Mastodon is extremely technologically stagnant, with one and only one player controlling the entire technical ecosystem.
There is a huge lack of interesting apps or innovation. The protocol is very narrowly defined, and does not have any Postel's Law characteristics. You have to use the limited API offered in the Mastodon form.
There is starting to be some interest from ActivityPub in defining their own client APIs. But this gathering together of people is only just starting, and it's unknown where that effort will go or when it will see traction in adoption.
It also sucks being moderated by whatever Fedi you are on. And the very poor state of account portability sucks.
There's also internally a very aggressive culture, against devs doing fun and interesting things with the social media data. The social rules of engagement seem to be that you can only enjoy Mastodon as a feed that goes by, and you cannot download or analyze deeply. You can't index or search, except with strong carved out cases. People building search or indexing tools are harassed and abused.
I strong recommend this excellent write up, which does a far better treatment than I could offer, on Mastodon: https://connectedplaces.online/reports/fr156-share-where/
Atproto has none of these issues, and has a rich ecosystem of devs building neat tools like RSS readers (skyboard), writing (greengale), book reading (bookhive), trail making (side trail), annotations (margin.at), events (smoke signal), chat (roomy), social/research bookmarking (semble), video streaming (Streamplace), media watching (pop feed), key attestation (keytrace), git (tangled), and yes, devs: containers (atcr).
I'm sorry but Mastodon is a dead end, a bad design, going no where, bereft of interesting dev engagement, with data that is hard to share & make interesting use of, data that is extremely narrowly constrained in shape/form.
I am envious of the dev-centric culture there. But it astounds me that devs would choose to exist on a place that is so technologically unalive & so low potential.
It’s weird that you think a social network needs to have some sort of technological drive or entrepreneurial spirit to be relevant. If anything, this will lead to its inevitable enshittification, as you can plainly see in Twitter and Facebook.
I agree that the sprawl of objective, trying to cover more and more, is quasi horrific in the ways we have seen it: seen through market captures and monopolization and acquisitions of Instagram, etc.
But for computing, not having a network fabric that is adaptable to purpose has kept people from having any starting point to play with. Every time you want to build a social system, you start with your own website (or go even lower level to reinvent), with your own bespoke back end, your own unique internal and or external API, rather than having set protocols you can adapt.
This isn't strictly true. There have been many attempts to build more good protocols, with endurance and reuse and adaptability. RemoteStorage, inventor of the web Tim Berners-Lee's Solid, HyperCore/Swarm, or Dat (which atproto cto @pfrazee built a browser for, Beaker). (Mastodon was never one of these & is not now.) This is not the place for writing a Speaking for the Dead for each of these, (and none of them are dead beyond grasp), but so far this question of what gets the most use, what has seen the most apps built upon it, the most end user adoption: it has a clear answer and it's atproto.
Atproto layers social networking right. You are self sovereign to a data store that you control (absolute authority, can-move/credible exit), that has your records in it (https://overreacted.io/a-social-filesystem/). It makes syndication cheap & easy. It makes relays that aggregated PDS's while presenting a like API surface to the individual user, to facilitate syndication en masse. It layers indexers (appviews) that provide aggregation etc. This mirrors a functional social network system, goes beyond protocols to be a complete system, that is imminently individually maintainable, at scale.
It provides a platform that anyone can use to be creative, in a connected fashion. Your phrasing hints at the whiff of terror that these large companies bring, the reign they have over our lives. The drive they have is to saturate, to take our attention. I see the role of atproto & it's drive as a network technology layer to allow new social technology layers to be created at very low cost. It allows individuals to build amazing services, very lightweight and client only / isolated web apps that speak the protocols and which can put or retrieve records or appviews, for any kind of social network technology.
That to me lacks the scariness you speak of. There is not one overarching drive or ambition. It is not centralized or directed intent, working cohesively to capture society. Both are expansive & space filling, but their characteristics are as far apart as I can imagine.
@pfrazee distilled it well in Atmospheric Computing (https://www.pfrazee.com/blog/atmospheric-computing), a lovely post. That to me resonated strongly everywhere but especially the cold start problem. That it used to be hard for people to make connected things. And now it's not. That's the essence of the cold start problem. And rather than control the things we create, the role of the creator is different: they facilitate creating records, tools to help people build Abramov's Social Filesystem (previous link). They don't nor should they want to own the network, the protocols, the data, the users, the accounts. Those are atmospheric ambient systems that exist.
> It’s weird that you think a social network needs to have some sort of technological drive or entrepreneurial spirit
I perceived a noted lack of techno-social substrait upon which individuals or entities could collaboratively build or explore. The systems humanity has built have been dead end systems, built for narrow set purposes. There has been a missing General Systems Research component to computing for decades. There has never been a successful General Network Research effort, beyond the web itself (what a hit), and the vision TBL had for the web as a bidirectional writable author able medium never materialized. BlueSky corporation has me at least convinced that nothing can stop credible exit, that this is not their network, and independent actors like BlackSky have somewhat proven that out already in astoundly sophisticated end to end fashion, with incredible thought to their own moderation systems etc.
Tl:dr: actual humans have lacked the means to explore and build the techno-social, and it is absolutely nuclear hot fire that I have all this from little apps a bunch of creative friendly awesome neat people built that lets me capture this stuff (https://pdsls.dev/at://did:plc:zjbq26wybii5ojoypkso2mso/). This is proof that we have a general social networking system right here today available for use, experiment, & play. Not proof, but imho the roots are strong & good. This is how we escape the capture-entrepreneurialism. We have lacked fertile soil & good foundation for new things to get started on. I'd characterize the distinction as alive, adaptable, usable systems, versus stagnant dead systems, and imo, stagnant fixed systems aren't serving us well now, and over time decay & en-badification to serve us worse. Hence my advocacy of aliveness & possibility. (And my disdain for Mastodon as barren anti-ecosystem.)