By nearly every measure Twitter is doing a lot worse than it was before those people were fired. It was growing in revenues and profit and since then has shrunk drastically in both measures.
Now one could reasonably argue that it wasn’t due to the firing. However, the burden of proof lies with the people claiming that the concurrent loss of revenue and profit had nothing to do with the firings.
And Twitter being a private company it’s unlikely anyone can ever get the data to support that claim. Further, Twitter’s unique nature as a vanity project for the owner makes even public statements bh thst owner highly suspect, since they have a vested interest in making this look like the right decision and disclosure laws don’t apply to private companies so he can basically lie and still be ok.
The people who said that had no idea what extent the Twitter tech team had gone to to build a robust application that wouldn't fall over. The fact it withstood 75% of the staff being removed is a testament to the good work they did, not a sign that they weren't needed in the first place.
>By nearly every measure Twitter is doing a lot worse than it was before those people were fired.
Care to share which measures are those you speak of?
From what I saw, Twitter's revenue went down only due to the post pandemic economic slump following the pandemic bubble, and due to advertisers leaving en-masse because Musk won't purge the platform of hate speech and other content that's non advertiser friendly, not due to the workers who got fired leaving.
> However, the burden of proof lies with the people claiming that the concurrent loss of revenue and profit had nothing to do with the firings.
No, if you want to establish correlation or causation, you have to prove it. Don't have access to the data to prove something? Sucks, but then you can't draw that conclusion, at least not definitively.
I can believe, though, that you're at least somewhat correct that the mass firings were responsible for Twitter's decline. But it's plausible that other things are also to blame, and possibly even primarily to blame: questionable product decisions made post-acquisition (Twitter Blue, lax moderation, requiring logins to view, ...), the volatility and offensiveness of the new owner causing advertisers to leave, etc.
From the 2019 annual report (one of the only years where Twitter posted any net profit):
> We have incurred significant operating losses in the past, and we may not be able to maintain profitability.
Since our inception, we have incurred significant operating losses, and, as of December 31, 2018, we had an
accumulated deficit of $1.45 billion. Our revenue has grown from $664.9 million in 2013 to $3.04 billion in 2018. While we
were profitable on a GAAP basis in 2018, we believe that our future revenue growth and our ability to maintain profitability
will depend on, among other factors, our ability to attract new users, increase user engagement and ad engagement,
increase our brand awareness, compete effectively, maximize our sales efforts, demonstrate a positive return on
investment for advertisers, and successfully develop new products and services. *Accordingly, you should not rely on the
revenue growth of any prior quarterly or annual period as an indication of our future performance.*
This is standard boilerplate that every public company puts in their official statements to pre-empt any shareholder lawsuits in case something goes wrong.
Imagine if a company told its investors: “Our growth is guaranteed! Nothing can affect our revenue and margins.” The only companies that give that kind of promises are ponzis (and to-the-moon cryptocurrency projects, which is not categorically too different).
It is actually quite remarkable that Twitter has kept functioning at a technical level. A lot of people expected the massive loss of engineering talent to doom them, but it seemingly hasn't.
However, the site is going through a serious cultural (maybe you could say spiritual) death, and that might have something to do with the lost institutional knowledge of what made Twitter tick at a level deeper than just the code.
On a technical level, I tend to agree. Never underestimate the skill of the employees to save their boss from horrible decisions.
That said, Twitter is no longer "open" like it used to be - I bet it just won't handle the public traffic anymore. That is directly tied to cratering revenue but it's hard to untangle from the owner's self-destructive drug-induced behavior.
It's a testament to the quality of the codebase in general. Good code is hard to kill, but it WILL be eroded to the point where it can no longer grow and be better.
That might be true about the traffic, but twitter wasn’t open before the musk purchase either, it would stop you from seeing posts if you weren’t logged in if you tried to scroll more than a few times. Its current behavior of showing just a few posts sorta, uh, randomly selected might be easier on the server, idk.
For most of its life, Twitter was completely open. You could see any tweet or user page with the URL, whether or not you were logged in. The only exception was if someone had “protected” their account.
Twitter as it runs now is far more locked down. And that happened after it experienced significant, noticeable outages.
Performance is not binary… Twitter is still “up” as a service but with a much smaller public footprint and handling much smaller amounts of traffic.
Do you know when they stopped being frictionlessly open? I’m curious when they started doing the to continue you have to log in pop overs. It preceded musk, as I said. I think every link still worked, but scrolling and navigating twitter like a typical website instead of typing in a link had log in gates.
That is NOT true. As a Twitter addict, I know that was not the case.
Also, Twitter had two notable Spaces problems, when launching the DeSantis campaign (into the ground) and the second fiasco, which I actually forget what it was, just recently. The system just cannot handle truly planet-scale traffic like before.
As someone who wasn’t a twitter addict, I know it was true because people would link me to twitter, I would see the tweet but attempting to scroll too much would result in a log in popover years before the musk buyout.
I'm skeptical, honestly. Within the first few months of the acquisition, it seemed pretty clear to my tech-friend group that Twitter was on its way out and was going to fail. But it's been 2 years, and many of those same people still use Twitter quite a lot. Maybe it'll still fail, of course... it'll just take longer than anyone expected.
I was never a big fan of it, and never used it much, so I can't judge any loss of quality over the past 2 years. And since they now require a login most of the time, and I don't feel like logging in, I don't bother clicking through links to Twitter that people post.
My friends who were passing around twitter links still seem to do it. And the content of those links hasn't changed (random nature/technical info that geeks like) so the twitter posters haven't left either.
The only thing that changes is i used to click on those that looked interesting and now i don't because i know i won't see the thread without a login.
I don't get why whole developer circles didn't leave the platform yet. You can't ask people something via DMs without paying for the blue checkmark and it's totally unhelpful, plus it sends the message you care not enough about the right wing messaging that is send by the platform now. Or these people are just ok with that, I don't know.
They have. Lots and lots of developers are now on the Fediverse. But there's a cultural schism between the types of developers that frequent Silicon Valley cafes and the ones that frequent the Chaos Computer Club.
It's not network effect. There is nothing valuable that happens on twitter that doesn't make it's way out of twitter to the rest of the internet and reality. Avoiding twitter is actually a great way to filter out literal nation-state produced misinfo and propaganda that doesn't pass the smell test, but seems rampant on twitter.
People are addicted to twitter because of FOMO, because god forbid they learn about breaking news an hour later than anyone else.
Software can work for a very long time with minimal maintenance. A different question is if it can keep making revenue without investing.
Quality content has disappeared on Twitter and there is a proliferation of Only Fans tweets. However if you love Musk and right wing conspiracies it is still a fine platform to use.
> A lot of people expected the massive loss of engineering talent to doom them, but it seemingly hasn't.
Not doomed, but lots of failures did follow. Timeline not loading or repeating forever, SMS auth issue, API performance, going down in Australia, etc. - they experienced quite a few problems early post-change, but managed to recover.
When SWEs in tech want to flex, they code Twitter on a whiteboard as an interview question. Major respect for the the work they're doing at xAI, but the hardest challenges building Twitter itself are social.
Uh, it has doomed them. Have you not heard about their massive drop in revenue as advertisers leave?
Sure, the advertisers say its because of the platform being a cesspool. But I mean anything on Twitter is on FaceBook or YouTube; you might have to look faster or harder. However, that stuff isn't where FaceBook / YouTube sends your ads. The targeting on Twitter is so bad compared to alternatives that advertisers just don't care to use it. This is an engineering problem caused by having nobody to fix it.
The revenue problem is mostly because of Twitter's new leadership, on the technical side the few people that remained have managed to keep the platform running quite well.
Sure, it has been plagued by the occasional random errors and downtime ever since Musk came in, but I don't expect most companies to remain operational on a technical level when 80% leaves.
An important lesson for tech companies: hire strategic H1B employees throughout your company so when times are real tough, you're sure you can maintain operations with desperate staff that can't afford to leave.
I was curious if this is true or not. Found this: https://www.investing.com/academy/statistics/twitter-facts-s... which came from before Twitter went private. It's technically correct that they did make a profit (e.g. 2022 Q1), but also made a lot of loss (e.g. 2022 Q2).
Comparatively, it's hard to find information about Twitter now. Most news articles mocks the revenue, but has no information about profit. However, given that Elon said Twitter "will be" profitable in 2024, I think it's safe to assume it isn't profitable yet?
This surprises me, I thought it became profitable from cutting expenses, but the truth is complicated. It used to be both profitable and not profitable at the same time (see the first link), but now it's mostly not profitable.
The massive debt Elon has set Twitter up with is probably driving their current losses. Advertisers wanting nothing to do with him after telling them to fuck off is one thing, but the interest on those billions of borrowed money aren't cheap to hold on to.
What also doesn't help Twitter's case is that employees are still fighting to get paid their severance fees. That's a couple hundred million dollars that Elon probably assumes he doesn't need to pay. Other fees on the order of tens of millions of dollars also include the rent that Elon decided he doesn't need to pay, as well as a bunch of other bunch of unpaid stuff.
My guess is that Twitter is banking on not having to pay all of its debts at once, slowly building up a profit and regaining its value so it can have the cashflow to pay back its arrears later, but that's a risky move that may lead to a debt spiral or bankruptcy.
The spam filtering system sure as shit isn't working. The amount of spam DMs I get is crazy. The trends section has also been busted for ages. Click on the "Show more" link and it shows you less.
- Scrolling after a while, the sound of a video doesn't stop anymore => force close
- A lot of violence, literally. I keep on blocking them ( recently logged in with another account due to a new phone and I had to do it again). It's nuts how much violence it promotes recently.
I'm pretty sure a lot of people who got fired did censorship and I miss it.
I wouldn't read too much into this, as OpenAI is a relatively big company with 3k+ employees (and I'm sure they'll reach 5-10k headcount soon), ie. they have enough firepower to run product experiments like this.
But, what I find interesting here is that this product could be developed by anybody using OpenAI (or other) API calls, ie. OpenAI is now experimenting with more vertical applications, versus just focusing on building the biggest and best models as fast as possible to keep outpacing the competition.
If this is more than just an experiment, which we don't know, that would be a very interesting development from the biggest AI/LLM player.
A lot of people are claiming that OpenAI has no moat, but so far they are clearly the market leader, and "chatgpt" has become an everyday word for LLMs, similarly to how "google" became an everyday word for search 20 years ago.
Why is Google still printing money on Search 20 years later? Surely at this point the know-how to build a search engine at scale is out there. It is a 2-sided marketplace, first they captured people's habits, then advertiser dollars. It could be that in the end LLM usage will also be ad-driven, in which case this will be captured by OpenAI most likely, similar to the Google case.
Another case. Why is Outlook the market leader for corporate email, even though email protocols are open standards, and there is no shortage of open source alternatives, etc. The reason is bundling of course and various other IT considerations, such as trainig/certs/control/security. Imo we don't really know yet how the LLM space will play out, what will enable (or not) OpenAI to win beyond the first years.
Of course there _were_ cases when the moat wasn't there, or was quickly disappeared, eg. Netscape's business melted away as soon as Microsoft bundled Internet Explorer with Windows.
Personally I think OpenAI still has a good 10x growth ahead (eg. 100M paid users for ChatGPT at the $20/mo-ish pricepoint) if they just maintain the current lead on the rest of the pack, and probably the API income can similarly be scaled up. At the slow-moving Retail company I work at, all the execs have been talking about AI for the last 2 years, but we still don't have a single AI feature in production in any of our apps, so we're not yet contributing to OpenAI revenues. But we will soon, as will 1000s of other slow-moving BigCos.
I watch CNN clips on Youtube with some regularity because sometimes it's stupid guilty fun to listen to a bunch of very well groomed "Senior Political Analysts" put on a show of moral outrage over something batshit crazy that Trump said that week.
But I would never pay for CNN. Their programs are high production quality, but the content is not high quality. It's mostly the same pool of center-left talking heads saying not very smart things about the random political happenings of the day.
If I want to know what happened in Israel today, I can just google news it, and maybe click through on BBC or AJ to read some details beyond the headline.
And if I want thoughtful long-from commentary, I will look for 30-60 minute YT podcasts with historians, retired politicians, etc. For example, recently The Rest is Politics has been great, and there are many others.
I'm not a follower of CNN, but I've heard that CNN USA and CNN International are pretty different. In Korea, CNN seems to have a large following because it allows people to get the news while practicing their English.
I have an actively maintained website [1], but it never got me anything. In fact, I even removed GA from it, because I don't care whether people read it or not, I do it for my own enjoyment.
Same for conference speaking, I never got a single good ping out of it (in that case, I stopped doing it, I hate flying).
I'm Hungarian, I hate Orban [1], but your comment is not accurate. We have a lot of problems in Hungary (see my other comment in the thread), but we're not Putin's puppet regime. A puppet regime would imply that Orban was put in place by Putin and/or is kept in place by Putin, and is purely doing Putin's bidding. I've never heard anybody serious say this, nor have I ever seen any evidence of this. Although rhetorically Orban is definitely pro-Russia, in the end (after wasting everybody's time for a while) he always goes along with EU resolutions against Russia.
Orban doesn't need Putin to stay in power, he has access to all of Hungary's tax income, and effectively also beyond that, due to the 2/3 majority which allows him to rewrite the constitution at will.
Likely reasons Orban is friendly with Putin: (i) access to Russian oil (ii) Hungary's one and only power plant is Russian tech run by Russian techs (iii) Orban's regime is like a soft version of Putin's, so criticizing Putin would in some sense mean critizing himself, it'd be counter-productive for him.
> Hungary is widely acknowledged to be in a state of nascent dictatorship anyway.
I'm Hungarian, I hate Orban [1], but your comment is not accurate.
It's not "widely acknowledged" and nobody is seriously using the word "dictatorship" to describe Hungary or Orban's regime, including Orban's fiercest opponents [in Hungary]. We have had various forms of dictatorships in Hungary, and this isn't one of them.
What we have today is most commonly referred to as an authoritarian form of democracy. But we have elections every 4 years, and if people were to vote against Orban, he'd be gone. The last one, in 2022, there were 20k opposition observers in the voting offices and they certified that everything was legit. Having said that there are multiple aspects that make the elections unfair. As we say in Hungary, the "football field is tilted to favor one side". For example, the election map is heavily optimized to favor Orban. Or, state media and 100% of state funds has been captured by Orban to run his misinformation campaigns, which does mean that by the time elections come, a significant % of the population has been indoctrinated and brainwashed. And so on.. But still, it's not a dictatorship. It's just that people are stupid, just as in every country, and unfortunately in 2010 Orban received a 2/3 majority which he used to accomplish the initial "tilt", and he's masterfully kept it up ever since.
Why is Hungary not a dictatorship? Your personal freedoms are not in danger, nobody has ever been killed or even hurt for political reasons in Hungary (unlike Russia). We have freedom of speech, you can go in front of the parliament building and say the nastiest things about Orban, and you'll be fine (unlike in Russia). We have opposition parties, who according to their seats in parliament receive state funding. We have fairly conducted and counted elections (unlike in Russia). We have tons of opposition media and they are free to operate, eg. the biggest Hungarian language political Youtube channels are opposition.
However, Hungary is definitely not a healthy democracy. We have widespread corruption wrt state funds. We have a significant imbalance in media reach in favor Orban. Orban is more than willing to run toxis misinformation campaigns and seed division and hatred that will take many years, possibly generations, to weed out. And, for still unknown reasons, Orban is in bed with Putin and runs Russian misinformation campaigns in Hungary. And so on..
Related to the OP, I can tell you that Orban and his cronies run a variety of misinformation campaigns [in Hungary], but nothing related to cryptography.
I will defer to you as you are Hungarian and obviously a lot more informed than I am, but I will say a lot of the things you bring up seem to be a matter of degree.
From what I can see out there on the internet, it's more than just the media - Orban's cronies seem to have captured the judiciary as well, which helps shield them from investigation. He also seems to be trying various tactics to suppress political opposition - https://www.theguardian.com/world/2023/nov/23/orban-accused-...
I think I'm going to stick by my evaluation of 'nascent dictatorship', though perhaps 'nascent autocracy' is more accurate if it's not all about the one man. And it does seem to be widely acknowledged that democracy has been undermined there, to the point it doesn't function.
> I can tell you that Orban and his cronies run a variety of misinformation campaigns [in Hungary], but nothing related to cryptography.
Yeah fair enough, it seems that a lot of other governments are all in on this sort of thing, and the headline is particularly misleading.
> From what I can see out there on the internet, it's more than just the media - Orban's cronies seem to have captured the judiciary as well, which helps shield them from investigation. He also seems to be trying various tactics to suppress political opposition.
Yes, this is true unfortunately. On the other hand, most of our political opposition [1] was so lame that Orban's baseline tactic was pretty much to just let them do their thing, and that repeatedly led to these parties to self-destruct.
To be clear, the situation in Hungary is shit, but it's not a dictatorship..
[1] between 2010 to 2023, before the new 2024 wave with Peter Magyar and his new Tisza party
> your comment is not accurate. It's not "widely acknowledged" and nobody is seriously using the word "dictatorship" to describe Hungary or Orban's regime, including Orban's fiercest opponents [in Hungary].
Yeah, exactly: in Hungary. The rest of the world is bigger than Hungary, so the terms used in the rest of the world are more wide-spread than those used in Hungary.
We agree that Orban is terrible for both Hungary and Europe. I hate Orban and have never voted for him. In fact, I hate Orban and his system so much that I refuse to pay taxes to him --- the last time I did was in 2015, since then I've been living abroad.
Having said that, Hungary is still not a dictatorship, inside or outside Hungary, per the Democracy_Index [1]. Your country, Finland, with 9.3, has one of the highest scores! Congrats, you have come a long way since the Cold War ended. Hungary has a significantly lower score at 6.7, which makes it a "Flawed democracy". If you check this list, you will find lots of other european countries in the "Flawed democracy" bucket, and with a lower score than Hungary.
"And Hungarians in general as right-wing nutjobs for keeping re-electing him" - I'm not offended by this, but surely you know that no country is a homogeneous set of people. Very roughly, Hungary has about 9.6M people, 8.2M in the voting age, in the 2022 election 5.7M voted, and about 3M of those voted for Orban, unfortunately [2]. (Among them was much of my family I'm afraid.) But you see, there are 5.2M people in the voting pool who did not vote for Orban, so it's not true that hungarians in general are "right-wing nutjobs".
Having said that I have a pretty low opinion of a significant portion of my Hungarian countrymen, but not specifically because they're all "right-wing nutjobs". I would say the root problem is that they (like Trump voters in the US) have access to an infinite stream of information on the Internet, some of it high-quality, some of it low-quality and misleading, and they should be smart enough to tell signal from the noise, inform themselves and have reasonable opinions.
"Why don't you get rid of him while it's still legally possible to do so?" - Naturally you don't follow Hungarian politics, but the way 2024 has been going so far, I strongly suspect that Orban will at a minimum lose his 2/3 majority in 2026, possibly the election. He and his cronies are making a lot of mistakes, there are scandals happening almost every week now. There is now a singular opposition party called "Tisza" which has about the same ~40% support as Orban today (so both are at roughly 40%). The only problem is, this new opposition party is so far a one-man show by a guy [3] who used to be in Orban's Fidesz party. So although he probably wouldn't be as bad as Orban, esp. at the beginning, he would still be right-wing.
> the way 2024 has been going so far, I strongly suspect that Orban will at a minimum lose his 2/3 majority in 2026, possibly the election.
He has a 2/3 majority?!? And won't (possibly) lose it until 2026? How long has it been known that he's a raging nutjob and asshole -- one decade, or several? Possibly losing his huge majority in over a year from now... That's far too little, far too late.
> But you see, there are 5.2M people in the voting pool who did not vote for Orban, so it's not true that hungarians in general are "right-wing nutjobs".
Enough of them are indifferent enough to right-wing nutjobbery not to vote against Orban's right-wing nutjobbery, which makes them enablers of him and it.
Over time I've come to think that all peoples are in the end responsible for their own leadership: If a majority of the people is against oppression, then they should vote against it. If they can't vote, they should rise in revolution. If they don't, they're not sufficiently against it to warrant being called civilised.
That's why Germans carried (and to some extent still carry?) a collective responsibility for Nazism (and anew do so for the AfD and BSW nutjobs they're voting into power nowadays); why Russians carry a collective responsibility for Putin (and before him, for Stalin. And Ivan IV); and why Hungarians carry a collective responsibility for Orban.
I'm hungarian, but I used to live in Irvine in the 90s, specifically on Turtle Rock. I went to University High School. I think Irvine is a prototypical rich US residential town: very boring, lots of doctors/engineers/lawyers, 2-3 car garage standard houses (like in the Simpsons), never meet the neighbours, you need a car to get anywhere, etc.
Nothing wrong with that really, can be appealing for adults in their 40s (like me today) but as a teenager I was bored out of my socks. I would probably not move to a place like that, I'd want my kid to have more stimulation. But happy to read that Irvine is doing well, I still have a lot of friends there.
Almost all of them have Python code to illustrate concepts.
-
1. Entropy of a fair coin toss - https://bytepawn.com/what-is-the-entropy-of-a-fair-coin-toss...
2. Cross entropy, joint entropy, conditional entropy and relative entropy - https://bytepawn.com/cross-entropy-joint-entropy-conditional...
3. Entropy in Data Science - https://bytepawn.com/entropy-in-data-science.html
4. Entropy of a [monoatomic] ideal gas with coarse-graining - https://bytepawn.com/entropy-of-an-ideal-gas-with-coarse-gra...
5. All entropy related posts - https://bytepawn.com/tag/entropy.html
reply