Beyond Platforms: Private Censorship, Parler, and the Stack

So this is just an aside in the article but they mentioned a new development in Uganda, that gave me pause.

Two days ago "Ugandan government blocks Google Play Store, Apple App Store, and YouTube (techjaja.com)" [0] Now "Facebook has shut a slew of accounts belonging to Ugandan government officials"[1]

Retaliation?

[0] https://news.ycombinator.com/item?id=25704433

[1] https://www.france24.com/en/africa/20210111-facebook-shuts-d...

This makes sense to me. I also take it to mean we should give "core infrastructure" sites like CloudFlare more leeway to continue protecting / not crack down on content that we may not agree with than a top-of-the-stack platform like Facebook.

This doesn't really seem to track with reality, and is a really disappointing distinction that the EFF is trying to make. In fact, there is way less competition at the "top" of the stack than the "bottom": there are many many many places I can take my personal website to other than AWS; it might cost a bit more or be slightly harder to use or be a bit slower, but I can make my site work almost identically on any "infrastructure" provider, because there are just so so so many possible points of connectivity. I love AWS, and have used it almost exclusively for all of my infrastructure deployment pretty much since it came out... but if I were to be banned from it tomorrow I'd probably be able to find another hosting provider somewhere. Hell: The Daily Stormer is still online somehow, accessible to the same audiences with the same tooling (at most with a different TLD).

However, if I were banned from Facebook, I'd be screwed... all of my friends are on Facebook, all of the University community is on Facebook, all of the local events are coordinated on Facebook. I am a local government official, and really the only marketing that our small government does that matters at all is on Facebook, as that's where our entire constituency "lives". Even in a microcosm: there's a Facebook Group called UCSB Free and For Sale, and I know someone who's banned from it... that ban alone means that discussions at our local meetings are really problematic for them as so much local discussion happens on that group (a really high-profile example that came up recently was discussion of someone who was shooting into a local building, which prompted discussion with the local police... this one person couldn't even see the posts I was referring to, as they were banned).

In that case, Twitter isn't a competitor to Facebook any more than the Google Play Store is a direct competitor to the App Store because the "relevant addressable market" is different: that network effect is really important. It is this overlap of closed ecosystems and aftermarket monopolies that makes companies being able to arbitrarily ban people so scary and so annoying to so many people. The reason the Internet works is because it is "the Internet" and not "a network"... being banned from AOL would be way worse than being banned from AWS, because now there are people whom you simply never will be able to contact... likewise, if I were banned from AT&T, I can still call AT&T customers from a Verizon phone (and vice versa); but if I'm banned from calling AT&T customers by AT&T then there are people in the world I can never call. This is the kind of distinction that needs to be worked on, not this subjective one about "infrastructure".

I mean, this is so weirdly circular and subjective that they then define Zoom as infrastructure, but not Twitter? The reason they cite is that Zoom doesn't normally do a ton of moderation... I mean, if you define something as infrastructure simply because it has historically chosen to be infrastructure, that's kind of begging the question, right? I think the big argument everyone is having here is actually whether Facebook should be infrastructure: half the people want Facebook to moderate less, and half the people want Facebook to moderate more; meanwhile, it isn't a given that Zoom shouldn't moderate and Facebook does: that's just an historical artifact! The people who like that Facebook moderates by and large also want Zoom to moderate, and the people who don't like that Facebook moderates by and large want Zoom to continue to not moderate... these are consistent positions.

The EFF trying to push this artificial distinction between Zoom ("infrastructure") and Twitter ("service") frankly just seems to be some attempt to force their stances on network neutrality (where they like to claim companies should not be able to discriminate traffic, whether or not it is profitable for them) and section 230 (where they specifically like to defend the right of companies to moderate haphazardly, defending the idea that no company should be required to do something that is unprofitable for them) to be a bit more generally consistent... but it is going to take a much longer explanation than this to make that make any sense (and I'd really like to see this argument be done as some kind of debate, as otherwise it is too easy for them to make false statements about relative competition in different market segments, which is what they seem to be doing here).