I'd like to talk about social host liability. If, as you read that term, you parsed it as 'social' as in social media, 'host' as in webhost, and 'liability' as in if a social media webhost does something negligent that causes harm to come to third parties that they should be held legally liable... then you don't know what social host liability is. You do, however, have an intuitive grasp of the point I'd like to make in this article.

Social host liability means that if you are serving alcohol at a party, and you serve one of your guests to a point of obvious intoxication, and that guest leaves and gets in a car accident or otherwise hurts someone where the proximate cause is determined to be intoxication, then you're legally liable. Dram shop laws are similar but apply to commercial establishments (e.g. bars) rather than private citizens hosting parties.

I mention this because we have a clear precedent in another context where we stop people from irresponsible overuse of a cognition-impairing addictive substance. In some cases we go as far to hold businesses and private citizens legally liable if they allow a person to continue to irresponsibly use that substance. 

While I've seen various criticisms of social host liability and dram shop laws, it's never along the lines of "we're violating people's First Amendment rights by prohibiting them freedom of expression through drunk driving." 

What is Twitter?

Twitter in 2021 is one of several businesses that have successfully developed a scalable way to surreptitiously give hundreds of millions of people a behavioral addiction and monetize it… oh and all the while making them believe they’re virtuously practicing free speech.

A more accessible answer to this question comes from an Irish podcaster, Blindboy (quote starts around 1:07:00).

Twitter is a video game that people don't know they're playing. Because the thing with Twitter is you're always engaged in an act of performance... Twitter encourages people to create a characterized version of themselves and to preform this character, like a role playing game. Twitter is a giant massively multiplayer online role playing gametext basedwhere it rewards hostility... You get rewarded on Twitter for having really good complaints. If you can think of a really good complaint on Twitter, you'll get lots of points in the form of retweets and likes. But the thing is, if everyone on Twitter is complaining, because this is what's getting you likes and points, then Twitter becomes an excessively negative place, which it is. 

I would only add, unlike other MMORPGs, how good (or bad) you are at playing Twitter can have substantial effects on your real world reputation and mental health.

A slightly more academic perspective on the topic comes form Shoshanna Zuboff's excellent book, The Age of Surveillance Capitalism. In the same way that capitalism optimized Doritos to be maximally addictive, it's done the same for casinos. Much of what was learned about casinos was repurposed for social media. Shoshanna is discussing Facebook here, but in terms of having a financial incentive to increase advertising revenue through engagement, Facebook and Twitter are isomorphic (hyperlinks are my own and added for clarity).

The hand-and-glove relationship of technology addiction was not invented at Facebook, but rather it was pioneered, tested, and perfected with outstanding success in the gaming industry, another setting where addiction is formally recognized as a boundless source of profit. Skinner had anticipated the relevance of his methods to the casino environment, which executives and engineers have transformed into as vivid an illustration as one can muster of the startling power of behavioral engineering and its ability to exploit individual inclinations and transform them into closed loops of obsession and compulsion. 

No one has mapped the casino terrain more insightfully than MIT social anthropologist Natasha Dow Schüll in her fascinating examination of machine gambling in Las Vegas, Addiction by Design. Most interesting for us is her account of the symbiotic design principles of a new generation of slot machines calculated to manipulate the psychological orientation of players so that first they never have to look away, and eventually they become incapable of doing so. Schüll learned that addictive players seek neither entertainment nor the mythical jackpot of cash. Instead, they chase what Harvard Medical School addiction researcher Howard Shaffer calls “the capacity of the drug or gamble to shift subjective experience,” pursuing an experiential state that Schüll calls the “machine zone,” a state of self-forgetting in which one is carried along by an irresistible momentum that feels like one is “played by the machine.” The machine zone achieves a sense of complete immersion that recalls Klein’s description of Facebook’s design principles—engrossing, immersive, immediate—and is associated with a loss of self-awareness, automatic behavior, and a total rhythmic absorption carried along on a wave of compulsion. Eventually, every aspect of casino machine design was geared to echo, enhance, and intensify the hunger for that subjective shift, but always in ways that elude the player’s awareness.

Schüll describes the multi-decade learning curve as gaming executives gradually came to appreciate that a new generation of computer-based slot machines could trigger and amplify the compulsion to chase the zone, as well as extend the time that each player spends in the zone. These innovations drive up revenues with the sheer volume of extended play as each machine is transformed into a “personalized reward device.” The idea, as the casinos came to understand it, is to avoid anything that distracts, diverts, or interrupts the player’s fusion with the machine; consoles “mold to the player’s natural posture,” eliminating the distance between the player’s body and frictionless touch screens: “Every feature of a slot machine—its mathematical structure, visual graphics, sound dynamics, seating and screen ergonomics—is calibrated to increase a gambler’s ‘time on device’ and to encourage ‘play to extinction.’” The aim is a kind of crazed machine sex, an intimate closed-loop architecture of obsession, loss of self, and auto-gratification. The key, one casino executive says in words that are all too familiar, “is figuring out how to leverage technology to act on customers’ preferences [while making] it as invisible—or what I call auto-magic—as possible.”

As far as I understand it, very few people even inside of Twitter and Facebook understand how their algorithms work in detail (if you have details, please share them in the comments). But we know the meta pattern of market forces evolving a business model, as it happened with Doritos and casinos, still applies. Those same market forces put constant selection pressure on Twitter and Facebook. We can expect their products have followed and will continue to follow a similar trajectory until their business models change.

Common carriers are not behavioral addictions

Infamously in 2012, Megaupload was shutdown because it was often used by people pirating media. Commenting on this event Steve Wozniak said:

If someone commits a crime shipping drugs on the sea, you don't drain the sea and say the sea is the problem. If they are mailing drugs through the post office you don't shut the post office down you try to get the people who are doing the wrong steps.

The implication here is that Megaupload was a common carrier, a neutral conduit, like the postal service.  The wrong-doing where it occurred was on the part of people using Megaupload illegally, like someone sending drugs through the mail. I agree here with Steve and this was an excellent point about Megaupload. Megaupload was much more like the postal service. 

Is Twitter like the postal service? We can answer that with simple thought experiments.

How many people, for example, regularly stay up until 3am compulsively stuffing envelopes with angry letters to strangers they disagree with about, say, K-pop bands? And, even if some eccentric types like that did exist, is the postal service in the business of amplifying their eccentric behavior and encouraging others to respond in kind? 

User content on Twitter isn't in an envelope or shipping crate. To the contrary, user content on Twitter is something like a prodrug, a substance that by itself is mostly inert, but is algorithmically metabolized by the platform to become “biologically active” and addictive. Jaron Lanier suggests this is done, to some extent, by separating users in to bins and finding what kind content engages those users then finding ways to get it in front of them

Consider how few people care about the press releases Trump has been writing since he was booted off the platform. Even though Trump's post-ban press releases are written like Tweets (down to the @ mentions), they’re relatively inert sans Twitter's "algorithmic metabolism." 

Trump’s derangement syndrome

Jaron Lanier often remarks on how he met Trump at several points in his life and watched Trump become progressively more deranged as Trump‘s Twitter use increased. As Jaron put it in Ten Arguments for Deleting Your Social Media Accounts Right Now:

As a Twitter addict, Trump has changed. He displays the snowflake pattern and sometimes loses control. He is not acting like the most powerful person in the world, because his addiction is more powerful. Whatever else he might be, whatever kind of victimizer, he is also a victim.

A better question than “why are we banning Trump from Twitter?” is instead “why are we encouraging high-profile politicians to engage with potential behavioral addictions?”

That's really a very serious question. Can you imagine a situation where large groups of people would demand we allow someone who still wields tremendous real world influence and power to become more regularly drug-addled? If substances are too much of a stretch, then substitute a behavior like gambling and the analogy still holds. 

The appropriate feeling towards Trump's ban from Twitter strikes me as mudita—a sympathetic joy that he's been given the possibility to dry out and sober up, as it were. I know that sounds like some hippy shit, but I'm not ashamed to say I believe it. He has a real chance for reflection and growth right now, and I hope he uses it. 

A bearded Trump deprived of salon appointments due to complications from eating a carnivore diet ferries the very online to the polarization realm on a boat named Larry... err I mean a 19th-century interpretation of Charon's crossing by Alexander Litovchenko

In all seriousness, "Trump in recovery" strikes me as something that would be powerfully good for the world. Where he once was Charon-like character ferrying people in to ever more polarizing hell realms of the Internet, he could instead become a something more like a Bodhisattva Boatman who "aspires to become a Buddha simultaneously with other sentient beings, sharing their difficulties and encouraging them along the way." ... okay, I'll admit I'm being slightly facetious here, but I still think it's a good idea. 

It's time to update

In January of 2021 a LW user, ragintumbleweed, lamented about how many rationalist-adjacent and rationalist people were against removing Trump from Twitter. The post currently has -3 karma.  

At the risk of writing a post that receives even more negative karma, I want to say ragintumbleweed and rationalists who want to see Trump back on Twitter have all missed the elephant in the room—they all seem to presuppose that Twitter in 2021 is a game that's safe enough for individuals and humanity at large. 

The seismic shift that’s occurred in the last 10 years is the ability of social media platforms to freebase user generated content and create serious behavioral addictions with very salient real world consequences. We‘re making a category error if we continue to discuss Twitter like it’s the same platform it was 10 years ago. 

Adding behavioral addiction to free speech is like adding raisins to cinnamon rolls and cookies


New Comment
39 comments, sorted by Click to highlight new comments since: Today at 9:37 PM

Wireheading is expanding rapidly.

At first it was drugs, with packaging and delivery carefully designed to maximize addictiveness.

Then gambling,  social media. 

More recently...

Stock brokers are increasingly leveraging the well-tested wireheading techniques used by casinos to make their customers into gambling addicts.

Stock brokers are increasingly leveraging the well-tested wireheading techniques used by casinos to make their customers into gambling addicts.

Confirmed. They've reached my co-workers. I can't get through a day without hearing these newly-minted day traders arguing about which crypto is going to do what. They specifically talk about how this is so much easier than going to the casino and still scratches the gambling itch. (sigh)

I completely agree, and wasn't aware that stock brokers were doing this. 

This is a minor point--I wonder if it's worth coming up with another term for what people typically mean now when they say 'wireheading?' Recently when I see people using  the term now, they almost never mean literally mean wireheading. 

I can't find the Tweet now, but I'm pretty sure I saw Roko describe masterbation as wireheading recently, when (absent some crazy tech you're applying to it) it's maybe one of the least cyberpunk things you can do in terms of our long human history of it.

People used to say things like "chasing the dragon's tail" or "pushing the envelope." I suppose they're both a bit more of a mouthful... 

It's a bit of a tangent, but just saying... 

EDIT: I suppose the article on it does say: "If we take wireheading as a more general form of producing counterfeit utility, there are many examples of ways of directly stimulating of the reward and pleasure centers of the brain, without actually engaging in valuable experiences. Cocaine, heroin, cigarettes and gambling are all examples of current methods of directly achieving pleasure or reward, but can be seen by many as lacking much of what we value and are potentially extremely detrimental."

I was taking it more as a "high-tech cyberpunky form of pleasure-seeking" but I suppose most people don't share the connotation.  

The problem with this argument is that it's implicitly assuming that all curves are lines, i.e. very roughly you argue that politicians on Twitter is bad, politicians not on Twitter is good, and therefore removing one politician from Twitter is a step in the bad->good direction so it's good. But not all curves are lines - e.g. if Twitter banned half of all politicians but chose only the conservative half, I claim that's worse than each of the endpoints of banning nobody and banning all politicians. (Well, except for the fact that with such an outrageous show of bias they might lose some of the hold they have over the world, which would be good.) So you have to argue more specifically that removing Trump and only Trump is still good. Might be doable, but I don't think you did it - in fact you specifically replace that narrow question with the general one ("A better question than “why are we banning Trump from Twitter?” is instead “why are we encouraging high-profile politicians to engage with potential behavioral addictions?”")

So you have to argue more specifically that removing Trump and only Trump is still good.

That's likely to be good for Trump. As you observed, there's much more work to do.

I agree with your point in general (not all curves are lines) so if the best possible state for Twitter is to have no politicians (and I would say ultimately no users, at least in its current form), then removing Trump may work like chemotherapy or something like that -- you feel worse before you feel better. 

A reasonable model of reality

Like a quadratic curve that drops below 0 and then rises back up as more people are removed.

I accept that as a possibility. 

My intuition would be that this was not the case here. Trump's mixed messages about the people invading the capitol building struck me as irresponsible (much worse than, say, drunk driving). Perhaps only a temporary ban was necessary there, sure. But ultimately there's very little question in my mind that Trump is healthier off Twitter. 

I'll also say the Donald Trump that left the Reform party when it was clear that party was falling prey to entryism is not the same Donald Trump in that was in office in 2020. I would love to hear anecdotes of people who knew him personally who could be somewhat objective about it (I realize Jaron is biased there, but it's still an important point) to see how various forms of reach-seeking (ratings, likes, retweets, etc) changed him.

One of the largest hazards influencers face is audience capture. I can only speak to what my gut says, but I think he painted himself in a corner where he had to cater to nearly mutually exclusive groups on the right.    

There are two possible meanings to phrases like "banning Trump is a good thing", and I think you are conflating the two in a motte-and-bailey. The motte is "the fact that Trump is banned is a good thing for Trump's health". This is likely true for the reasons you mention (Twitter is a drug etc). The bailey is "the decision process that Twitter used to ban Trump (which, as an ongoing process, has also been used to ban people in the past and will presumably be used to ban people in the future) is a good thing for the world". But to defend this statement, all claims about whether Trump himself is healthier off Twitter are irrelevant, because Twitter is obviously not banning people for their own health (they're the cause of the addiction!). Even if every individual person they choose to ban is personally better off banned, if the process is flawed (as our simplistic example, "only ban conservatives") then the process can still be bad for the world at large.

So which of the two versions are you claiming (or is it some third thing I missed)? If your claim is just "the fact that Trump is banned is a good thing for Trump", then fine, I don't disagree with you, but I'm also not sure why you'd bother to write this article - why should we all care about Trump's personal health? But if instead you're arguing that Twitter's process isn't broken, then stop hiding in that motte.

I wrote this primarily because there is a category error in terms of how people think about Twitter, and how Twitter presents itself to the public. I didn’t spend as much time on the latter case because I don’t know that many people reading this work for Twitter.

If I had the opportunity to write an essay for Jack Dorsey I would basically say “hey, you’ve created a behavioral addiction and you need to manage it like you’re managing a behavioral addiction and not pretend it’s anything but that.”

You can say what you want about alcohol and tobacco, but at least they acknowledge they’re peddling an addictive vice. They at least encourage people to be responsible, even if it’s nearly an empty gesture. Twitter made their own addiction and they won’t cop to it. They’re making decisions in a way that attempts to shift the blame on to the people being banned and it’s sort of like Purdue Pharma telling people if they’re hooked on oxycodone, it’s entirely their own fault. https://www.justice.gov/opa/pr/opioid-manufacturer-purdue-pharma-pleads-guilty-fraud-and-kickback-conspiracies

Twitters decision making is flawed largely because they’re pretending they’re not a behavioral addiction that rewards antisocial behavior from their users.

I perhaps should have been harder on them here if I have an impression otherwise.

Why isn't this an argument for banning all politically powerful people from Twitter?

Or maybe all people and we just have bots arguing for retweets and likes? :)

I think it is? That was kind of the implication that I read into it at least.

Seems to me that the question of whether Trump should be banned from Twitter is a distraction.

The real question is whether Twitter should ever use the means it uses to boost "engagement", or indeed should be allowed to use those means. If you solve the Trump problem, you still haven't solved the K-pop problem.

Also, I don't get the raisin analogy. Raisins in cinnamon rolls and cookies are nasty abominations that reduce addiction. As far as I can see, that's the only reason you'd put them in there...

That’s fair… I was trying to think of things that make something good (free speech) intolerably worse. There were both memes to that effect… so yeah.

But I’m sure there’s a better analogy there, I just couldn’t think of one.

I think the metaphor is that raisins are (reasonably) healthy but adding them to cookies won't make cookies healthy.

This title is something of a bait-and-switch. The headline talks about removing Trump from Twitter, but the actual argument presented is an argument againstTwitter in its entirety, at least as it currently exists. As far as that goes, I am in favour of keeping Trump off of Twitter if and only if it's a first step towards keeping everyone, everywhere off of twitter.

The seismic shift that’s occurred in the last 10 years is the ability of social media platforms to freebase user generated content and create serious behavioral addictions with very salient real world consequences. We‘re making a category error if we continue to discuss Twitter like it’s the same platform it was 10 years ago. 

Important point, and well-put.

The Jaron quote was also powerful; I hadn't heard that sort of thing about Trump before but it's not surprising. I personally think the highest reasonable hope would be for Trump to return to how he was in 2012--the birth certificate stuff was much less bad than the Qanon stuff and the capitol insurrection. That was less bad, but it was still bad and this might undermine a sanguine narrative of "Trump in Recovery"...but if it somehow didn't, then yeah, I'd be happy to see that narrative get some airtime.

Regardless of how the stories of Trump end up being told, I do hope that people start to see Twitter as the psychotoxic game that it is. I have expressed some optimism about this in Predictions for future dispositions toward Twitter. It's possible that tech companies will eventually try to sell cleaner digital ecosystems to conscientious end-users--I imagine a high-income, tech-savvy buyer paying extra for a well-integrated device-app ecosystem that tends to respect & enhance one's mental/emotional life rather than harming it. This could come to represent a high-status, post-Twitter lifestyle. Again this is optimistic, but perhaps worth hoping for.

I would hope it's as simple for Trump as "Banned from Twitter --> Path to Recovery becomes available", but I wouldn't give it particularly good odds (at least in the near-term). I suspect that at least some of the people he's made strong and very public connections with in recent years will tend to be large obstacles in that path, and may divert him from it entirely. And that's if he can even see the path in the first place.

May I be proven wrong!

In the bigger picture, I agree that we really need to deal with social media. The suffering-reduction that could be accomplished is substantial here. And the possibility of social media being a distraction from civilization-building enough that we end up self-destructing is pretty alarming.

I would charge a small, hormetic, amount of money for each post and a smaller amount of each action taken on that post. I believe people originally proposed this for email, but decided against it, if I understand it correctly in part because the implementation options were difficult at the time.

Micropayments have been difficult in the past, but there’s some cryptos that can do them now with very small transaction fees (e.g. $HBAR).

The some of the money could go to the poster so people could make income from social media, some to the platform so they would rely less on advertising.

This would also create small amount of friction for posting content and make users stop and think for a second as to whether it’s worth it.

You raise good points. Society would benefit through the re-aligning of incentives 

If this was done and it was done like with royalties, it represents a form of long-term income for people who create things of value and put them on the web and similarly discourages people from creating "the next clickbait hot take" that will typically have no lasting value and has no net benefit to society. The pro-social incentives are better aligned there.

Think of it like music (pre-Napster). Musicians either (1) wanted to make pop music that will get played to a large audience so they collect larger royalties or (2) if they can't go for mainstream pop they try to find or create a niche audience and dominate there. 

I'm not sure if you're an old like me. But it makes me think of this KMFDM track, Fairy, on their album Xtort. In 1997 playing that in one of your friends car discmans was this crazy experience of "Oh my god I can't believe they put that on an album!" But by comparison to most of the web now, it's so tame that it's boring. 

When incentives are aligned you're more likely to get more people trying to make an album like Abby Road, and a few people following the strategy to make albums like Xtort. The artists like KMFDM will be around but, will be in the counter-culture margins where they're still both cool and profitable. There is no mechanism to incubate a real counter-culture anymore--anything that's cool that gets created gets discovered and becomes part of the mush that is the completely permeable borderless Internet. It's like when someone has a broken jaw an has to eat through a straw, so they just blend what could be a four course meal in to a shake. 

When incentives are misaligned like they are now, art and creativity are mistaken for or replaced by vice and base pleasures. I'm not a prude, vice and base pleasures have their place, but there's good reasons why you shouldn't have chocolate cake and cocaine for breakfast.

Well, yes, basically. Here are some sugestions for exploration. I am not saying all of these are good ideas, and some of them conflict, but they're things you could look at.

  • Don't allow responses, by which I mean replying, retweeting, liking, forwarding, or whatever, until the base material has been up for something like a couple of hours. That includes responses to responses.

  • Adjust that delay so that a response that will be seen by few users can go through faster than a response that can be seen by many users... but there should always be at least a few minutes of delay for any response that is public, goes to a very large audience, or could be in any way be forwarded to become public or go to a very large audience.

  • Limit the number of responses a user can post per hour. Put heavier limits on responders who don't generate a lot of original posts. Put still heavier limits on people with large followings.

  • Combine the above so that the delay or audience limits applied to you depend partly on how many posts or responses you generate in general.

  • Downrank prolific posters.

  • Downrank clusters of posters who frequently amplify one another, especially if nobody outside the clique seems to amplify them to the same degree.

  • Aggregate reposts of the same link or substantially the same text, and treat them as a single object that is show to each user at most once.

  • When you're ranking material to display to a user, uprank material from accounts that user follows that have fewer other followers (like family and friends) over material from accounts that have more other followers (like politicians and media). On edit: heavily uprank posts from people who reciprocally follow the reader.

  • Provide downvotes, and have them actually sink material rather than upranking it as "controversial".

  • Uprank long posts and maybe posts with multiple links... especially links to sources that do not usually appear together in other posts on the platform.

  • Uprank material that's grammatically correct.

  • Eliminate avatars and emoji

  • Consider eliminating user identity and "following" in favor of an anonymous rumor mill.

You can provide overrides for those, but they should be things that have to be selected every time you visit the site. Better yet, provide access to all the content using an API, and allow users to use clients other than the official ones... including clients that aggregate different services.

I am suggesting applying things like that globally, to all users, not just users who have done something to get noticed.

Does 4chan have reply delays?

By the way, I think I'd like to amend that "response delay" thing to be "after the responding user first sees the material", rather than "after the material goes up".

I had been thinking of social media as "digital nicotine", but the metaphor of MMORPG/casino fills in a large gap in that picture.

Nicotine is a natural substance that presses certain buttons in our brains that make us crave more of it. Cigarette companies' whole deal is to figure out the optimal delivery platform (form, size, flavor, etc.) and make money off our brains' reaction.

But MMORPGs and casinos, they're designed by us to be like little theme parks for our brains. Instead of pressing buttons like "pleasure" or "chill", they press buttons like "family", "status", etc. Coupled with instant feedback loops that can optimize the input for each unique user in real-time, it seems to me that social media is on a whole different level than mere nicotine.

Your fundamental theory is that Facebook and Twitter, while not actually changing brain chemistry, are so addictive they are like a dangerous drug.

Is there some way to test for this? What allows you to know this? Is there a way to "test" an online service for addictiveness? Is Twitter more or less addictive than world of warcraft? How much does possible mental degradation in Trump's mind affect his propensity for addiction?

Your fundamental theory is that Facebook and Twitter, while not actually changing brain chemistry, are so addictive they are like a dangerous drug.

Why do you believe in dualism? (That would be required for reading something on a computer not changing brain chemistry)

Ok, externally changing it.

 The biological answers I'm aware of are centered around expression of ΔFosB and related activity in the nucleus accumbens, and are seen in both substance and behavioral addictions. 

From Wikipedia: "[b]ased upon the accumulated evidence, a medical review from late 2014 argued that accumbal ΔFosB expression can be used as an addiction biomarker and that the degree of accumbal ΔFosB induction by a drug is a metric for how addictive it is relative to others." 

So, I do think they actually "change brain chemistry." (Although as mentioned above, I think just about everything "changes brain chemistry" but I would expect overtime to see Twitter and Facebook to change it following the patterns seen with other addictions). 

So we could measure different kinds of social media usage in certain populations to quantify this, and I would expect to see a lower amounts in a control group and higher amounts populations that are "very online." (90%) Would also be good to compare to a never online population, if one could be found. 

Then we'd probably also want more evidence that social media is playing a causal role, and would want to make an RCT of people who have never experienced social media and pair them with something thought to be addictive (Facebook) vs. something more inert (e,g, The WELL?)... would that pass an IRB? I don't know. Maybe something more longitudinal with people in the first group to see if their social media use increased or decreased and if that was correlated with ΔFosB.

I read studies, but have never designed one, so I'd want to talk with some who has more experience. 

But, short answer: yes, my argument here is falsifiable.  

As far as what's more or less addictive, FosB may help out there as well, but I'm not as optimistic about it. I don't study addiction professionally, but I believe there's something like an interaction between an addiction phenotype and a particular drug or behavior. I may be reading that all wrong and talking out of my ass here, but at first glance it strikes me that way. So what's very addictive for someone may not be as addictive for someone else. 

Looking on Google Scholar, in the last 10 years or so it does seem like there's more interest in liability to addiction (aka addiction liability) as in who's predisposed to what. One specifically on IT here


But forget all of that for a second and just ask yourself "What incentive does Twitter have to not make their service as addictive as possible for as many people as possible?" Ethics? Guilt? Not wanting to zombify all of their friends and family? There's very few I can think of. The money is all in discovering how to get people more and more engaged. 

In some ways Twitter is worse than cocaine because cocaine is more or less always cocaine. Cocaine doesn't monitor you to see how you use it and change itself or the way it's delivered to become more addictive in ways that are intentionally imperceptible to you.

Right now, you may not fit a phenotype of someone who can become addicted to Twitter, but expanding the circle of the would-be addicted to include you is likely just an engineering problem. 

Then we'd probably also want more evidence that social media is playing a causal role, and would want to make an RCT of people who have never experienced social media and pair them with something thought to be addictive (Facebook) vs. something more inert (e,g, The WELL?)... would that pass an IRB?

One more ethical setup would be to use existing users and do a study of how many succeed with an intention of stopping to use a given website. 

If a lot of people who decide to do a two week Facebook fast fail to follow through on the intention, that would be a sign of it being very addictive. 

Your study can do that with a bunch of different behaviors and then rank how successful people are with the fast.

Ok. It's falsifiable but requires expensive evidence you don't have an easy way to collect. What do you propose we do? Like you said, the economics currently reward sites for being as addictive as possible. Moreover, better ai methods will allow generation of even more addictive sites. Or just optimization - compare tik tok to YouTube. Tiktok is a set of very short clips that have become super clickbait through careful winnowing. Every single clip has someone in the top 1 percent of attractiveness... maybe higher than that...or something immediately violent or dramatic or funny.

It would be like the kind of junk food you would get if you had realtime feedback to what people were eating and had millions of samples to pick from.

In the case of Donald Trump, supporting at least a temporary ban (I'm on the fence about whether a permanent ban is appropriate) is more optimally moral than continuing to allow him to use Twitter. I see two reasons for this. More explicitly than I stated in the article -- first, it's better for Donald Trump's mental health 

The second, I just re-read the reasoning behind the Twitter decision to ban him on the grounds of their glorification of violence policy, and in the context of where the country was on January 8th, it 50% makes sense (and that 50% is the part that counts in terms of the ban). The 50% good - It reads like a company that's self-aware of how dangerous their platform can be and is trying to reduce the proximate harm. The 50% bad - it's taking no responsibility for their hand in creating software that coaxed it to that point and profited from it all along the way.

I realize there's plenty of other accounts saying awful things on Twitter, but if you know Wikipedia policy that sounds like a WP:OTHERSTUFFEXISTS argument. That would be an argument where the existence of something bad somewhere else is used to justify allowing something else bad to continue.


In the case of social media at large... I share the view that Ted Nelson is perhaps the worst project manager in the history of software development, but his ideas for building a network that allowed users to generate wealth through micropayment royalties is one of the best yet unimplemented ideas. 

If you write a typical clickbait buzzfeed-esque article it gets shared, makes some money form ads, and then is likely forgotten... not to mention it has all of these negative externalities (e.g. using fear and outrage to get attention).

A micropayment royalty system would incentivize people to produce content that's valuable rather than clickbait... sort of like if you write a good journal article, it get's cited, quoted and has not just a long tail effect but a kind of a stacked cumulative long tail effect from not just the views on the original piece of content but from the views of everything that incorporated it. We want people to make more content like this and to get paid for doing it, but there's not good mechanisms to pay them for this now, although people are working on it. 

Of people working on this now, Leemon Baird seems to get this

Then you can talk about having very cheap transaction fees, which in turn allows you to do whole new kinds of transactions in the world, and so you can talk about a world where maybe instead of playing games and going to websites and watching videos where you're paying for [00:28:00] by watching ads, you pay for it by paying one hundredth of a penny. Does that matter? Well, guess what comes along with paying for it by watching ads? Whoever is showing you the ad, has a very strong motivation to spy on you, that's just inherent. People have said, “You know if you're not paying for this service, you're not the customer, you're the product."

I realize there's plenty of other accounts saying awful things on Twitter, but if you know Wikipedia policy that sounds like a WP:OTHERSTUFFEXISTS argument. That would be an argument where the existence of something bad somewhere else is used to justify allowing something else bad to continue.

When it comes to cases where a person has a duty to treat people fairly, other stuff exists is a valid argument. Black people are jailed much more for drug use then white people is a valid argument to be made against a black person being jailed for drug use. 

Do you think that Twitter doesn't have such a duty to treat people with equal standards?

A tweet by the PotUS has large effects, the same tweet by me would have none. Should the tweets be equally judged by just the actual words, or equally judged by the effects they are likely to have?

What has that to do with what I wrote and whether WP:OTHERSTUFFEXISTS applies here?

It was a response to your final question about treating people with equal standards, drawing attention to the ambiguity of the concept.

When it comes to cases where a person has a duty to treat people fairly, other stuff exists is a valid argument. Black people are jailed much more for drug use then white people is a valid argument to be made against a black person being jailed for drug use. 

Agree that case is a valid argument. More (but not perfectly) analogous to this situation would be if someone was arrested for drug use and said "hey, I know this other person that's been using drugs way longer than I have and they haven't been arrested, why are you arresting me why you should be arresting them for being the more egregious offenders?" 

Do you think that Twitter doesn't have such a duty to treat people with equal standards?

I'm not sure if you mean this in a normative or descriptive sense. In a normative sense, yes, and I would apply this to people and corporations.

In a descriptive sense, what incentives does Twitter have to treat people with equal standards? Being morally good? Will that make their shareholders and advertisers more money? Would the board of directors suggest it? Capitalism doesn't necessarily optimize for the moral goodness of products.

There are ways to use capitalism to improve capitalism. There's a lot of "Metcalfe's law determinists." Like "there is no alternative to this network because it's the large network." But any network is just a spot on a much larger canvas of possible networks, any number of which could make money and make it more morally. 

Clubhouse isn't a perfect example, but it does show that people are willing to use networks operating in different paradigms. Sure, Clubhouse didn't stick the landing after it's growth spike, but someone else could.

How is facebook not external?

New to LessWrong?