You might have noticed that a lot of software has been getting worse, not better. The writer Cory Doctorow has a great term for this: enshittification. The basic argument is this:

Here is how platforms die: first, they are good to their users; then they abuse their users to make things better for their business customers; finally, they abuse those business customers to claw back all the value for themselves. (Doctorow, 2023)

Capitalism demands growth of profits. And it demands it endlessly. It’s not enough for a product to be good, to attract those who are interested in it, and then plateau, comfortably working on making things better for the existing userbase. Profits must continue to grow. That basically means some combination of three things: attracting more users, getting more money out of existing users, and cutting costs.

The first of those – attracting new users – is great at first, and can align quite well with building a good piece of software. New users will flock to a good new product. But it hits diminishing returns. For an extreme example, imagine you’re Facebook. You have more than three billion users. Realistically, everyone who will ever use Facebook is already using it. You’re not getting anywhere near as much return on investment for attracting new users as you were in the beginning. But remember: growing slower is not an option. So that leaves the other two choices.

Cutting costs also has diminishing returns. You can streamline processes and make things more efficient to some degree, but eventually (or quickly) you run out of fat to cut and you shoot yourself in the foot.

So you have to squeeze more money out of the users you have. That can come in many forms. Standard options include: increasing cost, putting more things behind a paywall, moving to a subscription model or adding microtransactions. But we can also get more creative. Your users’ data can be sold, and more of it can be harvested. Instead of improving the core product, you can sidestep your product into other domains, like popular productivity app Notion adding calendar, mail and AI integrations which don’t really work properly and aren’t fully featured. Or Proton neglecting some key issues in their cloud storage drive (including the long-requested Linux client) in favour of adding an AI chatbot, a password manager, a cryptocurrency wallet, an authenticator and a video calling app.

A big impediment to profits here is also ethics. So we could just get rid of those to find some extra growth, too. Think lucrative contracts with controversial governments, militaries and institutions. Microsoft’s dealings with the Israeli Defence Force is a great example. Or OpenAI striking a $200 million deal with the US Department of Defence. Or Palantir being awarded a £330 million contract by the NHS. (Okay, that last example is kind of reversed compared to the other two, ethically. But still.)

The point is that enshittification and ethics end up intertwined. The same profit motive that means we can’t have nice things because they aren’t profitable enough also means that the now-not-nice things are also evil things.

What I want to do about it

So, I want to be more thoughtful and ethical in my choice of technology. Normally, I am a systems thinker. That is, I don’t think an issue like this should come down to individual responsibility. But given my position as an AI researcher, I think my individual choices should aim to set an example.

So I want to exorcise big tech from my life.

I don’t expect I will succeed. If we live, as former Greek finance minister Yanis Varoufakis argues, in a technofeudal world now, then the prospects of escape are not good. His point is that big tech now controls so much of the fundamental online infrastructure that it’s basically impossible to avoid paying them ‘rent’. No matter what software I use, I’m going to end up interacting with Amazon Web Services, for example. Nonetheless, I think the attempt will be illuminating.

This project is about shifting my workflow and personal technology use away from big tech companies: Microsoft, Amazon, Apple, etc. Wrapped up in that are other principles that guide my decision-making here.

Here’s a non-prioritised list of principles with which I will choose my software replacements:

• Good, robust software that fulfils the day-to-day user needs I have for it.
• Free, open-source software (FOSS).
• Ran in the interests of its users.
• Privacy and security focused.
• Not American-owned.
• Not big-tech-owned (primarily meaning Microsoft, Alphabet, Amazon, Apple and Meta, but also whichever others come up to this level or fit the vibe).
• Not run for profit.
• No links to or contracts with unethical partners, such as the military-industrial complex, big pharma, the Israeli government (especially the IDF and illegal Israeli settlements), or the Russian government.
• Does not harvest unnecessary data and/or sell user data.
• No/very minimal AI integration. What AI integration there is should be able to be switched off.
• Does not try to lock users in to a walled garden.

That all sounds a bit Goldilocks. I expect some compromises will have to be made. For example, my workplace (University of Bergen) – as well as many others – is well integrated into Microsoft’s Office365 environment, so that’s impossible to avoid entirely. Plus, I will not, for my sins, use software that sucks. I refuse to believe that good software and ethical software are mutually exclusive.

Where I am so far

Here’s where I’m at so far. I’ll continue to update this as my journey continues, offering my thoughts on switching some of these out, like going from Windows to Linux and from Office365 to LibreOffice.

I hope in particular that my experiences will be useful to students and academics looking to do the same. There are many lists of alternatives out there, but many of those write-ups (if they aren’t AI generated) are also very general, and don’t mention some of the specifics of this line of work that might influence our choices. One is exception is that Jesper Juul wrote about switching office suites last year, which is a very valuable contribution. Personally, I disagree with his choices and some of his requirements/expectations. E.g., with his assertion that free tools automatically means ads and tracking - not true in FOSS - and that we should therefore skew towards proprietary, paid, but European software. But it’s well worth reading his suggestions for yourself.

One huge caveat is that this mostly goes for my personal PC at home. My work laptop is still fully Microsoft-pilled because, as I mentioned, my workplace is far too integrated into that ecosystem. However, where I can – such as with the office suite and cloud storage – I am still switching to the alternatives below.

Over time, I’ll write up my experiences with each of these. Click on the links below to find the blog posts I’ve put out so far.

The Operating System

Last update: 22 January 2026.

General

Work

Personal