Bluesky's science takeover: 70% of Nature poll respondents use platform

I don&#x27;t understand all the comments promoting the openness of twitter&#x2F;x. There&#x27;s a ton of friction to interact with X without an account. Meanwhile, the alternatives mostly allow third party apps, cross posting, and logged out access.<p>I can follow mastodon and threads accounts on my bluesky timeline using one of many bridge services: <a href="https:&#x2F;&#x2F;fed.brid.gy&#x2F;" rel="nofollow">https:&#x2F;&#x2F;fed.brid.gy&#x2F;</a><p>I can merge mastodon, bluesky, nostr, and threads one timeline using one of many third party apps: <a href="https:&#x2F;&#x2F;openvibe.social&#x2F;" rel="nofollow">https:&#x2F;&#x2F;openvibe.social&#x2F;</a><p>I can unsubscribe from the default bluesky moderation service and choose one run by a third party: <a href="https:&#x2F;&#x2F;bsky.social&#x2F;about&#x2F;blog&#x2F;4-13-2023-moderation" rel="nofollow">https:&#x2F;&#x2F;bsky.social&#x2F;about&#x2F;blog&#x2F;4-13-2023-moderation</a><p>The only differentiator for X is the content, and that&#x27;s an easy choice.

BlueSky has its problems, but the problems are more bearable than Twitter. You can have conversations about science and technology without weird angry people&#x2F;bots popping up in the replies and shouting at you.<p>Twitter used to be great that way as well, but now it isn&#x27;t. Losing those communication channels is a big loss, no matter what you think of social media.

Unfortunately, this is a result of twitter censorship directly caused by harassment of the science community. A specific type of censoring people that is often ignored. In this way, Twitter practices more censorship than most platforms!

A bit off topic, but does anyone know why there&#x27;s only one Bluesky relay, or can direct me to an article? Even if the hardware requirements are high, I&#x27;d think someone would&#x27;ve made a second relay by now.

Can anyone recommend good science or ml people to follow on bluesky? I&#x27;ve moved from twitter, but haven&#x27;t quite found as interesting content yet, especially with ML - seems like they are still posting on twitter

You can read Bluesky posts without an algorithm, just like the good old days.

But at some point they will have to interact with people outside the hermetic filter bubble(s) of BlueSky - even if it relates more to the results and byproducts as opposed to the raw research.

This is in line with the ideal scenario of Twitter remaining online as a quarantine for spam and culture war stuff, but real traffic migrating slowly elsewhere, hopefully without any single &quot;winner&quot; ever becoming big enough to attract what was left behind.

Important to note:<p>&gt; Although the survey is not statistically representative of Nature readers or the scientific community at large<p>I found that the survey seems to have been advertised for with &quot;Has Bluesky replaced X for scientists?&quot; according to this article from Jan 14 [1]. This attracts people who switched more I think. So the 70% figure is nothing than hot air.<p>It is still interesting seeing the positive sentiments and reasons for switching analyzed.<p>[1] <a href="https:&#x2F;&#x2F;www.nature.com&#x2F;articles&#x2F;d41586-025-00037-y" rel="nofollow">https:&#x2F;&#x2F;www.nature.com&#x2F;articles&#x2F;d41586-025-00037-y</a>

Thanks for marketing push for bluesky via nature!

Being in academia (or very adjacent to it), there are two types of BlueSky users:<p>1) The ones that moved because of ideological reasons 2) The ones that moved b&#x2F;c Twitter became a desolate hellscape.<p>I&#x27;m closer to (2) than to (1), and the difference between 2024 and 2020-2022 twitter is stark. Before, if you had an interesting tweet about a new article you would get lots of engagement (retweets, discussions, people disagreeing and pushing back).<p>Now? Nothing, just a few random retweets or comments from non-academic people that almost feel LLM-generated. Compound this with network effects, and even my right-wing pro-Trump colleagues (and there are some) decided to at the very least dual-post.