Twitter was a valuable platform for scientific discourse in the past
jrhawley.caThis article makes excellent points in a cogent fashion.
I partly disagree with the author that simply pointing at the existence of an alternative is sufficient. One of the reasons I've been very lukewarm about Mastodon is that it really lack the networks effects that make Twitter special; federation is a nice idea but the pragmatic benefits are a lot less clear. Being able to follow, track, and have conversations with individual scientists/scholars on Twitter has been a huge benefit for me, and it's not obvious that federated social networking can reliably deliver that. Another issue is that while Mastodon started well out of the gate 5 years ago (which was when is first signed up for it), very little has happened since then. 'We're not those other guys' is not a sufficient recipe for changing the world.
An excellent point that I do agree with is how 'sticky' Twitter is and how (like many other big tech firms) the tools it gives you when you export your data aren't really that helpful/useful unless you have sufficient programming skills of your own to overcome the quirky formatting issues. It seems like there's an audience for a tool that leverages the Twitter API to scrape your following/follower data into a convenient format and perhaps automate the business of finding and reconnecting with those people on another platform.
I think it's reasonable to say Twitter's utility is rapidly waning, both as described and with each new day's manufactured drama. However, the network effect issue is a big one. If 'science twitter' decamps to 'science.social' it could quickly find itself effectively cut off from its public and derided by antagonists as a 'woke echo chamber populated by high IQ stupid people' to borrow a phrase from what passes for political discourse in 2022.
> it's not obvious that federated social networking can reliably deliver that
Why not? It's federated, not isolated. If enough people are available to achieve the network effects you're after, what would you lack? (Content discovery solutions are already being created for people who want to play with them)
> It seems like there's an audience for a tool that leverages the Twitter API to scrape your following/follower data into a convenient format and perhaps automate the business of finding and reconnecting with those people on another platform.
https://twitodon.com/ for mastodon already exists
> If 'science twitter' decamps to 'science.social' it could quickly find itself effectively cut off from its public
Why? What's different from frontend social, infosec exchange, etc.? In the general population is mastodon, I can't imagine science twitter being called out as woke.
Of all the tools for helping find your follows in the fediverse, I hope https://movetodon.org/ gets more attention. It is easier (no “download and upload this CSV”) and also serves a purpose over time to discover new migrants.
The site could do with some text explaining what it does, who operates it and a privacy policy etc. rather than just starting off by immediately asking for permission to access my Twitter account. Looks as shady as ... as it is.
>If 'science twitter' decamps to 'science.social'
There's fediscience.org and sciencemastodon.com already (Sean Carroll created account on fediscience.org and James Gleick - on sciencemastodon.com among people I follow on Twitter).
Author here, thanks for your thoughts.
I agree with you, just the existence of Mastodon/Discourse/others isn't sufficient. The onboarding and home/local/federated timelines of Mastodon can be a bit confusing at first, which does take some time to get used to. Just not being the other guys is the first of many necessary steps to being better.
And yes, scientists can't be entirely insulated from the wider public. Ideas that only stay in academia don't make the same impact that they need to in the wider world. Similarly, academics need to hear from the wider world to learn what problems are important and how their work can impact people. There need to be forums for interacting with large public audiences.
> federation is a nice idea but the pragmatic benefits are a lot less clear
> it's not obvious that federated social networking can reliably deliver
I love the idea of federations / federalism.
I also look at Mastodon and neither want to run my own instance or carry the mental ability to remember what instance I signed up on. Defending the idea of federalism but acknowledging that Mastodon isn't a perfect replacement for Twitter, I'd like to modify your statement:
Federation is a nice idea, but federations are only as strong as their members.
The Mastodon federation, combined, is smaller than Twitter.
So Twitter, while not being a federation, still has a better network effect.
One of the things I noticed and immediately liked about Mastodon was that I quickly found curious academic types who are interesting to chat with. And the whole vibe on Mastodon was different because you'll find quirky and idiosyncratic people who pay a lot of attention to a very specific type of intellectually curious conversation, and your more likely to discover that on Mastodon in my experience, so I think it is an excellent fit for academics.
I joined Twitter some years ago with the only interest of growing my network of functional programmers. I specifically only followed people who do Haskell, Scala, or some kind of abstract algebra. I eventually quit because I don't subscribe to platforms that promote cancel culture. I did try Mastodon, but didn't discover as many interesting people before I lost interest.
I haven't used Mastodon and so the impression I get is that it doesn't have an easy way to look up other accounts or discovery of accounts on other instances.
How does user account discovery work? How do you think it could be made better?
I found DeviantOllam's introduction on Mastodon/Fediverse[1] to be quite interesting and useful.
He covers contact discovery/migration from Twitter, and also mentions a fediverse contact search, although I forget what it was.
Seconding this
Two ways: find instances relevant to your interests a la Usenet If you already know the person's name, use the search field.
[Edit] this is not how to improve it. This is how discoverability in Mastodon currently works.
They should have left years ago, rather than tolerate the decades of sewage and outrage around the platform. The problem is some think the solution is to replace it with a 'federated' outrage machine (Mastodon) which discoverability is an eternal problem with instance moderators banning entire instances for any reason if they wanted to. For the example of journa.host [0].
The same can easily happen for a Mastodon instance for academics. The whole point of Mastodon is to own and self-host your own instance. If these journalists, academics, artists, etc are still not able to do that and are joining centralized instances; it is no better than being on a worse version of Twitter, but with a significantly limited reach and not truly owning your accounts.
It is fine to believe the delusion that Twitter will immediately be falling over 'any minute now'™ for 100% of users. But if I had to choose where to focus advertising my work on either Mastodon or Twitter it will always be Twitter; both before and after Elon taking over. The reality of the point is, there is still no viable alternative to Twitter that has the same reach and features that make it convenient for many to use.
But this time round, the ToS applies to everyone equally, scam bots are invisible in replies, normal bots are labelled as automated and tells you who owns it and much more. The exact opposite of what I have heard from the 'doomsters' and the screaming minority spreading misinformation about a so-called 'Twitter apocalypse'.
[0] https://twitter.com/ajaromano/status/1594432548222152705
Mastodon's at the "email in 1994" stage--the general public is starting to become aware that it exists, and folks who were running small servers for their own purposes suddenly find themselves in a position to earn social capital by handing out accounts. Obviously in a few years at most, this will get tedious for the admins providing all that undercompensated labor. I think the winning scenario will be old-school webhosts including a personal Mastodon instance with their default package, like email. If that happens, Big Tech adoption will follow and we're quickly talking about a real standard with a lifespan of decades, if not more
Twitter is just as bad as it was before Musk took over with one exception: they removed the blatant partisan censorship. If that makes it worse than before for academics there is more at hand than it just being a viral meme-amplifier unsuited for academic discussions.
You’re absolutely correct. Taking the breaks off of American anti-science wingnuts is indeed the “more at hand” that has made it worse than before.
That is not that much of a problem, is it? Any annoying American anti-science wingnut (and what about non-Americans, are they not annoying?) is easily blocked so in that respect not that much has changed. The widening of the Overton window will bring in more of those who previously were censored but that does not make them all - or even most of them - anti-science wingnuts. They just hold opinions which hitherto were deigned off-limits on ideological grounds. As far as I'm concerned this should improve the quality of the discourse instead of diminish it since ideology tends to be more of a hindrance than a help there. That is what the scientific method is about after all: careful observation, applying rigorous scepticism about what is observed, given that cognitive assumptions can distort how one interprets the observation. I want my observations to be scrutinised by those who look at things from a different perspective because they may see problems which I'm likely to overlook. The opposite is true as well, I will see holes which they are likely to overlook.
The bots are also removed. It's great to see human answers again even to the most provocative tweets (about help with Metamask and ETH just as an experiment)
That didnt happen.
I don't know and you don't know but xiphias2 seems to know. Can you (xiphias2) describe your 'experiment' and its outcome?
If the bots are 'gone' - are they already, all of them? - that would certainly make it a better place. I'll still not get an account as I find the format not conducive for the type of conversation I'm wont to engage in but it is clear that a bot-free (or at least bot-labelled) and bias-free (as far as possible, there will still be a weird [1] bias but I can live with that) Twitter is less of a problem than the biased bot-hive it used to be.
I was looking at this Tweet:
https://twitter.com/coinbureau/status/1602172317022445569?t=...
I don't have more data, and I'm sure there are more bots, but the replies look like they are from humans which is very different from a few months ago.
It is still biased and it is still filled with bots. It will most likely continue to be these 2 things. If anything, chatGPT will make sure its more of both moving forward.