HACKER Q&A
📣 8f2ab37a-ed6c

Twitter was unanimously predicted to implode. Nothing happened. Why?


Twitter laid off 50% of its staff. Principal and staff engineers made lengthy Twitter threads predicting the collapse of the system given its nuance, complexity, and how their positions were key to its survival. Pundits predicted that the site would likely not survive a few weeks, that there was no fat to trim at the company and all the workers let go were essential for daily operations.

Now that it's been three months, what happened to the dreary prognosis? Where is the continuous downtime? Where is the permanent feature freeze? Where is the imminent collapse we were told about? I noticed that nobody is talking about how none of those predictions seem to have manifested. Why?


  👤 steviesands Accepted Answer ✓
If you are interested in the genuine impacts and not a hyperbolic, emotionally charged discussion best left to pundits, then below are some areas I would suggest researching and conducting your own thought experiments. A good exercise is to pretend if you were a CTO presenting to the CEO the tradeoffs and risks in cost cutting.

- impact on reliability (sum across users for key metrics, not anecdotal evidence)

- systemic risk changes, measured using probability of a rare event

- feature velocity, volume of features, 'width' of features (how_many_bets x feature_surface_areas)

- the durability of systems engineered for reliability or "did we build systems that will survive if everyone leaves"

- revenue

- advertiser satisfaction with the technical platform

- impacts to internally maintained systems

- impact of loss of institutional knowledge

- impact of further attrition post-RIF

- compliance

- rate of progress vs competitors as an advertising platform

- rate of progress vs competitors in social media

- (related to above two) ability to maintain competitives advantages with reduced headcount

- probability of success for new revenue streams

There has also been recent reporting on Twitter: https://nypost.com/2023/01/18/twitters-daily-revenue-plunged... https://www.axios.com/2023/01/29/fidelity-cuts-twitter-valua...


👤 hayst4ck
I'm not sure many people who actually worked on reliability expected imminent implosion.

It's more like thinking the seed of eventual destruction was probably planted.

From a reliability perspective, there's probably two interesting forms of failure here.

One is latent, like shutting off backups. Like yeah, you can shut them off and nothing bad will happen imminently, but eventually something bad will happen, and that will be a very bad day. Firing an entire team of experts in a system is kinda like getting rid of backups or probably closer to making backups but not testing them. If you don't test your backups, maybe you will be fine, or maybe you won't.

The other is corrosive. Imagine upsetting a companies most impactful devs while simultaneously shrinking growth prospects. How are you going to hire amazing devs when the devs you have left are kinda mediocre or working due to coercion (h1b) rather than more positive reasons? What do you think Twitters sales pitch is to potential new hires?

Reliability/security/etc are all types of maintenance where most of the time you don't have a problem, but sometimes the little hook on your power line fails, resulting in the live cable hitting trees that weren't trimmed, on a forest rife with fuel that results in billions of dollars of damage. The lower the investment, the higher the chance of failure, every individual failure has the chance to compound with other failures into catastrophic failure.


👤 qup
Because the predictions were mainly people hating on rich people. It's very en-vogue. We're mad at all these [celebrities, athletes, CEOs, trust-fund-babies, techbros] for all these "reasons," like I really care if the guy who scored the touchdown is a good dad or something. And just look at the legions of people ready to call anyone a total moron who is on display.

So, Musk bought Twitter with his 44 Billion, and legions of people jumped in line to talk about what an idiot he was for buying it at that price, as if they (or anyone else) know how many billions twitter is worth, or as if they know how billionaires should spend their money.

Then they predicted he would fail immediately because, since he's an idiot and just fired pretty much the whole staff, he can't possibly run something as complex as Twitter.

But then he does, and so they start talking about how he "lost 100+ Billion" between the purchase price and what they perceive as a resulting dip in Tesla stocks. Statistically, 60% of these people can't handle a $1000 financial emergency in their life, but they're experts on big business and billion dollar companies.

Then they started saying they were going to quit twitter, and switch to Mastodon. Some even made the switch, although most seem to be posting in both worlds. But they were being smart, establishing before the inevitable collapse of Twitter.

Then they quit talking about it, because it didn't happen.

The thing is, this whole thing really has little to do with Twitter or its operations. Most of the people commenting on Musk/Twitter have zero fucking clue what twitter is like internally, technologically, what it needs to maintain (tech and culture both), how to improve it, etc etc. They are just hating on rich people.

Ridiculing people on the internet is also a really easy way to get a lot of engagement.

See also: constant disparaging comments about pg (pseudo-intellectual), Bezos (tiny man), Bill Gates, all past presidents, basically anyone making public decisions who is wealthy, all the conspiracies about eating babies and shipping children via wayfair, and on and on for days. This is also largely extending to any unliked groups, like "Russians" and "republicans" will be thought of as less-competent because "we don't like them." Some of these hated groups constitute about half the population. They can still be perfectly competent.

It didn't happen because it turns out Musk probably does know how to run companies.


👤 stillkicking
Twitter hired tons of employees since 2020 so the apparent massive layoffs simply weren't. And there is a lot of dead weight at tech companies regardless.

Political pundits had a direct interest in sabotaging the perception of Twitter as an authoritative source of new information, and the coordinated attempts to get advertisers to pull out early on also point in this direction.

The reason they aren't talking about it anymore is because, by and large, this plan failed, and those predictions weren't predictions but wishes. Some groups slinked off to Mastodon, which works for niche communities, but doesn't work for the site at large.


👤 klyrs
It's only been 3 months. Their ad income is reportedly in the pits, and $8/mo from users isn't going to cover that loss. Twitter took on loans to cover its own buyout, service of which is around $1B/y. He's being sued for not paying rent, and various issues around layoffs. Musk has been talking about filing bankruptcy to skirt financial obligations. So, count me among those who see Musk's endeavor as fraught. But I wouldn't bet on a timeline. Maybe he'll follow through on his poll, and relinquish the CEO position to an adult, and I might change my mind.

👤 matt_s
From a technical perspective it hasn't been long enough for systemic failure to happen, financial failure might happen sooner.

Think about how much day to day effort is spent on keeping systems you work on running. Not new feature development or bug fixes or running reports or weird user errors but actual effort to keep the system "up". Its not much from my perspective. It will take time, perhaps long periods of time, before the tech stack starts aging past EOL dates for supporting elements it runs on (OS, DBs, messaging stacks, web frameworks, etc.) If twitter runs on their own built hardware vs. cloud services that are maintained by others then they probably have even more runtime because another entity won't be rolling cluster and other upgrades to their infrastructure.

The gloom and doom predictor people won't want to admit that twitter can train and staff up people over time so that site reliability won't be a factor. What should concern folks here is non-technical management at other places thinking they can do things like this as well. Don't think for a minute that what we do is specialized to the point that they can't let you go immediately.


👤 rvz
> Now that it's been three months, what happened to the dreary prognosis? Where is the continuous downtime? Where is the permanent feature freeze? Where is the imminent collapse we were told about?

I remember that. False predictions like this: [0]

Because their false prophesies of the imminent and total collapse of Twitter has not come true, the same pundits are still tweeting on Twitter and are moving goal posts; expecting 1000% uptime. One outage and it is the end of the world.

To those pundits, Twitter now has a new standard of uptime. While the 220M+ users are still using the platform, the so-called competition or 'Twitter-killers' have failed to make a meaningful dent on it's daily active users.

Perhaps too many screaming voices got in the way of the fact that Twitter was actually better for less than 1,000 employees.

[0] https://www.technologyreview.com/2022/11/08/1062886/heres-ho...


👤 mikewarot
As Grady Booch pointed out, Elon is testing in production[1]. Even Elon doesn't know where all the filters are in the algorithms used.

The absence of a total collapse does not imply all is well.

[1] https://twitter.com/Grady_Booch/status/1620720537805922306


👤 barbazoo
> Twitter was unanimously predicted to implode.

There's nothing unanimous here


👤 zamalek
Last I read, Twitter has about 2 years of runway in the bank (based on the loan repayment costs). They will be forced to deal with bankruptcy eventually, there is little chance of anyone scaling it to what is required to pay the interest on the loan. As of right now/short-term, though, it is financially safe.

To make matters worse, Musk has not financially scaled it in the direction it needs to go in order to pay the interest once the runway runs out. Revenue is currently trending downwards, mostly due to scaring off advertisers (who were basically solely shouldering the cost of keeping Twitter afloat). It seems that he doesn't understand that he needs to scale revenue upwards, never mind the historically steep gradient of growth required.

The odds of the banks shutting it down are slim, they do ultimately want to recoup their costs. I imagine that they'll out Elon when they have the legal tools to do so, and replace him with someone who at least understands that "LESS PROFIT BAD, MORE PROFIT GOOD, ME NOT PISS OFF IMPORTANT CUSTOMER."


👤 advisedwang
All those SREs have been spending the last X years making sure twitter does NOT rely on some humans doing regular maintenance to stay up.

People who predicted it's demise were foolish and were not a consensus at all.


👤 celticninja
We don't know that nothing happened. I'm guessing that whilst they reduced their costs by sacking staff they also lost a lot of ad revenue. They can paper over cracks for a while and still be imploding. It's notable that 'new' features have been about charging users or cutting costs, e.g. blue checks and less moderation

👤 rsynnott
Well, if I look at "for you", it's showing nothing newer than 11 hours. And it's mostly just this random crap. The chronological feed seems to be working, albeit slowly, though it was just spinning forever earlier (though, it seems... a bit empty; I'm not sure if the people I followed have mostly quit or if it's just broken). The "notifications" section is just showing me a bunch of posts by one account I don't follow. The ads are all for either scammy-looking stuff or, weirdly, lots of small obscure AI startups.

What I'm getting at is, define 'implode'. Does it literally still work? Well, yes, mostly. Is it working at all well? I mean, no, clearly not. Is it the sort of site I can see myself still using, after 15 years of tweeting? No.


👤 dxbydt
Taleb says the turkey lives happily for 364 days. Then, there's Thanksgiving.

👤 pedalpete
Though the tech layoffs have hit the development and infrastructure teams (I mean all the tech layoffs, not just twitter), it is my understanding that the majority of people laid-off are in sales, management, and administration.

Users won't see the impact of this as much. Perhaps some slower pace of releases (twitter was already painfully slow from what I understand), slower hiring, etc.


👤 FrontierPsych
The revenue is down 35% from last year, but what is their profitability? It's not a public company anymore, so revenue goals...who cares about them? A private owner is more concerned about profits. So did the shedding of all those positions either give profit now, or set up profitability in the future? I think so.

Planning for long-term - who cares if 500 advertisers left? The exodus of advertisers is why other advertisers should expand their advertising. Less eyeball competition, and better deals.

90% of the Fortune 500 companies that were at the top 50 years ago are no longer there, or whatever that number is. They stagnate. They don't shake themselves up. They go along to get along.

I think that a lot of people want Musk to fail, just because. People take glee in the hope that he will fail. Especially now that they don't agree with him politically. Which, for the record, I think that it was monumentally stupid to bring politics into it, on his side.

The reality is the guy spend $44 billion of his own cash. He has built many HUGE companies that he owns. I think the guy knows what he is doing. He might fuck it up, but he knows more about running huge companies than you or I, that's for sure. Could he fail? Sure, of course. But the guy put his money where his mouth is. They guy is not full of shit, he ponied up $44 billion.

He has done things that you, nor I, nor anyone else could do in 100 lifetimes. Oh, I've read the nay-sayers - he gets the public to fund things, the government to fund his things, and this and that. Well, why don't you or I, knowing this, go out and make a bunch of multibillion dollar companies that change the course of society the same way that he did? Not possible, otherwise you and I would have already done it. Except, of course, if we actually did, everyone would LOVE us and our decisions, because we would be pure and holy. WE would do it right. hahahahahaha.....sure, sure.

I'm not a Musk fanboy, but I think he did some impossible things. I just think the whole hoping he fails because of his politics is hokey and immature.


👤 aappleby
Complex systems decay over long periods of time.

👤 MagicMoonlight
It’s a finished product, you need like 10 employees total to run it. Techlets just don’t want to admit that they don’t really anything.

👤 jamesboehmer
Yet.