Looking back it was a pretty good dozen years in many ways. From when I started using Twitter for EU political purposes early in 2009 until it all started to go wrong throughout 2022 I managed to achieve a hell of a lot on that platform.

But it was based on a misconception. That Twitter was a social network, not digital public infrastructure.

It all started innocently enough. I was in the middle of a participative and decentralised blogging scene in the years 2005-2009, and Twitter started out as an addition to that – the place for the quick interaction that did not justify the time and thought a blog post required. And – unlike Facebook and LinkedIn that pre-dated it, at least for me in terms of when I signed up – Twitter had a lot going for it. “Follow” rather than “Like” gave the interactions a neutrality that suited public political discussion. Being able to interact fluidly and flexibly with people you did know offline as well as those you did not know in “real life” was a plus as well. It allowed me to take what I was talking about to new audiences, and to put together and take apart communities with all sorts of interests.

The EU institutions jumped on the Twitter bandwagon, driven in part by Neelie Kroes’s commitment to it. It hit a kind of high in 2014 when a new Commissioner – whose Twitter name I had pre-emptively registered – had has staff contact me to “officialise” this account. A Commissioner has to have a Twitter account the staffer told me, and I chuckled and nodded, and handed the account over to them.

Lots happened. I broke political scandals on Twitter. I ran informal events known as #EUTweetup. I coordinated protests against Berlusconi on it. I was part of a remarkable community of Brexit experts on it.

Professionally I managed to get a visiting lecturer gig at the College of Europe largely because I could teach future Eurocrats about Twitter. I have run dozens of training courses for NGOs, trade associations and EU institutions about it. I have made professional connections and what will be life-long friendships through that tool.

But we were building Twitter into something that EU politics nerds could not do without.

We were building digital public infrastructure – on an American, private platform.

The frivolity of the early years – where Berlaymonster could share a joke with Cecilia Malmström on the site – came to be replaced be a professionalism and seriousness that was not all for the best. Sure, Twitter was still the place for breaking news – a deal at a European Council would be tweeted by Donald Tusk before it was told to the press. But that also meant that for senior people some of the humanity was lost. Make an error on Twitter and that could easily turn into a major political scandal. Better get your staff to do it for you rather than do it yourself.

There were of course warning signs. The massive diversity of Twitter tools from the early days was restricted even during Jack Dorsey’s time as CEO (before Musk finished the job), but I stayed. Twitter has to make money, I told myself. And savvy users could still get chronological non-algorithmically filtered timelines if they knew what they were doing (and I taught my clients how to do this), and could avoid the most obvious advertising in their browsers if they used an ad blocker. I’d have paid to use Twitter in the good days to avoid this, but there was not a way then.

I had a parody account summarily suspended with no way to get it back – an interesting warning I think. And of course there was the mounting amount of trolling and bot-generated content, peaking in my timeline around the Brexit referendum.

But the crux was that despite it all, it was reliable enough, mainstream enough, predictable enough to justify still using it – until Musk came along.

And when he did, and swiftly broke it, it was not too hard for for me to escape (as I have outlined here).

Leaving Twitter behind was not without some personal cost, but whether I am on Twitter or not is not really central to the network in EU politics. Sure, I can convince myself that a few people there might miss my tweets once in a while. But I am not a politician concluding an agreement at a summit, a comms unit of a DG of the European Commission that thinks it has a responsibility to inform a wide public of what it is doing, or a journalist who is at least partially paid based on the numbers of eyeballs on what they produce. For them Twitter is infrastructure that they think they cannot do without.

The most used alternatives look little better. TikTok has espionage allegations against it, Meta (that owns Facebook and Instagram) is likewise a centralised American-owned platform so has the same weakness as Twitter. LinkedIn – owned by Microsoft – might manage to be a bit more resilient, but is not necessarily a better bet either. And that’s before we come to what Cory Doctorow calls the enshittification of these platforms.

So what is the answer?

European publicly funded digital infrastructure (as outlined in papers like this). Efforts to line up the values of public institutions with the technology that they use – as outlined by Ethan Zuckerman and others here. Decentralised, open source tools – that are not controlled by one central entity, one unaccountable likely point of failure. Build up things like social.bund.de on Mastodon instead.

And at an individual level, those who can take more ethical tech choices need to vote with their feet – and leave the centralised unethical platforms behind and do their bit to build up better alternatives.

When – in the 2010s – European politicians were aghast that Europe had no massive tech firm akin to Facebook, I shared that concern. Now – in the 2020s – I see it very differently. Decentralised, open source, interoperable, a mixture of commercial and non-commercial is the way to go – and with an inherent understanding that building public infrastructure on privately owned territory, in the digital as well as in the physical world, is not a clever call.

We are only at the start of our thinking about this. But that late 2022 period when one tech bro, in one fell swoop, broke a lot of what we had been building on for years, shall act as a cautionary tale for years to come.