Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Elon & Twitter

This site may earn commission on affiliate links.
Status
Not open for further replies.
Last night I spoke with a Facebook employee who survived the recent 11,000 person layoff. He had a view that I thought was interesting. I asked if he was relieved about keeping his job and he was relatively ambivalent about it. That surprised me, because he's a pretty high earner. When I pressed about it, he told me that very few people who work in social media are passionate about their job, because, well, it's social media. The pay is good, and that's why people stay. But most employees recognize that it's not some great mission. It's a job.

I asked how he thought that would play into the current Twitter situation and his feeling is that Musk has had a lot of success driving people to do great things because they were working towards great causes. Launching rockets and being the underdog in a new electric car company are mission driven projects that can ignite passion in employees. He has worked with people from all over social media and his feeling is the same kind of passion will be significantly harder to generate about a social media company like Twitter.

That might be one difference of this particular undertaking that we haven't yet acknowledged in this thread.
The difference is that Elon sees Twitter as the town square for the world, but speech was not free with the former brass of Twitter. Old Twitter would shadowban people for wrongthink, or ban them outright. They would suppress responses or hide them from the original author. They were wildly offended by inconvenient or 'controversial' facts. There was the 'algorithm' which was clearly biased heavily in favor of Leftist/feminist ideology. SMR (Steven Mark Ryan) talks about this frequently and has shown the evidence.

So for Elon, Twitter is not just a place to make an idle Tweet but a locus of free speech. All of the monetization and verification stuff is just noise. Advertisers are a necessary parasite but he doesn't want them overly influencing or dictating speech. And, Tesla doesn't advertise so it's not like Elon is big on advertising in the first place, but that's probably to save money and because Tesla doesn't need advertising. Their products are their advertisements, and every stoplight is a potential commercial.

Elon freed the bird. That was and is the primary mission and no matter what noises we make, that's an important goal. That's why he spent the 44B. That said, I think Twitter is going to be way better under Elon and if anyone can aggressively iterate to monetize it while improving the signal/noise ratio, he (and his team) can.
 
The difference is that Elon sees Twitter as the town square for the world, but speech was not free with the former brass of Twitter. Old Twitter would shadowban people for wrongthink, or ban them outright. They would suppress responses or hide them from the original author. They were wildly offended by inconvenient or 'controversial' facts. There was the 'algorithm' which was clearly biased heavily in favor of Leftist/feminist ideology. SMR (Steven Mark Ryan) talks about this frequently and has shown the evidence.

So for Elon, Twitter is not just a place to make an idle Tweet but a locus of free speech. All of the monetization and verification stuff is just noise. Advertisers are a necessary parasite but he doesn't want them overly influencing or dictating speech. And, Tesla doesn't advertise so it's not like Elon is big on advertising in the first place, but that's probably to save money and because Tesla doesn't need advertising. Their products are their advertisements, and every stoplight is a potential commercial.

Elon freed the bird. That was and is the primary mission and no matter what noises we make, that's an important goal. That's why he spent the 44B. That said, I think Twitter is going to be way better under Elon and if anyone can aggressively iterate to monetize it while improving the signal/noise ratio, he (and his team) can.
Your unwavering confidence is noted. I admit I am not as confident but certainly willing to watch without a calcified position.
 
He kinda does since he fired everyone else :)

Again mixing physics and social sciences as if they're remotely similar.

I mean... he's like 7 years late on FSD and it remains unsolved-- but apart from that, the other two are vastly easier as keeps being pointed out to you.

There's basic laws of physics that tell you they can be solved and roughly how-- then you just need to execute the details. FAR from easy, but you can reason to the solutions with universal laws and principles. Elon did that.

There's no such laws about human behavior or social media. There's no first principles or universal laws on this that humans inherently all obey reliably.

it's a MUCH harder problem, and in a vastly different field from anything he has worked on. Hell Tesla had their own forums once- they were shut down because of how garbage they were, no Elon magic fixed that.

<citation needed>

Eli Lilly would certainly disagree.

So would the current financials with all the ad revenue pulling out.

Was this week needing to dump billions in TSLA to "save twitter" after already dumping 44B into it part of his "well thought out 4D chess plan"?

Like this?

Elon literally didn't fire 'everyone else'. So, that's just false. You don't need to be so hyperbolic.

Nobody said Physics and Social Sciences are remotely similar, and this has exactly nothing to do with the point of fixing Twitter. Elon understands free speech vs. LegacyTwitter™ and that was the primary driver for his purchase. Nobody is gonna spend 44B to verify us commoners. That's completely mundane and secondary noise. The main goal was to free the bird, and he did that in short order. Next he'll work out some kind of viable verification system (if anything), fix the algorithm, get his moderation council in order, and find monetization through subs and content and new advertisers—but that will take time.

FSD being late, well it takes as long as it takes. Solving real-world AI is not easy. That's why nobody else has done it nor will anyone else beat Tesla to the punch here. Not sure what your point is here, but if you know better how to solve this then you can go work for Tesla.

Human behavior has been studied for generations. While it's not a 'hard science', nobody is claiming as much. However, we do understand enough about human behavior to make reasonable predictions. Why are you so down on Elon? Why am I so up on Elon? A lot of this has to do with the big five psych traits and our respective worldviews. Some people are inherently pessimistic, or optimists, agreeable or disagreeable, skeptics or gullible (two extremes, with a spectrum between), etc.

Intelligence plays in here too, and that's obviously heritable but with environmental factors being very important as well. And, there's no tabula rasa or 'Blank Slate'. Steven Pinker debunked this idea in this book, 'The Blank Slate', among others. Some people are offended by the very idea of heritability, so we have ideologies that spring up based on ideas that don't match the data simply because some are offended by these differences. This is in part why people adopt weird ideologies, aside from religious ideas. Ideology in my view is a type of malware. One should follow the facts where they lead, because ideology can derange people in some absurd ways that may seem 'obvious' to an outside observer.

Why are some humans so resistant to facts or changing their minds in the face of new data? Humans aren't all that resilient with some rare exceptions (like Elon), and tend to defend the familiar rather than changing a fundamental precept. This is all a light overview but human psychology is fairly well understood, we have gobs of sample data, and the rabbit hole is deep. How many Flat Earthers would change their mind in one sitting? Almost none. It kinda of proves the point. Same with any weird ideology that flies in the face of settled science/fact. Is it easy to convince someone that psychics aren't real? No, but it should be. This is why partisanship and ideological ingroups can be so toxic, because it becomes an entrenched 'defense' of home and hearth where one can become impervious to new data when it conflicts with the familiar or the dearly-held tradition/notion. Just ask Galileo.

One thing I figured out early is this; intelligence alone is never enough. One must also possess courage, intellectual honesty, and humility to follow the facts where they actually lead. Elon (being on the spectrum and all) and being insanely smart is quite good at this, but he wouldn't be able to harness his genius without intellectual honesty AND a huge helping of courage. He has a perfect storm of traits which makes him a rare but effective disruptor. Other people with these traits may simply lack the resources to do much about it, or the drive and knowledge to execute.

Why does one person work in a factory until they drop dead and why does another start a billion-dollar business? It's all in those big five psych traits, and more. Environment has a whole lot to do with this....all of which is studied under the purview of psychology and the nature/nurture debate.

Nobody's claiming any Elon magic. But, he's aggressive and isn't afraid to break and disrupt a system to improve it, along with deleting unnecessary staff, parts, processes, etc. THAT is the magic.

Ad revenue is the old model. Elon is working on the new model and he doesn't want Twitter content beholden to corporate interests. Do you want Gillette vetting posts that are skeptical of feminist ideology? I don't. That's assuming the post would even survive with LegacyTwitter™ and their 'algorithm' (read: some moderator in Asia with a political axe to grind). The same censorship in various forms plagues Facebook and YouTube, and Google search results.

The 44B was to free the bird. You don't get to fire the top brass at a company if you don't own it. Elon did that, and he's dumped some staff. He's working fast and sure there will be mistakes (admitted already by Elon) but you need to sometimes start from the foundation to build a better house. That's what's happening now. Maybe you'll get some sweet Schadenfreude or Elon will redeem himself (in your eyes) but give the man some time to work. It's not like he doesn't have a great track record of being a successful disruptor (Tesla, SpaceX). Even Tesla insurance is groundbreaking, but nobody talks about that either.
 
Last edited:
Nothing.

Its just a way to oppose liberal ideologies you don't like - call it "woke".
I oppose all ideologies because I prefer facts, data, and the scientific method. "Woke" is just shorthand for one particular type of Leftist ideology and I don't use 'Leftist' as a synonym for 'liberal' (small l). There are also progressives, Marxists, Communists, Greens, Antifa, BLM, etc., and people may take on more than one label for themselves. It's just language and most people understand the context. Feminism is often combined with these labels, and there are gender feminists which are different from equity feminists and TERFs, to name a few. A lot of ideologies have overlap and exception with each other, much like religious sects do.

There are names for ideologies on the far right too; MAGA, Alt-Right, conservatives, religious conservatives, etc.. We have names for religious ideologies, and pretty much any ideological system under the sun.

I call myself a left-leaning moderate atheist, but I'm not 'woke'. See, I even have a definition for myself. That phrase tells you a lot about my mindset, but not about my opinion on every issue. Someone who doesn't even know what 'woke' means either means they are woke or very not woke, lol. This is why free speech is important. Labels are one thing and can be strong predictors of our mindset/opinion on various topics, but it doesn't tell you what we think on every topic. This is why conversation, nuance, and context is so important, along with a commitment to not misrepresenting people. To that end, it pays to know our logical fallacies lest we fail to avoid them or detect them in our adversaries.

We even have labels for the food people will or won't eat; Omnivore, vegetarian, vegan, keto, paleo, pescatarian, raw vegan, etc. Labels we choose for ourselves are just symbols, and a starting point. We can always change our minds and our positions about any topic. Sometimes people don't even get the label right if they don't understand the fundamentals, so defining terms is often a good launch point.

Quick, since I define myself as a left-leaning moderate atheist, what do you think my opinion is on private gun ownership? How do I feel about reproductive choice? I don't categorize easily, and this can offend people who don't want to exert the mental energy to actually employ some nuance.

Elon doesn't categorize easily, but the Left keeps wanting to paint him as some alt-right blowhard who likes to waste billions on a social media platform. I see him as someone who genuinely wants to enable free speech on the world's most used social media comms platform. Why would he waste 44B if he didn't have some grand plan in mind?
 
Last edited:
No he didn't. Speech is still limited on the platform and it always will be.
Well if you agreed with LegacyTwitter™ and their dubious practices then of course you may not see a difference. I see a change of management with a fundamentally more honest and healthy approach to the plurality of opinions one might express on the platform.
 
Most accounts I've read indicate that Twitter (and other social media giants) are overstaffed. Twitter was/is also hemorrhaging money. This being the case, staffing cuts are a natural and expected course of action. Very few people would dispute that. The source of contention is how the cuts were done. Employees found out they had been cut when they couldn't check their email. Ya think maybe a more professional route would have been to actually call them? Or send them a letter or and email before you cut them? Even forgetting about professionalism, just in terms of getting the work done, suddenly cutting people off leaves everything in the air with no way to make a controlled landing.

Then there's the reports of them going back to some employees and saying 'oops, we didn't really mean to fire you, can you come back?'

Finally, if half of the employees were actually bots like Elon claimed, why didn't he just fire the bots first? Seems like that would be a totally free way to cut costs.
Do you think Elon did that himself or was someone assigned that task?
 
Well if you agreed with LegacyTwitter™ and their dubious practices then of course you may not see a difference. I see a change of management with a fundamentally more honest and healthy approach to the plurality of opinions one might express on the platform.
I haven't seen a difference, other than the huge increase in parody accounts of course.
 
Genuine question. What would you train an "anti-bot" NN on? And what would they be targeting? Would the inference always have to run the bot's entire profile and post history as a sum, or could it "target" individual posts? I'm asking because it seems that bot posts could simply be tailored to be less verbose, hence with less text for "flags" that the anti-bots could identify. Less rope to hang themselves with. Similarly, fewer posts per "bot" could make it much trickier to analyze larger posting patterns.
I agree that the bots will innovate to resist detection, but this remains the task to stay ahead.

The training is the same for any NN, you feed it labeled data that demonstrates ground truth for bot or human. Then it can predict which posts/accounts are which. The trick is to label it with enough detail so that you can stay ahead of the bot innovators.

And the beauty of NNs is that I don't know what is the best way to label the data, you have to let the model tell you as you see how precision and recall fare.

When it gets good is when you can just crowd source, from trusted sources, what is a bot or human and then, it becomes an auto-labeller like Tesla has for FSD. Bot models should be small by comparison, training on like 100 GPUs for a day. Then you could potentially train everyday for about $10k per iteration. Just guessing...
 
I haven't seen a difference, other than the huge increase in parody accounts of course.
But that's your own anecdotal data. A fundamental change in leadership is a HUGE difference, because the brains and ideologies of the old brass vs. the new brass (Elon) are very different. It's night and day, really. A lot of people are simply trolling Elon because they're upset, which was totally predictable.
 
Again false

Again, not.

As proven to you repeatedly.

Jesus has $8 and an iOS device and was "verified"

Or, wait... are you under the impression that's the actual ~2000 year old Jesus?



. As I already said, abuse of the system required a throwaway Apple account, credit card, and phone number. Those took some effort for Jesus Christ to obtain.

So then, just like I said he was not actually verified

How bizare you tell me I'm wrong then 2 sentences later confirm I'm right.


Because if you put a tiny amount of effort in, you can get the check while having a 100% fake identify. And tons of people did.

Which is the opposite of them verifying your identity...and why they had to turn it off.


Heck it was explained, repeatedly, in this very thread how trivially easy it is to do. I can get a bunch of throwaway CC numbers in like 60 seconds from multiple banks happy to give them to me so I don't need to use "real" CC numbers for example.
 
I'm just saying the only functional difference I'm seeing is the massive trolling.

Sure, the old algo is still in place. But, Elon is running things and it's early days so we know things are now going to skew away from a Woke narrative (with attendant censorship) and toward more free speech insomuch as it's legal. Twitter's standard for 'free speech' wasn't even using a legal filter but an ideological one. Why else would Elon have paid 44B for Twitter? Freeing the bird was the goal, and Twitter will be better for it once enough time has passed to update the algorithm and moderation team.

With LegacyTwitter, you couldn't appeal to the brass if you were being censored or shadowbanned (or banned) because the ideological brass was the root of the problem, assuming unreasonable suppression of speech. Some people really do earn their bans, but even this is up for interpretation and debate. The hammer of 'justice' wasn't remotely fair under the heavy Leftist bias of the old guard.

For those who've listened to the Joe Rogan podcast featuring Tim Pool (acting as complainant v. Twitter), former Twitter head of 'Trust and Safety' Vijaya Gadde, and founder Jack Dorsey—it was pretty obvious how biased Twitter was. I'm guessing Elon heard this podcast too. Seeds were planted, I imagine.

It all comes down to whether you think Elon will be fairer than the outgoing Twitter brass, and for me there's no contest.
 
Last edited:
The training is the same for any NN, you feed it labeled data that demonstrates ground truth for bot or human. Then it can predict which posts/accounts are which. The trick is to label it with enough detail so that you can stay ahead of the bot innovators.
This trivializes the problem quite a bit. I have some limited experience with machine learning and NNs, as I used some of that in the business I ran, and then considered a new venture that would be heavily dependent on machine learning of different types. I went to a number of conferences, chatted with educators, tried to solve some classification problems using NNs, etc. But I admit it's been some years at this point, since I never truly embraced the concept I was chasing and decided to walk away.

What I know from that experience is that more labels can be worse. Frequently, simpler input layers result in higher accuracy. Some classification problems are challenging to the point of requiring so much more effort at getting to a reasonable solution that it'd be cheaper just to hire a team to do the work.

I watched a company have their NNs go from effectively printing money in the prop trading space to being upside down, all in a matter of weeks. The hard part about NNs is that they're black boxes, so if someone is able to game the input (think bot programmers here), they can rapidly crash an already successful network. And in my space, this was a huge data set of live streaming market data, much of which is constrained by rules of the exchanges and SEC. In Twitterspace, free speech will mean that breaking a working classification model won't take an awful lot of effort.

I'm not saying it can't be done, but it's important we don't overstate the power of NNs. They're magical, but they're not magic.
 
The difference is that Elon sees Twitter as the town square for the world, but speech was not free with the former brass of Twitter. Old Twitter would shadowban people for wrongthink, or ban them outright. They would suppress responses or hide them from the original author. They were wildly offended by inconvenient or 'controversial' facts. There was the 'algorithm' which was clearly biased heavily in favor of Leftist/feminist ideology. SMR (Steven Mark Ryan) talks about this frequently and has shown the evidence.

So for Elon, Twitter is not just a place to make an idle Tweet but a locus of free speech. All of the monetization and verification stuff is just noise. Advertisers are a necessary parasite but he doesn't want them overly influencing or dictating speech. And, Tesla doesn't advertise so it's not like Elon is big on advertising in the first place, but that's probably to save money and because Tesla doesn't need advertising. Their products are their advertisements, and every stoplight is a potential commercial.

Elon freed the bird. That was and is the primary mission and no matter what noises we make, that's an important goal. That's why he spent the 44B. That said, I think Twitter is going to be way better under Elon and if anyone can aggressively iterate to monetize it while improving the signal/noise ratio, he (and his team) can.
He didn’t “free the bird.” He hasn’t changed enough yet to claim that.
More to the point, that wasn’t the mission.
That was just the rationalization for the hasty impulse purchase. And wastin all this time on something so esoteric.
 
Status
Not open for further replies.