site banner

Culture War Roundup for the week of November 14, 2022

This weekly roundup thread is intended for all culture war posts. 'Culture war' is vaguely defined, but it basically means controversial issues that fall along set tribal lines. Arguments over culture war issues generate a lot of heat and little light, and few deeply entrenched people ever change their minds. This thread is for voicing opinions and analyzing the state of the discussion while trying to optimize for light over heat.

Optimistically, we think that engaging with people you disagree with is worth your time, and so is being nice! Pessimistically, there are many dynamics that can lead discussions on Culture War topics to become unproductive. There's a human tendency to divide along tribal lines, praising your ingroup and vilifying your outgroup - and if you think you find it easy to criticize your ingroup, then it may be that your outgroup is not who you think it is. Extremists with opposing positions can feed off each other, highlighting each other's worst points to justify their own angry rhetoric, which becomes in turn a new example of bad behavior for the other side to highlight.

We would like to avoid these negative dynamics. Accordingly, we ask that you do not use this thread for waging the Culture War. Examples of waging the Culture War:

  • Shaming.

  • Attempting to 'build consensus' or enforce ideological conformity.

  • Making sweeping generalizations to vilify a group you dislike.

  • Recruiting for a cause.

  • Posting links that could be summarized as 'Boo outgroup!' Basically, if your content is 'Can you believe what Those People did this week?' then you should either refrain from posting, or do some very patient work to contextualize and/or steel-man the relevant viewpoint.

In general, you should argue to understand, not to win. This thread is not territory to be claimed by one group or another; indeed, the aim is to have many different viewpoints represented here. Thus, we also ask that you follow some guidelines:

  • Speak plainly. Avoid sarcasm and mockery. When disagreeing with someone, state your objections explicitly.

  • Be as precise and charitable as you can. Don't paraphrase unflatteringly.

  • Don't imply that someone said something they did not say, even if you think it follows from what they said.

  • Write like everyone is reading and you want them to be included in the discussion.

On an ad hoc basis, the mods will try to compile a list of the best posts/comments from the previous week, posted in Quality Contribution threads and archived at /r/TheThread. You may nominate a comment for this list by clicking on 'report' at the bottom of the post and typing 'Actually a quality contribution' as the report reason.

12
Jump in the discussion.

No email address required.

Twitter dies for good in the next six months: 80% probability

By now you know that Elon gave staff a deadline of today (Thursday) to either commit to being "extremely hardcore" or leave (source). Unsurprisingly, most people - roughly 75%, according to some Internet rando - didn't take him up on this. Elon blinked and apparently people still have access.

That won't do much (WaPo):

“I know of six critical systems (like ‘serving tweets’ levels of critical) which no longer have any engineers,” a former employee said. "There is no longer even a skeleton crew manning the system. It will continue to coast until it runs into something, and then it will stop.”

But that's not even what I was going to write about, just what happened while I was composing the post. (Also let's put aside that he said "microservices are bloat" and then they killed the microservice serving SMS 2-factor login.)

To me, the biggest news is that he axed 80% of the 5500 contractors (source, Casey Newton, or someone with a premium account impersonating him I guess).

The contractors were responsible for things like moderation (source: what are they gonna do, use salaried employees?). If you don't have moderation for basic things like CSAM, you're boned. I know a thing or two about moderation, and if you let the Internet type into a text field, you get some dank shit. And crucially, you can't automate it away, because there's a human on the other side working to defeat whatever you're doing. I mean, the YouTube comment section probably has some of the most expensive automation on the planet working on it and the spam still gets worse every day, and I'm talking the obvious stuff like "HIT ME UP ON TELEGRAM <number>". The only thing that saves you is humans clicking buttons (and getting PTSD, but let's skip that for now). Google had 101k employees but 121k contractors as of March 2019, and that's what the contractors do, click buttons.

If you don't have moderation, you don't get the YouTube comments section, because they at least have contractors backed up by code (at the cost of many expensive engineer-years). You don't even get 4chan, because they at least have Those Who Do It For Free. You get some ungodly shithole most younger Internet users have never experienced. You're getting... the virtual equivalent of your local Greyhound terminal. Whatever happens to someone's chat room side project that gets posted to /b/. Sludge.

Twitter will have to either restrict posting to an unbearable degree or watch as the remaining users get tired of slurs in their replies and bounce.

Remember when Elon was just going to clean up the bots on Twitter?

(Reason for posting: I saw some takes elsewhere on this site that apparently Musk would lead Twitter to success or at least improve it or something, and disagreed.)

People trust Musk because he has been instrumental in creating three huge businesses (and another that would be impressive for most people). There is a track record here. Either musk is the luckiest man in the history of the world or he has business chops in the top 1

% of the 1%.

I'm not sure why it can't be a mixture of skill and chops, or why chops in rocket building would necessarily transfer to social media, or why a talented person can't make big mistakes. It's equally plausible to me that Elon Musk's ego has swollen to the point where he thinks he can run Twitter with him +50 people, as it is that he really can run Twitter with him +50 people.

From a technical standpoint is running Twitter on 50 people so implausible? I imagine new features would be released very slowly, but I think that's enough to maintain the site, so long as moderation is relaxed quite a bit and mostly delegated to AI.

They have their own datacenters, at least three of them, with hundreds of thousands of servers. 50 people would barely be enough to do the hands-on work of managing one datacenter.

Just swapping failed hard drives would probably take about two people datacenter. Assume 300k machines, 100k per datacenter, with ten percent of that footprint being machines with lots of disks, let's say thirty disks per machine. AFR for drives that are being constantly hammered as I would expect them to be is about 3%, especially if you try to stretch the lifetime of the drives.

10,000 machines * 30 drives per machine * 3% AFR = 9000 drives per year, or about 24 per day. Let's say each drive swap takes 30 minutes from the point of receiving the ticket, picking up the replacement, performing the swap, and some wiggle room for complications / ticket re-opens.

So an eight hour shift can do about 16 swaps per day. Typically datacenters don't staff drive swappers around the clock, so you need at least two people just for drive swaps.

Those people have no knowledge of how the replacement drives are stocked, someone else needs to do all the ordering of spare parts, and at least one person needs to receive and stock them. So at least three people per datacenter just for drive swaps.

Then someone needs to handle the relationships with Western Digital, Seagate, Hitachi, etc. If you don't want to be surprised as you roll out new drive models you'll need at least a couple of people whose job is to qualify new drives by abusing them for a few months and running real workloads on them.

And someone needs to handle the OS / software side of the drives. Handling firmware updates, drive settings, all the automation that goes into failure detection, handling replaced drives, investigating problems, etc. That's at least a couple of people.

Someone needs to write nad run the inventory system that cuts tickets and registers spare parts as well, let's say another two people.

We're up to 9 people at the datacenters and 7 people outside. 16 total. We haven't even written the software that's going to be using the drives yet. And this is just for hard disk-based storage, and I'm being extremely conservative. In reality two people is not sustainable for any team that needs to be oncall, you'll burn out quickly if you're oncall half of the time.

I would be pretty surprised if these jobs weren't handled by contractors. I haven't read about any ex Twitter employees talking about how difficult it will be to replace their invaluable hard drive swapping experience.

Drive swappers almost certainly are contractors, but they're still people. The claim I was responding to was about people, and most claims of this sort are also about people.

The distinction between contractors and employees is pretty arbitrary, and makes for impossible to disprove claims. If I can sneak in contractors I could plausibly say that Twitter could be run by one employee (and 10,000 contractors).

I think it's totally plausible from a technical standpoint if those 50 people are talented and hardworking. All the hard stuff for a basic scaled social media platform is a cloud API or open source library these days, and Twitter is past its startup design sprint anyway. A low cost structure would give him a lot of headroom to ignore the problems that his employee bloat is aimed at addressing right now.

The issue is that a 50-person giant social media empire isn't a stable equilibrium. It's similar to why you rarely see charities that spend ~100% of their funds on the cause: if spending a dollar on marketing gets you more than a dollar in increased contributions, then why wouldn't you do that? Likewise if you had a scaled 50-person giant social media empire, then the return of hiring a marginal employee is much greater than the cost.

The issue for Musk specifically is that he needs to pay a billion dollars a year in interest payments for the leverage that he took on in order to acquire Twitter. So he doesn't have the luxury of running the shop at low cost and telling advertisers to take it or leave it.

Since he needs advertisers to pay his interest, he needs to solve a lot of the messy social problems that drove a lot of that employee bloat in the first place. He needs good sales teams. He needs good marketing. He needs good moderation to keep the tone of Twitter consistent with advertisers' brand expectations. He needs giant compliance teams to keep up with the onerous, schizophrenic, internally inconsistent and offensive regulations imposed by the likes of Europe and India. None of those are purely technical problems; they require giant teams of people, and (pre-AGI) they always will. Why? Because they aren't static goals; they're adversarial goals with elements of competition. They're basically a policy market, in the sense that if it could all be automated, advertisers and regulators would have more headroom to increase the onerousness and contradiction of their demands until it couldn't. The only check on advertisers' demands of Twitter is how they compare with other social media platforms in terms of brand safety, and the only check on regulators' demands of Twitter is how onerous their regulations can be before Twitter will go dark in their country (or before the US government initiates WTO actions against them, and no one is betting on Biden's willingness to bail out Elon Musk in the international policy market).

I dunno if he'll pull it off. I suspect he will, but no outcome here will surprise me.

One outcome that particularly wouldn't surprise me is if Musk capitalizes on the chaos to threaten his lenders with bankruptcy, and uses that threat to buy out his debt for 20-50 cents on the dollar. Good luck marketing this debt, guys: no one has ever demonstrated that Twitter can have positive economic value, and a high-profile failure by Elon Musk isn't going to increase anyone's estimation of those odds. Then his cost structure becomes much simpler and he can tell advertisers and foreign regulators to get fucked, the prospect of which at this point I am sure provides him with near-sexual arousal.

Reddit had 30 employees in 2012 and it was a far more complicated application then than twitter is today.

When Instagram was acquired for $1 billion in 2012 it had 13 employees.

Twitter by this point has a lot of technical debt and cruft, so 50 does seems like too few. But less than 1000 seems very doable. One reason that so many people want this to fail is that they're afraid that Musk is right about these workers being worthless.

I didn't say it was implausible. I'm 50/50 on whether running Twitter with 50 people is possible or not. But I don't think it's convincing to appeal to Elon's genius to argue that it proves that it's possible.

Very implausible. Sure, a single person can build a Twitter clone (timeline of Tweets from people you follow) in a weekend, but there's a whole lot more to it than that. Realistically something like Twitter might be doable with 500 engineers (hard to say; not sure of the entire set of features), but you need a whole lot of moderation, marketing, business associates, etc. on top of that if you actually want to make money.

To elaborate, you need people handling identity, authorization, infrastructure, payments, internal tools for CSRs and moderators, customer/business facing tools, third party integrations, APIs, auditability, security, ad placement strategies, site reliability, data stores, build/test/deployment systems. Take any of those away, and you don't really have a viable business. And that doesn't even touch on new feature development (which to be fair hasn't been that important once Twitter found its niche).