI is for the public. Coded right, it’s for democracy. We have to resist the consolidation of these tools into the hands of the same people who consolidated everything else. The same people who turned music into a rental service and social networks into surveillance systems and the web into five websites filled with screenshots from the other four.

This isn’t just about convenience or being a cranky Gen Xer (though I am both). It’s about building parallel infrastructure that can’t be captured. It’s about demonstrating that another way is possible.

Digital sovereignty. Personal infrastructure. Your data, your tools, your terms.

Let’s make dumb people smart again. Let’s make the web weird again. Let’s own our own stuff again.

Unsubscribe. Resist. Build.

---

Want to follow along? I’ll be posting updates on specific platforms I’m ditching and how I’m replacing them. No promises on schedule—I’m doing this on my own time, on my own site, and nobody’s algorithm gets to decide when you see it.