Thanks to Gabriel Tyler at Knight Institute for a nice summary of this stimulating expert convening on the question, “Can Middleware Save Social Media.” While it captures much of the spirit of this lively discussion very well, there are some points I alluded to that deserve reinforcement. (Full video of these sessions is now online.)
My reinforcements here relate to just what middleware is, how it can be help save social media, and whether it can do that without government intervention to facilitate and mandate interoperability with the dominant platforms.
At the highest level, I want to reinforce my point that there are "two ways you can come at middleware," and that one way is exemplified by Bluesky as “a greenfield development” of a new social media ecosystem. Proponents see this new architecture for social media (based on the innovative generativity of open protocols) as having potential to mature and grow enough user acceptance to replace the dominant platforms, even without government intervention.
There was little discussion of this second path after my comments in the first panel, and admittedly, it remains very unclear whether this kind of ecosystem might succeed at obsoleting the platforms. And of course there is no question that government intervention could greatly facilitate both paths to middleware. But current US policy trends do not augur well for such intervention.
That leaves this glass as arguably half empty, but given the forward-looking work being done in the Bluesky ecosystem that I alluded to -- including the accelerating impetus to build out Eurosky as a counter to the power of the dominant US-based platforms -- I see that glass as half full, even without US regulatory change. If Eurosky development scales successfully, that could spill over to attract demand, and thus supply, for similar services in the US and elsewhere. (The other speakers may be less hopeful about this path than I am, but it would be ironic if it turns out to be Europe that saves US democracy from social media.)
The other point I would add context to is that, as Gabriel put it, “middleware alone cannot resolve the deeper forces shaping online discourse.” That is true, but we should not view this as a matter of "middleware alone."
This bears on the issues discussed in the keynote conversation that followed: concerns that middleware alone cannot save social media -- that instead of a focus on user control, regulators should focus on restriction of addictive or deceptive design practices and commercial surveillance, or supporting other prosocial remedies. The pillars framing emphasizes that middleware (in the narrow sense of the technology enabling interoperation) is not a solution in itself, but it enables all manner of solutions -- middleware-enabled services (not just for user control, but also for social mediation and socially-derived reputation) that can work in concert through an open networked ecosystem. This open market could offer tools that are less addictive and that do not extract and abuse personal data, and that support greater community roles in social media that would encourage people to demand such tools and choose to use them wisely. Government can certainly facilitate movement toward this more open architecture, but user desires for better social media, alone, might drive this shift. That has already brought over 40 million users and growing independent development into the Bluesky's AT protocol ecosystem, even in its highly immature state.
Whichever way we get there, the real promise of middleware is to reduce platform control over the media tools that human discourse depends on. No permission from platforms is needed – if you build it, they can come -- and it can work seamlessly (with all of a user’s data and their social network). You just have to get them to come, and to find a sustaining revenue model.
