Twitter CEO: “Let’s talk about spam, with the benefit of data, facts, context” | Hacker News Reader

Twitter CEO: “Let’s talk about spam, with the benefit of data, facts, context”

twitter.com

61 pointsminimaxir67 comments

Great to see King Musk, the future arbiter of our public discourse, responding with a deeply thoughtful poop emoji.

The constant shitposting is exhausting and I wish he would just go away for a month (or year...).

Edit -- this is a mean tweet, I know, but it's the truth. You can't just respond to a thoughtful thread like that with a poop emoji. His plan for Twitter is probably going to be just as narcissistic as his tweets.

He did reply more: https://twitter.com/elonmusk/status/1526250477456965634

At the end of the day if you buy Twitter ads and Twitter is getting paid depending on traffic, they have no incentive to remove spam accounts.

> This is fundamental to the financial health of Twitter.

Staggering how transparently he is mounting the court defence for him to wimp out of this deal, all while maintaining that he did everything he could to save the world and twitter.

> The constant shitposting is exhausting and I wish he would just go away for a month (or year...).

Outside the US, where there's sanity around the concept of money , there is a deep desire for the wealthy to be invisible & charitable.

Only in the US you get this sort of phenomenons where rich people mount this cult of personality around them and peasants trying to impress them in order to get some crumbles.

If you are a billionaire in Switzerland, Germany or France and try to do what Trump did or Musk is trying to do...it would actually be other rich people who'd ostracize you and distance themselves.

Italy too! After Berlusconi they developed antibodies

There are few, if any, billionaires in Europe with the kind of assets that US (and Chinese and Indian) billionaires have.

There might be 3 in the top 30 here:

https://www.forbes.com/billionaires/

If you ignore Russian oligarchy

I think he knows they knew there were way more than 5% bots on Twitter and wasn't happy about getting bullshat by his new employee.

I think he knows there are less than 5% bots on Twitter and wasn't happy about paying 44bn for a 20bn company.

I think you might work for Twitter or have significant Twitter holdings! LOL

Yes, I have a significant holding in twitter - short.

>The constant shitposting is exhausting and I wish he would just go away for a month (or year...).

Have you considered not following people whose posts you don't like?

Title slightly editorialized to include that it's the Twitter CEO as it's contextually relevant and would make the title confusing if not present. (particularly as this is a direct rebuttal to Elon Musk's claims about bots on the platform)

Elon Musk has responded: https://twitter.com/elonmusk/status/1526246899606601730

> [poop emoji]

Almost no data here, the only new bit of data to me is that he is defending a system that classifies "FirstnameBunchOfNumbers" as human and not spam. Considering I can buy followers, and twitter removes tons of accounts as spam and he admits spammers are trying to constantly defeat their detection algorithm, some % of those are probably spam I'd guess.

https://www.denverpost.com/2021/10/27/buy-twitter-followers/

See timestamp 23:55 in this tech talk:

https://archive.org/details/hopeconf2020/20200726_2000_Peopl...

FirstnameNumbers is the default Twitter username generation policy. It doesn't imply it's a spam account.

> he is defending a system that classifies "FirstnameBunchOfNumbers" as human and not spam

That's a signal that a user is new or inexperienced, not that they're a bot. Twitter uses that pattern when automatically generating usernames for new users.

I can believe under 5% of all active accounts on Twitter are spam bots.

However, I would bet money that they account for a significantly higher percentage of user interactions, particularly ad interactions, and most notably link clicks.

If you run an ad on Twitter it will generate an unnatural amount of web traffic without much in terms of actual interactions on your website. No other digital ad I have run behaves in quite this way, which leads me to believe it is almost all bot activity and accidental link clicks.

Eh, more likely, pure botting is less than 5%. What's a greater amount is mechanical Turks operating hundreds of accounts using enhanced identity management and vpns to astroturf comments. It's been documented multiple times in connection with Russia and their Internet research agency.

I remember further back in 2008 the CIA themselves had a RFP out for "identity management software", so it's been decades since governments recognized the value of astroturf and making social media accounts at volume to persuade public opinion.

I'd love for this argument be put before the court threading the need between bots and mechanical Turks

Related ongoing thread:

Nearly 20% of active Twitter accounts likely to be fake or spam - https://news.ycombinator.com/item?id=31397137 - May 2022 (305 comments)

It's a tool that enables someone who is too dumb to understand how dumb he is, to find the largest group of people on the planet, who are dumber than him. And exploit said group.

In the past you needed an army or huge amounts of corporate capital to find and exploit large groups of people. Twitter has shown you just need to pander.

Such a tool can't be "improved" or "fixed". It needs to be shutdown and banned.

You just described all of social media

The irony of the CEO of twitter presenting information supposed to add nuance and context to a debate… using a twitter thread where every paragraph is an open invitation to misinterpretation.

The twitter format is not conducive of a nuanced data-based discussion and never will be, by design.

Nor is there anything in that thread by CEO that shows us he has no incentive to lie and hide behind "personal data not accessible to outsiders". But trust us, they're human! You just can't check!

It's mind-blowing how anyone can be naive enough to believe that, without some sort of personal incentive (e.g. Twitter employee or some dir-level whose bonus depends on the buyout)

I'd say his word is more convincing than the exact opposite position as similarly-baselessly-claimed claimed by Elon, who similarly has a conflict of interest incentivizing him to be dishonest, as well as a vindictive personality driving him towards the same

I dunno, is it a CEO's duty to conceal certain company factors that could have a negative influence on the shareholders?

In other words, is it OK to straight up lie to fulfill the fiduciary duty?

If the answer is yes, then the CEO is less believable.

AFAIK, the duty of the CEO to lie for the shareholders is far outweighed by the duty of the CEO to not lie to the shareholders via SEC filing (or earnings calls, etc)

How is this different than me quoting only bits of a blog post? You can say that you can just reply with the whole quote , but I can just easily quote tweet reply the tweet that gives context. I feel fact the quote tweet is even guaranteed to be unedited and linked to the rest of the conversation.

The difference is that a blog post is presented as a whole, in long form. The author can construct context to have a debate informed by things harder to put into an individual tweet. Long form communication is where nuance lives. Twitter actively seeks to reduce and diminish that because any single tweet has to summarise its ideas in a single character-limited block. I personally think it’s a net negative for the world

Ironically we have had this discussion in tweet sized chunks. Guess I’m a hypocrite, eh.

So weird to see people want to have identitiy requirements to post online. I'd prefer to be anonymous, even this account is pushing being to attached to my real life identify.

I think the middle ground, pseudonymity, would be way better than the true form of either end. I think there was some research done that shows how it's also a bit less toxic than the other two.

At some point most accounts reveal enough bits to identify someone. To actually keep anonymity is difficult and also comes with the abuse being discussed.

Unfortunately I'm not aware of a zero-knowledge real person mapping system. But that would work to a reasonable extent. Any person (no info who) could pick one (active) username on a website. It could curb abuse and doxxing both.

People want authentic social media and realize anonymity does not provide that in the face of any monetary value to the content creation and distribution.

The world you're pining for does not exist.

don't we have facebook for that. i don't think its easy to have anonymous accounts there.

reddits pretty anonymous, seems to mostly work, and imo better then twitter

I disagree, but I would be interested in seeing an experiment where users can choose if they want to only see information from accounts that have had their real world identity verified. Hell, if it's important to enough people make that the subscription based slice of the site.

We need both. We need places where people can speak as themselves online and places people can speak and communicate anonymously. We need both because online is taking the place of face to face conversations more and more than it ever has. There needs to be a way to know that the equivalent of a random person you meet in the bar is actually a human.

> Unfortunately, we don’t believe that this specific estimation can be performed externally, given the critical need to use both public and private information (which we can’t share). Externally, it’s not even possible to know which accounts are counted as mDAUs on any given day.

Thank you Parag, for admitting that Twitter needs someone else to clean up the bots and spam accounts plaguing the platform. Like it or not, it will soon require KYC to get that silly verification checkmark at this point and bots that still exist there would need to be on a leash, tied to the KYC'd account to know who is running it and where it is from.

Don't believe me? YouTube, Facebook and Instagram are already ahead with this.

I won't be surprised to see this implemented in Twitter. If they don't? Good. Either way, I don't care if they do it or not since I never had an account in the first place and by just lurking on Twitter, it's apparent that is an infestation of brainless parrots not worthy of a sensible conversation anyway.

Someone else can cleanup that mess and if Elon wants to do it, fine.

> Thank you Parag, for admitting that Twitter needs someone else to clean up the bots and spam accounts plaguing the platform.

This made me chuckle a bit. It’s like a manger thanking to their employees for admitting that a program needs a different team to be bug free. As I see it there is only one absolute: there are no absolutes.

I honestly don’t believe Twitter needs clean up. Rather user needs to use the tools at their disposal. Twitter has the best user-moderation tools and if it’s insufficient I see no reason why a third party client is not a sufficient solution.

I must use twitter differently to everyone else because I don't see the 90% of all twitter is bots and spam.

My timeline is people I follow; sometimes they retweet absolutely demented retarded shit that makes no sense but they nevertheless believe in.

"The algorithm" just brings the more engaged stuff from people I follow to the top, however, I don't use it because while I don't care about someone I follows hot take on thing happening right now, I care even less in 2 days time. As far as I can tell it hasn't turned me into a socialist / white supremacist.

Occasionally I will take a look at "trending" - Yes 90% of what is posted is utter garbage, but its not coming from bots and spam, its coming from people.

only 5 percent is real users IMHO

It's actually completely irrelevant how many there are.

There's no real difference between a spam/bot account and some random troll with too much time on their hands who just retweets posts from a flame war.

It’s relevant to the advertisers who are paying per impression. Bots are not going to buy their products.

It's relevant because if there are more they lied to Musk and he can pay less money, possibly much MUCH less.

and if his unsubstantiated pretext to back out of the deal remains unsubstantiated, he cannot

That's why Twitter will be made to provide better information to Musk as this goes along. There are many ways of determining if some account is a bot for sure, Twitter likely tracks all kinds of standard things like mouse position in the window, scrolling behavior, etc. which could easily discern a bot from a human. He's clearly not falling for the 5% number, and he shouldn't. 90% sounds low.

musk can request however much info he wants, but no amount of bad faith requests serving as a pretext for what is clearly a deep, economic-based regret will give him control over the deal now that he's signed a contract

if you ask me, "I can't tell if there are bots on twitter" seems like a really good reason to not tender an offer until you can, but then again, I'm not as impulsive as Elon

in any case, didn't he claim he was doing this for free speech and not money? Why should bot counts and advertisee eyeballs matter in that case? Not that I am one of the people who believes his oft-outlandish assertions and poop tweets anyways.

Heh, who cares if Twitter was knowingly misleading investors for years!

Tell that to paying advertisers

What are they paying for though? Influence on twitter? You don't need real users for that, just promotion with the algorithm

Musk has a point - the reality is that each and every Twitter user wades through obviously fake accounts every time they use the service.

this is a problem that hasn't been solved. all explanations around why that is are meaningless compared to real action / perceptible change.

Musk is the largest beneficiary of all the fake accounts pumping his follower count.

He basically lives on twitter so he knows everything perfectly. He should pay the 54$/share as per the contract he signed.

The stock market crashed and it keeps going down so he wants to negotiate a different price, that's all there is to it. Buyer's remorse due to deterioration of macroeconomic enviornment.

And the credulity of people believing a billionaire would come to save them because he's "not caring about economics at all"...

> Musk is the largest beneficiary of all the fake accounts pumping his follower count.

Fake accounts don't pump TSLA stock. They have no purchasing power. And he has a large enough following that it really doesn't matter what the number next to his name is, whether it's 30 million or 90 million.

But I agree with you that it sounds like he's got buyer's remorse. Seems like free speech (or whatever the public reason he stated for buying Twitter) isn't worth that much to him.

They might have purchasing power, how do you know?

How do you know they are obviously fake? That's the point the CEO is making. From the outside you can't always know. A lot of accounts that get fingered as 'obviously' fake aren't, in reality.

Edit: I guess I'll add some background here. I worked on anti-spam for a few years at Google and basically agree with everything Parag is saying here (although we'd disagree about many other things!). What he's laying out is spam fighting on social networks 101. Stuff that outsiders often latch onto, like user name patterns, are ~useless. You need a whole lot of signals many of which simply cannot be replicated by outsiders to detect spam. For example, one set of signals that Google was largely ignoring when I first joined the team was protocol deviations. We built an infrastructure to force and detect them, which was effective and is still in use.

For some years now I've been writing about the plague of academic "research" into Twitter bots that use completely invalid methodologies to try and detect spam accounts. There are over 11,000 published papers on this topic, which is absurd because very close to none of them are sound.

https://blog.plan99.net/fake-science-part-ii-bots-that-are-n...

https://blog.plan99.net/did-russian-bots-impact-brexit-ad66f...

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3814191

Tech talk for those who prefer video:

https://archive.org/details/hopeconf2020/20200726_2000_Peopl...

I'm pleased to see that the idiotic 20% claim is getting shredded on another HN thread, perhaps now people are waking up to the weakness of these sorts of claims more authors will stand up and publicly debunk them. As far as I can tell only myself, Florian Gallwitz and Michael Kreil have been pointing out the problems with third party Twitter spambot investigations in recent years.

Although I'm definitely a Muskian on free speech topics, I still feel a lot of sympathy for the spam fighters working at Twitter. The extent to which their work is second guessed is astounding, the fact that the second guessers are often peddling pseudoscience with institutional credentials just makes it worse.

Yeah. For a community that constantly berates major tech companies for FPs in policy enforcement, it is baffling to see people say that various users are just obviously spam.

Right? If there’s one thing we should know in this industry is that unless we have all the facts , we know nothing , and even when we do, we still know nothing.

False positives are expected. It's another cost to handle them gracefully so that the people you affect are affected in the least bad way, and can get back up and running quickly.

I would expect this post to not go over well on another thread about an incorrect policy action. It is definitely not the norm that this community expects false positives.

In part this seems like a denominator issue. Many real Twitter users lurk, or post very seldom. Bots probably don’t lurk much - what would be the point?

If you took the likelihood of a tweet being spam, it’s going to be much higher than a user being spam, when “users” includes the lurkers you rarely see.

For the user experience, what you see is what matters and lurkers don’t count. But counting lurkers might matter more for ad impressions and revenue.

I wonder sometimes if Musk knows this stuff and enjoys playing dumb to troll people, or if he really is out of touch. Maybe focusing on the user experience is the right move regardless of how you get there?

The conversation we should be having is where are we now, and what is good enough?

Having worked in large scale anti-abuse detection for most of my career (~18 years), the points mentioned line up in the Twitter thread align with my experience. Scaling in this area is hard. 99% efficacy sounds great, until you say 99% out of millions/billions. The amount of FNs ('bad' or unwanted things) is still substantial enough for users to notice. Taking a 229m active user count [1], 99% fake account detection efficacy sets you at 2.2m fake accounts. Looking into tweets/day you've similarly large numbers if you want to look at content detection.

Twitter can most likely do better given the right resources, people, and leadership support (Facebook has similar problems aligning all 3 of those). Once they have those, the open question is how much better they can get. Each incremental increase in efficacy gets more expensive.

To top it off, as detection gets better, you think those abusing Twitter will sit still? Of course not, they'll change tactics (content, usage of hacked accounts, etc.).

[1] Twitter 10-Q 2022-Q1 - https://www.sec.gov/ix?doc=/Archives/edgar/data/0001418091/0...

I am 100% sure that if the stock market rallies back to March levels all this nonsense would disappear and Musk simply buys Twitter for the agreed upon price of 54$

Musk is the perfect CEO for our era, this guy doesn't brush his teeth unless the stock market gives him a solid round of applause accompanied by a 10$ rally in TSLA stock price.

Behind all the boldness and bravado there is a constant need to re-assured that his decision is the best decision. Warren Buffett on the other hand salivates every time the stock market goes down, rooting for it to go down at times so he can buy shares in the companies he believes in.

But why is spacex private then? And Didn’t Tesla and spacex almost both go bankrupt in 2008?

If the email that came out right after Thanksgiving last year was accurate then there is a decent chance that Raptor 2 engine development might bankrupt SpaceX which makes him spending so much on Twitter when his dream is SpaceX / Mars / space travel a bit strange.

Do you have a source, or at least a good set of keywords to google?

Here you go https://www.cnbc.com/2021/11/30/elon-musk-to-spacex-starship...

It's really bad management to send such an email the day after Thanksgiving to.

Ah, thanks.

I wonder what degree he believes that SpaceX is in trouble vs. just not thinking people should take holidays off.

Or maybe Musk took the billions set aside for funding to short the stock to get the rest of the funding. We don't know what's going on behind closed doors. I would assume he knows more about investing than you are giving him credit for.

I would, on the other hand, NOT assume that, since I have not seen enough data to support such a bold assumption

> The hard challenge is that many accounts which look fake superficially – are actually real people. And some of the spam accounts which are actually the most dangerous – and cause the most harm to our users – can look totally legitimate on the surface.

Regular users don’t care if a low-effort post is from a bot or a human. They want their feed populated with interesting content.

Also, freedom of speech is good, but for most what actually happens is they post whatever “hot take” they have and nobody or like 2 people notice.

What’s really important is the ranking algorithm your site uses, and whether it upweights spam posts, “pop culture” human posts, or whatever posts are most relevant / interesting to each user / community.

I feel like Elon complaining about spam is really him just making up an excuse for suddenly backtracking his offer.

Reply on news.ycombinator.com