4 ms·

> and at least half of those 600 million are bots, I don't think anyone disagrees Not that I disagree (or agree) but I would like to know (out of curiosity) ho

> and at least half of those 600 million are bots, I don't think anyone disagrees Not that I disagree (or agree) but I would like to know (out of curiosity) how would you (or the source of this information) got the number > 300M bot accounts. Do this include inactive accounts? Does this include people with multiple accounts? And how would you determine that without access to the database? Is there a tools that can help you derive this number based on some statistical analysis? I would be very interested in the details. Hint: I am not trying to start a debate around X policies or Musk. I just find that it could be interesting the way we can get into this number if it is supported by any rigorous methodology.

ck2 2y ago
there have been several serious studies over the years trying to determine the bot count but of course the api lockout by Musk now makes it impossible * https://www.pewresearch.org/internet/2018/04/09/bots-in-the-twittersphere/ https://www.pewresearch.org/internet/2018/04/09/bots-in-the-... * https://www.theguardian.com/technology/2023/sep/09/x-twitter-bots-republican-primary-debate-tweets-increase https://www.theguardian.com/technology/2023/sep/09/x-twitter... * https://www.google.com/search?q=study+of+bots+on+twitter https://www.google.com/search?q=study+of+bots+on+twitter but it is clear the signal-to-noise ratio on twitter is now unmanageable especially after tools like blocking are taken away
- elashri 2y ago
  That's interesting although from a quick skimming it seems that these studies are focused on the activity of the bot accounts. This might give a hint on their relevance but due to the fact that their activity is probably above average it doesn't tell much about their percentage of accounts (in absolute numbers). So I am not sure if there is a studies that shows that something near 50% of accounts would be bots. The number is huge and that's why it is interesting for me to see if it is true and how could we determine that. But I also don't think that the blocking change here would affect signal to noise ratio. You still cannot engage with accounts who blocked you but can see their public posts which you could do by logging out (before X required account to show tweets) and create alt account for that. Blocking access to the already public posts/tweets didn't make sense anyway. And I don't think this is even restricted to X or any particular website.

ck2 2y ago

there have been several serious studies over the years trying to determine the bot count but of course the api lockout by Musk now makes it impossible * https://www.pewresearch.org/internet/2018/04/09/bots-in-the-twittersphere/ https://www.pewresearch.org/internet/2018/04/09/bots-in-the-... * https://www.theguardian.com/technology/2023/sep/09/x-twitter-bots-republican-primary-debate-tweets-increase https://www.theguardian.com/technology/2023/sep/09/x-twitter... * https://www.google.com/search?q=study+of+bots+on+twitter https://www.google.com/search?q=study+of+bots+on+twitter but it is clear the signal-to-noise ratio on twitter is now unmanageable especially after tools like blocking are taken away

elashri 2y ago
That's interesting although from a quick skimming it seems that these studies are focused on the activity of the bot accounts. This might give a hint on their relevance but due to the fact that their activity is probably above average it doesn't tell much about their percentage of accounts (in absolute numbers). So I am not sure if there is a studies that shows that something near 50% of accounts would be bots. The number is huge and that's why it is interesting for me to see if it is true and how could we determine that. But I also don't think that the blocking change here would affect signal to noise ratio. You still cannot engage with accounts who blocked you but can see their public posts which you could do by logging out (before X required account to show tweets) and create alt account for that. Blocking access to the already public posts/tweets didn't make sense anyway. And I don't think this is even restricted to X or any particular website.