Bluesky now platform of choice for science community (opens in new tab)

If the poll was done _properly_, that sample size is _fine_; there’ll be a decent margin of error, but not as much as you might expect. 1k people is a fairly standard size for polls, with even very high quality ones rarely doing over a few thousand.

The real consideration was whether the poll was done properly.

What is the sampling bias? You dont explain what it could be, and your quote doesn't give any clues about what the bias could be.

There's exactly three reasons for people to stick to Twitter:

* They don't care/agree with the policies of the guy running it.

* Legacy reasons; either they have no reason to leave (automated org accounts keep running until something in the workflow breaks) or they have an existing community that doesn't want to move. This group will eventually leave but is currently stuck with inertia. Most "public service" accounts are in this category.

* And finally, for artists, Bluesky is undesirable as a platform because it has some very aggressive image compression compared to Twitter (2000x2000 is the absolute limit). Some are dualposting to Bluesky, but are unlikely to fully leave Twitter for this reason.

Finally, I'll note that while accounts are generally abandoning Twitter, this doesn't automatically mean they're moving to Bluesky either. A lot of those service accounts just up and vanished and just said "well, go visit our website".

X really sucks in it's current state, but it's where the things I'm interested in happen or are discussed first (eg AI state of the art, bootstrappers). There's a bunch of tech people I follow who aren't on BlueSky, Threads, etc.

Interestingly, when I glance at my Bluesky feed once a month or so, it's a lot of complaining about everything. I think I hear more about Elon on Bluesky than I do X. And yeah, I follow reasonably high-value people.

That said, I keep some sort of X exit plan in place, and I look at it a lot less than before. When the signal vs noise value shifts, I'll be done, but I'm not quite there yet.

It's surprising that any serious organization used it at all. It was never a good place to spend your time really.

It's sad that the science community is just moving to another walled garden rather than spawning its own network of federated ActivityPub services (eg: mastodon).

Bluesky seems to be based on an open protocol (AT Protocol), but how actually interoperable is that ? I can't find a list of non-bluesky AT protocol servers that can interoperate with Bluesky.

No true scotsman's fallacy - no serious person would use X; if you use X you aren't serious.

I don’t know that “AI hostile” is quite accurate. It’s more like it’s not blindly accepted and praised, which is different. Generally any hostility comes down to cases where it’s replacing or could replace people instead of bringing a new capabilities to the table — AI helping detect cancer or discover new treatments for example is a lot more accepted than AI taking the jobs of book cover illustrators, for example.

That… kind of makes sense? It’s logical that applications with significant downside, particularly that which impacts peoples’ livelihood, would get greater questioning and pushback. If anything I’d call into question a platform where nobody is asking these questions and wants to charge ahead with zero regard to potential ramifications.

I don't have that experience. But perhaps I am far less forgiving of the christofacist community on twitter that seeps through EVEN though I follow not one of them and none follow me. It's easier to read past the artsy crowd for me at least even if I don't agree. But AI hostile? Guess it depends if you don't post AI 'artwork' all day but just interesting things you are doing then I have not seen anyone complain.

That's been my experience of most Mastodon instances, too (except the ones explicitly created to counter that, which don't tend to get much traction).

I follow some actual far-leftists on Mastodon, and they would probably find your assessment of Bluesky to be pretty funny.

im there, I use Claude every day, am not a far leftist. come on over and there will be two of us.

You forgot to add "yet" unless you can predict the future will 100% accuracy. Only once it's sold to someone else can you really say he didn't do it.

The data's longevity was probably helped by being a potent source of hate to power Musk's AI

so deeply controversial when a guy doing nazi salutes in front of crowds and illegally dismantling key pillars of US foreign policy finds himself unfavored by the media

X hides discussions. If somebody sends you a link you can't see the discussion without an account and logging in and being extensively tracked.

Not sure what you mean about anti-Elon bias. This was straightforward reporting of the truth. If reality has an anti-Elon bias then perhaps it's not bias.

owned by conde nast its just normal MSM stuff

Yeah just reading comments on Arstechnica articles like this one you can see the same baying mob they have on Bluesky. Elon bad. X nazi. Meanwhile I closed my Bluesky account because every other post is calling for violence against political opponents, gleeful over assassination and unrest and the recent beating of a doge employee.

It's so odd that they still get to enjoy their status as a "decentralized" "open" social network when it's really just a centralized platform by definition (and it's not even their fault, there is simply no way to do what they want to do without centralization).

Sure, someone may say "the AT Protocol is open", but that means nothing because the AT Protocol is not Bluesky, Bluesky is one centralized platform that happens to "talk" that protocol (well, of course, since the "protocol" is literally defined by whatever they happen to be doing), it still controls who can be inside and who can't.

TL;DR: Nostr is a much better option for most use cases, sadly for some unfortunate reason it never got to enjoy too much attention from a wider technologist community.

If you pull out to view the full history, you'll see it's a constant series of huge spikes (e.g. in late 2024 the site more than tripled in size over about a week) each followed by a slow trend downwards, then another huge spike up further, etc.

I suspect that's pretty common for something that's been in the news quite a bit: you get occasional big jumps in attention & usage, and then only some smaller percentage of users will stick around longer term. When you're getting such big spikes in signups this is unavoidable I think - even with new users coming in, the descent from the spike overwhelms any other trends.

The interesting question is whether that settles down into a slow steady sustainable state eventually. Looks plausible but still unclear imo.

This is extremely interesting. Didn't know of these stats. Would be great to be able to have twitter, mastodon and bluesky on the same graph (perhaps with different y-axes to allow comparison of their relative trends)

Any news article that could be interpreted as critical of Musk these days seems to have a broad army of commenters that swarm to add low information critique of critique. This article is likely to be flagged off the front page very quickly too.

Edit: yes, and as predicted, this straightforward reporting on the scientific community's migration of social networks has been flagged off the front page, because too many HN folks can not distinguish bad news for Musk from biased against Musk. There's no reason for this to be flagged except for the political motivations od the flashers.