The theory of this post is this - what can I do over the next year to make myself and my loved ones more free? By free, I mean free as in freedom, not free as in beer.
I think this is different from resolutions in that it’s more about broad philosophical strokes rather than discrete actions. But also, I want to try and avoid the idealistic, unachievable, and self-flagellating-in-failure nature of your typical set of New Year’s resolutions.
Maybe this can take the form of a retrospective (stop/start/keep doing)? Or, perhaps am I just writing OKRs? Oh god… let’s get this thing started!
The broadest stroke of the tech in the 2010s was intermediation. By this, I mean the change in the nature of business as being mediated by middle-men to one being marketed to and bought directly by consumers (what a horrid thing to refer to people as). Some examples?
The capitalist would say that these are good things - there was inefficiency in these industries, and technology helped them to become more efficient. Who among us hasn’t had a weird or dodgy experience in a taxi, hotel, or buying insurance? Perhaps a little competition would drive everyone to be a little better.
And yet, the result of disintermediation could be more accurately described as theft. Maybe you’re reacting harshly to that word, thinking, well, if Uber hadn’t created a vast underclass of independent contractors making poverty wages, taxi drivers would still be mean to me.
This analysis fails, however, when you consider that nothing about distintermediation requires any particular technology. Yes, I know that AirBnB, Uber, et al produced all kinds of technology (much of it open source), but none of it was crucial for the act of disintermediation. It’s hard to disprove a counterfactual that taxis might have ended up installing credit card readers simply because the cost of credit card readers came down due to better technology.
Furthermore, perhaps, in the absence of “competitive forces”, we simply created regulations that govern the behavior and business practices of taxi drivers? Yeah - such things already exist, and we didn’t need to create said underclass just to have them.
The point is this - even if you believe that competition drives people to make better products, the undeniable outcome of distintermediation shows that we’ve come too far, and now the pendulum is beginning to swing back the other way.
“Enshittification” has come for the disintermediators, again demonstrating that they didn’t bring any actual innovation or efficiency to the table, they just bought money to burn; money which they used to buy a monopoly, and we’re now entering the exploitation phase of monopoly. When VCs talk about “moats”, they’re talking about the means and methods to survive long enough to achieve monopoly.
I want no part of that. The actionable part of this resolution is simply to consider my relationships with disintermediators and either replace them with sustainable alternatives (including, but not limited to, self-hosted solutions) or cut them out completely. It’s certainly true that my life will be a little less convenient and more expensive, but as I said, these are choices made in service of freedom, not convenience or economy.
As I write this, I am struck that many of the most harmful disintermediators fall into the category of marketplaces. For example, consider Amazon as a marketplace for consumers and manufacturers to meet, or AirBnB as a marketplace between (again) consumers and lodging providers (primarily homeowners).
I also consider the vast amount of time an energy being spent on “AI” (really, fancy applications of LLMs) that are being marketed directly to consumers - think Apple Intelligence, or Gemini, ChatGPT, and the like. These are all things that nobody needs, nobody asked for, and are now being aggressively pushed in search of shareholder value. It’s too easy a resolution to just not use these services - but rather, in 2025 I want to build a habit and framework for evaluating technology, and especially ones that are targeted directly at consumers, for the risk of harm implicit in their offerings.
I want to talk a bit about why I’m singling out technologies that are targeted at consumers. The reason is this - if you define your market as all consumers (again, ick), then the potential impact of the choices you make are much higher. Because the thing you’re building is for regular people, you will inevitably affect regular people. This should not be taken lightly - only those that have confidence, or better yet, a structural guarantee that their business model will do no harm should play in this space.
A very easy to understand example of this is personal lending. Banks have long been allowed to market their services directly to consumers - “Get a car loan from XYZ Credit Union”, “ABC Bank has HELOCs”, etc… The harm and risk of harm resultant from a loan is also intuitive - if a consumer can’t pay back a loan, then they are financially harmed. And so, there is a (sometimes tepid, but nonetheless existent) regulatory framework that governs how much interest lenders are allowed to charge, and the terms on which they’re allowed to do so (things like usury laws, amortization schedules, prepayment penalties, etc..).
Imagine instead if the only remedy to a usurious lender were for a slightly less usurious lender to drive them out of business, and then systematically fail to resist the temptation to become as usurious, or more so than their former competitor. Of course, yet another slightly less usurious lender may rise to the occasion, but how many people were harmed by this cycle of increasingly usurious lenders while waiting for the “free market” to get off its ass? Is it not much simpler and effective to simply make and enforce regulations?
Interestingly, I think that this does merit a lower bar for B2B companies; while they can be somewhat banal, there is less risk and motive to harm, and that is preferable. At the end of day, corporations are not people, no matter how much the US Supreme Court wants them to be. This being said, I do want to keep some space for the risk of labor disintermediation, though - to the degree that “AI” startups in particular seek to “replace workers”, I think that they will be ultimately unsuccessful simply because the quality is not there, and never will be.
My mom recently came to visit me, and it happened to coincide with the weekend over which I built Local Library. When I had finished deploying the first build to my local network, I encouraged her to check it out, as she has always been an eager reader.
To my surprise, she found the concept compelling, and downloaded a few books to read during her plane ride! I have always been the kind of write software for myself, and I found it encouraging to see that other people are using my software. For the same reason, I suppose, I sometimes find myself checking the number of stars that Precis has on GitHub - knowing that you’ve made something useful to others feels good! The flip side, of course, is that often your needs turn out to be somewhat unique, and you build something that only you have a need for. It happens, but the sting of rejection as you realize that is an unpleasant feeling.
In 2025, I resolve to lean into that feeling - to seek out opportunities to share what I’ve built with others, and to build a tolerance and resistance to the sting of rejection. At the end of day, I’ll have built something that I find personally useful, even if others do not.
In the same way, when we think about liberating ourselves from disintermediators who’ve proven themselves to be monopolists, the degree to which self-hosted infrastructure plays a role merits consideration. My goal for this year is to provide at least one service for another person. I’m not going to count my mom, unless she decides to join the tailnet! But, a regular visitor that uses something while they’re on my network, perhaps?
That’s it! What else is there to say?