This is a longer version of a post upcoming on LSE Impact Blog
In October last year Elon Musk’s long predicted takeover of Twitter finally occurred, leading to a gradual withdrawal of academics who had come to rely on the platform. In some cases this has meant deletion of their account but more frequently it has involved disengaging from a service which is increasingly beset by problems. Under these conditions Twitter increasingly fails to serve the purposes it once did for academics, such as developing communities and circulating information within them. If there had been an obvious replacement for Twitter then the speed of this departure would have been hastened. But there are obvious limitations to existing platforms like LinkedIn or emerging platforms like Mastodon and Threads which make it difficult to see them as a direct replacement for Twitter. This has left us with an increasingly fragmented social media landscape where academic audiences are split between multiple platforms, eroding the vibrancy which characterised Twitter when it was effectively the default platform for academics. I suggest that Bluesky is the emerging platform best placed to function as a replacement for Twitter. However I remain doubtful that we should be looking for an immediate substitute rather than reflecting more deeply on the role which social platforms now play within higher education. In this post I will outline why increasing numbers of academics are turning to Bluesky, as well as offering a note of caution about where this might be leading in the longer term.
Bluesky originally began as an internal project at Twitter to build a decentralised protocol for social networking. It was driven by former Twitter CEO Jack Dorsey, who initially backed the deal with Musk before more recently criticising his leadership, who now sits on the Bluesky board. It is a public benefit corporation: a legal form that seeks to make a positive impact on society alongside commercial success. The leadership of Bluesky are active on the platform and it is clear that a sense of mission pervades the organisation, even if I have concerns about how this might develop in the longer term. Their mission revolves around the decentralisation of social media in a manner analogous to, but distinct from, Mastodon. Dorsey’s intention was that Twitter would eventually operate on this new protocol but given the doom loop in which it now finds itself, we should be glad that it was spun off into a standalone company.
The interface is effectively a clone of Twitter which makes it easy to grasp how to use Bluesky, with established habits easily translating into this new environment. However there are a number of subtle differences which reflect Bluesky’s commitment to learning from the experience of Twitter. For example the default setting is not to show other people’s replies in your timeline unless they receive a certain number of responses. These are subtle shifts in the conversational architecture which can initially be slightly opaque to a longtime Twitter use. But it important to stress these are customisable settings reflecting Bluesky’s commitment to offering users control over their own timeline. Rather than having a ‘master algorithm’ imposed by the platform operator to further their own commercial interests, Bluesky is working towards a ‘marketplace of algorithms’ in which users can select a method for sorting their timeline which works for them. These custom feeds are likely to become ever more central to the platform over time. At present there are a range of custom fields which academics have created for different fields and disciplines. For example I follow feeds for Sociology and Science, Technology & Studies which have led me to consider establishing one for Digital Education if someone else doesn’t do this in the near future. This is a relatively complex process for the time being but we can expect automated solutions which streamline it is as the platform grows in popularity.
These custom feeds mean you can switch between multiple timelines in ways which will feel powerful and useful once the platform has grown significantly. At present the small scale of the network means these feeds are more useful for discovery rather than filtering. They are a useful way to find accounts to follow and the relevant content they are posting. But if Bluesky grows in a manner akin to Twitter, where regular uses routinely follow thousands of people, custom feeds will be an essential way of filtering this complexity by switching between communities at will rather than having these mashed together in one timeline. It takes the best of the Twitter lists feature and builds it into the core of the platform, rather than leaving it as an awkward appendage the developers didn’t know what to do with. Even if I find the terminology off-putting, I can see how the ‘marketplace of algorithms’ could prove to be enormously beneficial to a diverse research community with different needs and preferences about how to use social media.
I feel a sense of ambivalence about the likely growth of Bluesky over the coming months and years. Since I first joined in 2010 I have run 30+ Twitter accounts with follower counts ranging from a few thousand to over sixty thousand. There is a liberating feeling to finding myself engaging in a network of a couple of hundred people, each of whom I have thought about before following and whose interests match my own in various ways. It is difficult to disentangle the small scale of the network from the design of the platform in making sense of how it feels to be a Bluesky user at this stage. It seems clear that both are playing a part, creating a sense of a welcoming and conversational space in which users are genuinely inclined to engage with each other rather than the hyperactive clout chasing which came to define far too much of academic Twitter.
Over the 2010s I went from being hugely enthusiastic about social media to being exhausted by it. Richard Seymour’s invocation of Paul Klee’s painting The Twittering Machine captured this perfectly for me. The critic Kay Larson describes birds which “whir helplessly, their heads flopping in exhaustion and pathos” defined by a “grim fate – to chirp under compulsion”. This is what academic Twitter felt like to me for a long time as our use of the platform became entangled with systems of evaluation and repetitional hierarchies within the academy. What had originally felt like a smorgasbord of intellectual delights and thought-provoking conversation came to feel like a wearying chore, in which I jumped through hoops in order to sustain the visibility on which my career depended, particularly when I was precariously employed. In contrast my first weeks on Bluesky have reminded me why I was enthusiastic about Twitter in the first place: an interesting mix of people whose interests overlap in different ways with my own, sharing an eclectic range of reflections coupled with a willingness to talk about them. There are far fewer external links in my timeline than I would find if I scrolled through Twitter. This is not to suggest there is anything inherently undesirable about sharing external links through social media, simply that prioritising personal voice leads to a more engaging and conversational platform. It is a breath of fresh air to use Bluesy after years of Twitter and it has reconnected me with people I had lost touch with and topics I had fallen out of the habit of thinking about.
It will be interesting to see how Bluesky develops over time. While I am extremely enthusiastic, it is clear these positive elements reflect the small size of the user community and my own network within it. There are reasons for optimism about how it will grow over time. The invite code system means the network is growing in a slow and purposeful way. The decision to grant each user one invite code per week encourages people to be thoughtful about who they decide to share these codes with. This means that networks are reconstructing themselves based on existing connections which the user values i.e. I care enough to share a code with you. There is a further element of self-selection in the users who are inclined to use Bluesky in spite of the relatively early stage of its development. This network structure is manifesting itself in a user culture which repeatedly jokes about ‘the other place’ and takes a reflective stance towards the norms emerging on the platform. For example there is a drive towards ensuring that alt text is provided by all users for images, with many committing to not resharing posts which lack it. This is one example of how what Jean Burgess and Nancy Baym describe as ‘public pedagogy’ is extremely visible on the platform. Existing users narrate how they approach the platform and explain its functionality in order to ease the transition of new users into the network. This increases accessibility of in a way which can be extremely welcoming, particularly given how it fits within existing networking which are (re)constructing themselves through the access code system
However I wonder if this user culture might sit uneasily with incoming users as the platform grows. Bluesky have suggested the allocation of unit codes might increase significantly at a later point in the platform’s development. This would mean uses will be less selective about who they invite, lessening the extent to which the strength of existing connections and the associated public pedagogy can help socialise new users into the platform’s culture. While I find the normatively focused and reflexive user culture of Bluesky refreshing, I find it easy to see how some incoming users could find it overly directive and possibility grating. The culture clash which was seen on Mastodon in late 2022, “with its purposeful complexity, designed to be antiviral” yet initially positioned by many as a replacement for Twitter, could easily be replicated with Bluesky in the coming months. The fact that Bluesky is, at least to some extent, positioning itself in this way only amplifies the potential for culture clash. What happens when users are not self-selecting and joining through the existing networks of early adopters? How much of the positive user experience which academics currently find on there will survive the transition?
Furthermore, the relatively shallow network which exists at present means the instrumental possibilities for academics are relatively limited. It is not a good place to shout loudly about your work in pursuit of academic celebrity because most academics have few followers and professional narcism is a poor way to build them up. This means that Bluesky’s culture is a breath of fresh air for those of us who felt choked by the self-promotional culture of late stage academic Twitter, while also feeling hypocritical for simultaneously participating in it. This is likely to be a check on its growth at present given how much competitive individualism, encouraged by the real but unpredictable rewards universities offer for platform popularity, has driven the uptake of social media by academics. But this competitive individualism will eventually arrive on Bluesky, possibly at the point where Twitter simply becomes untenable, even for those academics who have built up massive social capital on there over the years which they have been loathe to part with.
This is not just a matter of how shifts in the network will impact the user culture. Bluesky is in such an early stage that they are without a business model and much will depend on how this plays out over time. Early suggestions are that Bluesky might eventually move towards a subscription model. In an interview with The Verge in April, CEO Jay Graber suggested that “In an open marketplace, there will very likely be value-added services that people find worth paying for”, though stressed that the firm’s immediate focus was on the challenges of moderation and growth. In fact Elon Musk’s announcement that he planned to charge all Twitter uses actually led to a spike in Bluesky’s membership, with 53,858 new members (5% of the platform’s entire user base) having joined by the end of the following day. This vividly captures how Bluesky’s positioning of itself as the replacement for Twitter is resonating with the legacy platform’s user base, long attached to the service but increasingly unwilling to tolerate how it is unravelling under Musk’s leadership.
There is nothing inherently problematic about charging for subscription, particularly if it provides an escape from unsustainable advertising models which incentivise surveillance practices to enhance the targeting offer they make to corporations. What makes Musk’s Twitter so toxic is the manner it which it hybridises a failing advertising model with a failing subscription model. It might be that Bluesky persuades a growing user base this is the precondition for retaining a service with the features which they value. But it is difficult not to wonder how clarity about their eventual monetisation strategy might shape the the willingness of ex-Twitter users to see hope in Bluesky. The firm describes its strategy at present as being to experiment with business models. As they put it in a blog post, their fundraising was motivated “to find new partners and to give ourselves room to grow the network and experiment with new business models”. One such experiment is their domain service which enables users to establish a custom Bluesky handle attached to their own domain without having to manage the domain themselves. It obviously won’t provide a viable model in itself but it gives a sense of the creativity which the firm is bringing to their business planning.
There is an experimental ethos at the heart of Bluesky, reflecting its origin as Twitter’s skunkworks. It represents the best of a fascinating, flawed yet hugely significant platform which is being run into the ground by a billionaire’s hubris. While it is fascinating to watch Musk burn his way through billions of dollars of investment, it makes me deeply sad as someone for whom Twitter was once a hugely positive part of life. I know I’m not alone in this. For this reason I sincerely hope the Bluesky experiment succeeds, keeping alive a user culture which is at present so reminiscent of Twitter’s golden years. It has a talented team who are clearly driven by a mission. However we should not be optimistic purely on this basis. Bluesky is a public benefit corporation but OpenAI was partly a non-profit and look how that is turning out. My concern is that platform capitalism has taken the form it has because this is how tech firms scale in the environment they confront, rather than reflecting the malign intentions of bad actors who simply lacked a real commitment to making the world a better place.
If the academic community is looking for a replacement for Twitter then I am increasingly convinced Bluesky is what we have been waiting for. However I hope our embrace of it does not prove to a mistake in which we once more outsource the digital social infrastructure of the research community to a private firm rather than finding a way to build it within the sector. I cannot shake the feeling we might be entering into a dynamic of digital migration of academics responding to waves of what the internet scholar Cory Doctorow calls enshittification as one platform after another becomes unusable. If social platforms have become indispensable to our work as academics then perhaps it is time for the sector to have a much deeper conversation about how we make them work for us, rather than leaving it to individual academics who have too often found themselves working for the platform instead.
Mark Carrigan 