4 ms·

Your argument seems to be entirely financial rather than what twitter _does_. My understanding is their MAUs and DAUs have fallen. Fewer people are looking at i

by abeppu 2y ago

Your argument seems to be entirely financial rather than what twitter _does_. My understanding is their MAUs and DAUs have fallen. Fewer people are looking at it or posting on it, less often. https://soax.com/research/twitter-active-users https://soax.com/research/twitter-active-users https://mashable.com/article/twitter-x-daily-active-users-drop-under-elon-musk https://mashable.com/article/twitter-x-daily-active-users-dr... I think it's interesting that you mention bloomberg as saying that things are going ok. My recent read was that they thought twitter was just messing with numbers. https://www.bloomberg.com/news/articles/2025-02-20/behind-musk-s-x-turnaround-pick-and-choose-financials https://www.bloomberg.com/news/articles/2025-02-20/behind-mu... Key quotes: > its improved fortunes owe much to heavily adjusted financials and investors’ fear of missing out > The 2024 figures weren’t audited, but the 2023 figures were, said one of the people. None of them would qualify for generally accepted accounting principles, also known as the GAAP standard that the US Securities and Exchange Commission requires for publicly traded companies. ... i.e. the recent numbers aren't trustworthy > Even under those unusual calculations, X’s debt burden was roughly nine times adjusted earnings — far beyond the baseline of six that bank regulators characterize as risky. > The threat of litigation — and Musk’s advisory role to Trump — has been a topic of conversation among some advertisers as they decide whether to return, according to several ad industry sources. While few believe X’s product or ad offerings have improved in the past year, marketers are weighing the possible downside of ignoring X and what that might mean in terms of retaliation from Musk. ... i.e. to the extent that some advertisers have returned, it's b/c they think they'll be punished if they don't, not because it's a good way of reaching an audience or promoting their brand. I think the actual revenue story is worse than the usage story. My first searches show that their Q4 revenue was $2.5B, roughly half of where it was in 2021. https://www.businessofapps.com/data/twitter-statistics/ https://www.businessofapps.com/data/twitter-statistics/ Analogously, I don't think any of the recent sudden moves mean that the federal government at large is _over_ ("doomsday scenario") except for the departments that the administration actually wants to kill, but I think the current administration is setting the government up to do less and less well. Possibly it will be a great financial outcome for Musk. (edited to add missing bloomberg link)

mirekrusin 2y ago
My argument is around cutting workforce and requirement to be present in the office in context of efficency in general. Sentiment of people predicting total failure back then feels like flashback now.
- pseudalopex 2y ago
  I saw few predictions of rapid collapse. Many predictions of slow or delayed reductions in revenue, active users, content quality, service quality, and ability to deliver new features. abeppu's argument was efficiency and health are not the same.
  - mirekrusin 2y ago
    One off, politically driven reductions in advertisers and user groups is not caused by unhandled load from reduced workforce or implications of work from office policy. Despite this out migration, so far their health looks exceptionally good. They're running at 20% of original workforce (!). Just think about it (and I really mean it - taking aside personal dislikes, political views etc) – you have a company, you cut 80% of workforce and it still seems to work very well. This is insane. It's also not one off example - he is well known for speedrun stunts (ie. e2e completion of colossus supercomputer in memphis in 4 months <<!>>). The argument is – can you do something similar with government as well? As a body that is notorious for inefficiency it seems plausible the answer is yes.
    - abeppu 2y ago
      > One off, politically driven reductions in advertisers and user groups is not caused by unhandled load from reduced workforce or implications of work from office policy. I think this is basically backwards; aggressive cuts and short-term revenue attempts alienated advertisers, casual users and influencers with large audiences. In particular: - they abandoned content moderation, including flagging misinformation, hate-speech, and targeted harassment. This obviously saves money, b/c it eliminates systems and ML models that needed to examine tweets/messages and accounts in real time. These are adversarial problems, so you can't just create the model once and leave it alone; you need to pay a team to keep finding the current abusive patterns. But abandoning this means that you have a network where (a) any post you see is more likely to be manipulative misinformation (b) you will see slurs and hatespeech and perhaps receive them and (c) if you're determined to be in some way objectionable to an army of trolls, you may be endlessly harassed. So it's a kinda awful place to engage. - they tried to sell blue checkmarks. This had multiple bad effects. It let any rando with a few dollars impersonate large companies and brands. This alienated those large companies who felt understandably burned, and it made casual users less trustful of content even that appears to be marked as legitimate by the platform. Further, b/c they were trying to force previously verified accounts from notable people to start paying them, they alienated celebs and influencers with large audiences. It's not just that people who disagree with Musk politically were moving away from twitter in some slow-rolling boycott -- the choices Musk made have transformed twitter to a place where it's harder to get real information, it's hard to have a conversation with your community without being attacked, it's hard to curate your feed (Musk posts may be injected whether or not you follow him), and far from being a "free speech" paradise, it's one where a bunch of people don't feel safe engaging at all.

by abeppu 2y ago

mirekrusin 2y ago

My argument is around cutting workforce and requirement to be present in the office in context of efficency in general. Sentiment of people predicting total failure back then feels like flashback now.

pseudalopex 2y ago
I saw few predictions of rapid collapse. Many predictions of slow or delayed reductions in revenue, active users, content quality, service quality, and ability to deliver new features. abeppu's argument was efficiency and health are not the same.
- mirekrusin 2y ago
  One off, politically driven reductions in advertisers and user groups is not caused by unhandled load from reduced workforce or implications of work from office policy. Despite this out migration, so far their health looks exceptionally good. They're running at 20% of original workforce (!). Just think about it (and I really mean it - taking aside personal dislikes, political views etc) – you have a company, you cut 80% of workforce and it still seems to work very well. This is insane. It's also not one off example - he is well known for speedrun stunts (ie. e2e completion of colossus supercomputer in memphis in 4 months <<!>>). The argument is – can you do something similar with government as well? As a body that is notorious for inefficiency it seems plausible the answer is yes.
  - abeppu 2y ago
    > One off, politically driven reductions in advertisers and user groups is not caused by unhandled load from reduced workforce or implications of work from office policy. I think this is basically backwards; aggressive cuts and short-term revenue attempts alienated advertisers, casual users and influencers with large audiences. In particular: - they abandoned content moderation, including flagging misinformation, hate-speech, and targeted harassment. This obviously saves money, b/c it eliminates systems and ML models that needed to examine tweets/messages and accounts in real time. These are adversarial problems, so you can't just create the model once and leave it alone; you need to pay a team to keep finding the current abusive patterns. But abandoning this means that you have a network where (a) any post you see is more likely to be manipulative misinformation (b) you will see slurs and hatespeech and perhaps receive them and (c) if you're determined to be in some way objectionable to an army of trolls, you may be endlessly harassed. So it's a kinda awful place to engage. - they tried to sell blue checkmarks. This had multiple bad effects. It let any rando with a few dollars impersonate large companies and brands. This alienated those large companies who felt understandably burned, and it made casual users less trustful of content even that appears to be marked as legitimate by the platform. Further, b/c they were trying to force previously verified accounts from notable people to start paying them, they alienated celebs and influencers with large audiences. It's not just that people who disagree with Musk politically were moving away from twitter in some slow-rolling boycott -- the choices Musk made have transformed twitter to a place where it's harder to get real information, it's hard to have a conversation with your community without being attacked, it's hard to curate your feed (Musk posts may be injected whether or not you follow him), and far from being a "free speech" paradise, it's one where a bunch of people don't feel safe engaging at all.