0.0 Context Setting

Wednesday, 14 January 2026 in Portland, Oregon where while the conditions in the U.S. may not be argued to be fascist as such (i.e. the government is losing in courts and appears to be following, to a certain degree injunctions), it’s not as if the government isn’t trending in a fascist direction. Sure the water’s not boiling, but it’s also not getting hotter, and just because there are still institutions that are slowing things down, slowing things down is different from “stopping”, and “stopping” is different from “reversing”, and “reversing” is different from “the degree to which you are confident that systemic changes will be put in place to prevent This Sort of Thing from Happening Again”.

This episode is another accidental lots of words, sorry.

0.1 Events

Other Events

Alongside the Han to Hon Combat I co-host with Ted Han, I am also yoloing another podcast!

Courtney Eimerman-Wallace of Wildcard will be co-hosting The Turnaround where we’ll be talking about product turnarounds with founders and leaders who’ve watched their products fail and then figured out how to fix them. If you’d like to take part, drop me a line and we’ll chat.

1.0 Some Things That Caught My Attention

This Thing That Caught My Attention is based on a thread I wrote the other day on Bluesky¹, then realized I should just continue it here instead.

1.1 What does it mean to be friends?

Here’s the deal with social networks.

People treat social networks -- the apps themselves -- as different physical places.

Certain people (say, software developers and product managers, I guess) see a bunch of inefficiency here because ultimately what’s going on here is that there’s You and then there’s Other People and the way we experience most social networks, the big ones at least, is that you are either Friends with a thing or you are Following a thing. (Fine, with LinkedIn you have Connections, which is actually interesting from the point of view of most plainly describing the relationship between two nodes in a graph)

I’ve written about this before: I present and act differently on LinkedIn than I do on Reddit than I do on Instagram than I do on Metafilter than I do on AVForums than I do on Bluesky than I do... you get the idea. We contain multitudes.

Google+ tried to reflect this in attempting to do Circles, which was a way of divvying up All The People You Know into different groups that could overlap. That way you could stay in the One App, and then share photos with friends and family (e.g. Flickr), post to professional groups (e.g. LinkedIn), and also shitpost (uh, everyone else?). But you’d still be you and there’d still be one account.

Google+ Circles was trying to solve the problem of people creating pseudonymous alt accounts in part because having all those alt accounts is a bit frustrating because of the friction.

But consider: the friction is a good thing because it enforces separation and a different mental model, it makes clear that the context is different.

Apps that want to simplify the type of relationship you have with people or things all the way down to Following in that way contribute to Context Collapse. On Bluesky I am doing a combination of talking out loud, talking to a close circle of friends (by, for example, making an inside joke). I personally don’t have a problem doing that in public. When you get to an account that’s Too Big, that’s when I think you commonly encounter context collapse because, through quote posting at least, Total Randos Who Know Nothing About You have the opportunity to willfully misconstrue what you’ve said because they’re tankies.

Bluesky -- which is more interesting because it’s a social network built on ATProto, which for regular people I guess is equivalent to “the really interesting stuff you can do with HTML², ie a thing that you can use to make lots of different things -- which is an attempt on one axis to solve this issue of one app having a silo of all your relationships.

One of the technical discussions going on right now is this: the way Bluesky works right now is that you’ve got, like, a box of stuff and right now that box of stuff is in a storage unit run by Bluesky the company. (An aside: you can take that box of stuff and move it elsewhere, which is Good, because nobody else lets you do that right now).

Now, this box of stuff is open. Anyone can look inside and see the stuff inside it.

One of the things in that box of stuff is a notebook. You can let anyone put stuff in that notebook. If Reddit wanted to put stuff in the notebook in your box of stuff, they totally could. Right now, Bluesky is the only practical one putting stuff in there.

Because Bluesky’s the only one putting stuff in the notebook, then Bluesky has put together a bunch of rules for how they write things down, like the column names in a spreadsheet or whatever.

One of the things Bluesky has put in your notebook is a list of all the people you’re following. One of the other things in there is “all the stuff you’ve ever posted to Bluesky”.

Anyone can read that notebook. Right now, that notebook is unencrypted so anyone can read it without needing a key. You could totally put encrypted stuff in there if you wanted.

Say there was a New Better TikTok. Right now, instead of doing the regular “import contacts from your phonebook”, a New Better TikTok, could just take a look in your box of stuff, see the notebook, and read the list of people you’re following that Bluesky put in there. Then it can show you that list and ask if you want to follow them in New Better TikTok, too.

This is better than what we have now because pretty much all the social apps out there jealously guard the other people you have relationships with because “the value is in the graph, which is true, but also not. It’s true because social apps don’t work if, well, there’s nobody to be social with and nothing to be social around.

Now. There are two ways this could work. New Better TikTok could also use that notebook in your box of stuff to put stuff in its own section. or it could also kind of just start its own notebook in your box of stuff, too. That stuff would also default to being public and open to anyone to read, just like Bluesky’s stuff, unless New Better TikTok encrypted it.³

This would seem a bit inefficient and redundant. Now you’ve got two lists of contacts in a notebook. Some of those contacts might be the same. Wouldn’t or couldn’t it be more efficient if you just had one list of contacts, and then for each contact you had like a notes field and the notes field said “following this thing on Bluesky and New Better TikTok”, instead of alist for each one?

I mean that’s one way you could do it. (I am still thinking this through). What’s easiest right now though, because there maybe isn’t enough coordination, is for different apps to their different lists in the notebook in your box of stuff.

One of the ultimate issues to have to accept is that a computer can never, ever, capture the full richness of your relationships with people or things. Those relationships are so context dependent, so context sensitive, they change over time, and so on. All we can do is come up with simplifications that are good enough to get the job done.

Right now the way this works is that if you normally import your contacts and then match against those (which Bluesky recently implemented) or you are lucky and can import from another app. In both cases, you are practically hunting through a shitload of things and the context is being collapsed, again.

I guess a point here is that there’s inherent metadata in the source of the graph/relationships. At the very least you’d record that “these people were imported from Reddit” and remember that so you can distinguish the Reddit influx from the LinkedIn influx.

There are other interesting ways this could work. This is one way to definitely not do it: the notebook in your box of stuff, which is effectively “your account” could also list all the places where you’re active. It could say for example, that you have a LinkedIn account and a Reddit account and a TikTok account and for the purposes of this example, a FetLife account because what we’re talking about here is making it “easy” for a person to be present in different places. But I imagine that for most people who have a FetLife account they wouldn’t want anyone by default to know that they had an account here? Most contemporary cultures are pretty judgmental about that kind of thing, especially if anyone could see your membership there.

But that’s the thing about doing it again this time round: Zuckerberg was on record as saying people should know you have a FetLife account or the different kinks you’re into because that’s your True Self, and hiding that information is basically dishonest and you’re a bad person. If there’s anyone who’s a bad person here I think we can agree it’s Zuckerberg.

So with hindsight, we should design for the difficult edge cases, and in this case it might actually be that you want as much separation as possible between one identity and another -- that you’d still have an alt or a burner for something that for you has a catastrophic consequence if the relationship is revealed, even if it’s a very low probability of that happening.

(I like to remind people that probability of risk only makes sense if you also take into account the magnitude of the consequence of that probability. Sure it might not be likely, but is it the “I would die” not likely, or is it the “my social security number, which is already out there, would be out there slightly more)

One other thought -- we have some examples of general purpose apps allowing for account switching and showing different context. In most browsers there is a stark difference between private/incognito mode and Regular Mode, practically like dark/light mode. Google Chrome uses different, er, window chrome colors to distinguish between different ur-Google Accounts (my work account in Chrome is purple, my personal one a sort of yellow).

There are different facets to this. There’s the “I want to post to different people in the same app”, for which equivalents would have been Google+ Circles, but more recently posting to Instagram and choosing whether it’s also posted to Facebook. (Again, interesting because those are different space). There’s a loose exploration of this in Bluesky where you’re kind of implicitly posting to different audiences by limiting not who can view the post, but who can reply to it. I’m at a bar having a conversation and it’s possible or likely that someone might overhear it, but it’s really for the people at my table.

I wanted to show through this my thinking out loud about how technical implementation of identity might need to work in conjunction with how people work, or to support safer more useful ways of communicating with people. There are so many layers here, and so many competing/different factors each with their own policy factors and requirements.

Some of these ideas are necessarily re-implementing what’s worked or been tried before. I don’t think there’s anything wrong with that. Tumblr, for example, had a pretty good mechanism for switching between your different Tumblrs. There’s no reason not to use it -- just because Tumblr has (whatever that means) doesn’t mean there isn’t anything to learn from its design. I suppose the point there is to also try to remember our history so we’re not doomed to repeat the same mistakes, or even better, to learn from it and improve.

I think the significant problems here are safety and portability. Portability leads to lower switching costs (moving to different apps) which leads to less lock-in which leads to more competition. If it were easier for Facebook groups to en-masse leave Facebook for other software, then at least they’d have that choice and it deals with part of the cold start problem.

So yes, maybe we need a way for apps themselves to not have singular control over the storage and representation and management of your myriad relationships. Maybe that should be in a notebook of yours, in an open spec, in a box of yours, that you can move to any other storage provider. Now that you’ve got the information about your relationships then you can decide how or whether or what to expose to other apps -- communities, really -- when you want to join or use them.

One point here for anyone who’s not as immersed in the tech world: this whole mess could’ve been stopped with legislation and regulation. It was a governmental policy decision that prevented portability and choice, in the ideological name of market competition -- if a regulation about portability had existed, about reducing the ways in which companies could innovate to develop their moats, then that would have reduced our choice of social networks. If Facebook et al couldn’t have restricted your portability and staved off competition, then why would they bother investing? But then we still managed to get cellphone portability, in the E.U. pretty much from the beginning, and in the U.S. after, well, a while. And even that’s being rolled back thanks to industry lobbying. But hey. Choice!

Only ~2,300 words this time so depending on your outlook, potentially an improvement over the last episode.

How are you doing?

Best,

Dan

Let’s fix things together

Aside from my regular consulting, I also do team workshops and individual coaching based on the workshop curriculum. Get in touch if you’d like to find out more about how to spend that newly reset professional development and training budget you’ve got.

You can also support this newsletter by becoming a paid subscriber!