This morning, Kathleen Fitzpatrick drew my attention to a Mastodon post by Babette Knauer about plans for the University of Groningen to get away from Big Tech. The article itself is well worth a read, with ambitious approaches to digital independence. I had some not-very-original thoughts on it.
At first glance, this is great stuff. Yeah! Take that, Big Tech. That said, I have been taught well by Geoffrey Bilder to be sceptical around phrasings of “digital sovereignty”, “digital independence”, etc. and the way that they have various nationalistic connotations attached to them. At this political moment, what does such isolationism and boundary-setting say, politically?
Nonetheless, it will be interesting to see how far Groningen will go with it. For instance, will they really (I mean, really!) replace Apple and Microsoft operating systems with Linux? I’ve used Linux on my desktop machine since 2008 and have never looked back. It’s doable. But it’s a change for those used to Windows and Apple. And people who are not tecchy hate this technological change. So there would be resistance.
There’s also costs. Changing the things people are used to comes with a whole different set of support requirements, for instance. Suddenly, your Windows support assistant that you hired does not have the tech knowledge to support a Unix-based system and you either fire them or retrain them.
Also, as with all self hosted open source stuff, when it goes wrong you will basically be fixing it and tinkering with the setup on your own/with online community support. Unless you pay for proper enterprise support - which is what I think should happen in this instance, like the Red Hat model - you are assuming that the free alternative will just work.
Basically: there are costs involved in going to free and open-source software. I am not saying “don’t do it”, but what I am saying is: “don’t underestimate the cost of ‘free’”. Support it properly. So many OSS switches fail, I gather, because people think: “ah, we are going to ‘free’ software! This will not need any technical support because it’s open source”. Then they wonder why it has failed and conclude that it must be worse than the proprietary alternative.
Finally, scale is a huge problem in these breakout activities. So many of the world’s contemporary problems rest on scale. Plagiarism checking, for instance, needs a huge model of all the things you want to check for plagiarism against, and a vast infrastructure that can run the inference on it at speed. So these types of problems require BIG data. And big data tends to be managed by… Big Tech, since you need to be big to be able to handle the big data on a big infrastructure. How we build infrastructures for open-source, free, community-owned versions of these scaled infrastructures remains to be seen.