In this case, the third option is what has happened.

That's good advice.

What did you think an open protocol meant? Vibes, papers, essays?

Losers.

They should bring him back if he removes PII from the dataset.

Although what's once published, can't really be unpublished. So I dunno.

BlueSky gets away with firehose broadcast by having a ToS for all consumers to be responsible to honour EU, California and Brazil privacy laws.

I think the assumption here is that more speech leads to better speech, while I find that the greater a speech's audience, the more extremist it tends to be.

Especially with the intense war to win the human attention span done across platforms and within.

Well, the default BlueSky app was always just the "reference" implementation created as more as a dev tool.

No reason why alternate apps can't be launched, although the AT proto does have usage compliance guidelines that stop half-hearted attempts.

newsletter.pragmaticengineer.com/p/bluesky

I suppose the data needs to have PII removed as a necessary condition for republishing data.

This way, there's the interaction graphs but nothing personal.

eg: This study is funded by EU - Horizon Programme, so I presume someone in ethics has approved it.

paperswithcode.com/dataset/blue...

Technically impossible to get banned from the AT protocol, but can be restricted on the primary BlueSky frontend apps.

🇺🇸