Facebook's problem has never been engineering. Blame its leadership and the product decisions they made (like dumping $4B a quarter into a "metaverse" that nobody wants).
Heck, a big part of TikTok was Musical.ly, an app for lip syncing videos. That was a huge stroke of luck/foresight because it gave TikTok a lot of the music rights deals they would need to allow people to put videos to music.
Are you really arguing an app around individual videos set to music (which, in my experience, is still the vast majority of TikTok videos) is somehow an "incremental product change" to a friends-based social network?
1. Content that is profitable from an advertising perspective
2. Censoring certain content/only showing major/authoritative sources.
TikTok changed that by allowing grass roots content to go viral, which from an algorithmic perspective is less profitable. But that doesnt matter, its backed by the CCP. Theres alot that goes into why tiktok is better, and alot of it is monetary/business related, not computer science.
But beyond that, FB/Google just got lazy and stopped pushing for new things. Wildly successful and presumed monopolies, they left a massive product gap (allowing grass roots content to go viral), which TikTok filled.
They hype around it is because it is the current thing.
[0] https://futurism.com/tiktok-spy-locations-specific-americans
Totally unimpressed.
With all due respect — Facebook (the product) is one of the social platforms furthest from TikTok. Even roughly equating them sounds like you might not use, know, or understand any of them. Vine, YouTube, or Snapchat maybe — they all differ in different ways — or even Insta I get ... But Facebook?
This is straight up wrong. Both companies ask difficult algorithms questions in their interviews.
This is going to be the next famous “nobody wants Dropbox” comment.
I don’t know if Horizon Worlds is going to be successful, but Reality Labs is much, much bigger than that. Do you honestly believe no one wants VR and AR? Seriously?
Think about gaming alone. Think about sitting on an airplane with three 50” monitors overlaid with your vision, where you can actually code, or do whatever. Travel, education, shopping, entertainment. The possibilities are endless.
This stuff is basically already possible today, and maybe a couple years away from a legitimately good experience.
Meta is way out in the lead on this. When the media talks about the spending on Metaverse, they’re always imagining Horizon Worlds. I think it’s entirely possible Horizon Worlds becomes successful. I mean, my company does some watercooler meetings in gather.towns, where you have a tiny little 32x32 pixel avatar. It’s entirely possible in a few years we’ll be doing these in Horizon Worlds. But even if we’re not, and that’s entirely bunk, Reality Labs is still the leading VR headset manufacturer, with a ton of software to make it all possible, still has an entire games division in Oculus Studios, and much more.
People definitely want this stuff. It’s early days still, and Reality Labs is definitely burning cash rapidly, but they also sold a billion dollars of headsets over the past year.
Until then, VR is a nausea inducing gimmick like 3D TV that adds precisely nothing to any kind of human interaction.
Augmented reality... that has some actual uses. VR is not even that good for games, let alone the fantasy of having a meeting enhanced by using it.
To put this in perspective, there were 8 million PS5 sales in 2022, and 6.5 million Xbox X.
I realize this is not quite an apt comparison, and I’m lumping all VR headsets from all manufacturers together. But this is rapidly becoming a big deal.
Look, I’m not necessarily claiming Quest VR headsets will be the next iPhone. But I think it’s pretty likely this will be a big and important product over the next few years.
That's not important question, that question is, do enough people want this that it will pay off.
I mean I heard that 3D TV was the shit a decade ago, but we don't much of that around.
What will anyone do with that, except watch funny cat videos?
The resolution is not quite there yet, but it’s getting close.
Horizon Workroom, along with Horizon Remote Desktop are tools that allow you do real tasks in VR/AR.
Funny cat videos it is.
The common feature of metaverse bulls seems to be asserting without evidence that pixel density will increase massively from its current state, such that emulating several monitors will be viable.
I see no reason to believe this is true. It’s not the case that all technology you can imagine will eventually exist. Things always level off eventually, which is why I don’t have a 10 GHz processor today.
If they can achieve that, people are going to start doing real work on the headsets.
Note that the 1200 DPI is per eye.
I would much rather look out the window, read or sleep.
Did you bet money on that?
e.g by purchasing meta stocks
Facebook makes zero effort to retire questions that have leaked on Leetcode and similar sites, so memorization really does give you a huge advantage. All the coding questions I regularly asked people were literally on the “Facebook” section on Leetcode.
Basically it will filter more bad than good but it will still filter good.
- Cassandra
- Hiphop VM
- Flow
- React
- GraphQL
- zstd
- PyTorch
- rocksdb
I hate Facebook to my core, but this is pretty undeniable. You gotta re-examine your priors on this one.
I don't deny the engineering innovation of Facebook but it must be moderated to some extent by the way ideas are flowing between engineers at mega-corps, probably from at least Bell Labs, if not earlier.
Amazon does not produce the same level of open source, or anywhere close, but the products are more varied and far better.
> I still, after all these years, hold the belief that facebook failed to innovate because the engineering interview is based on memorization. Having worked with those types of engineers, they can absolutely build exactly what is needed.
I think the fact that so much engineering innovation has come out of Facebook absolutely disproves your thesis. Also, while I've interviewed for Amazon (but not FB), I don't think Amazon's technical interview questions were really any different from any other big tech companies I've interviewed at. They certainly didn't seem to optimize for any sort of "innovation" in their tech questions (to be clear, I'm not sure how you'd do that).
Meta is product-driven not engineering-driven. How engineers are hired doesn't really matter because it's generally not engineers at the helm.
Meta is in the position they are because post-WhatsApp there has been no company strategy. First it was the post-2016 election fallout of fighting misinformation, which was never going to work and isn't really the problem anyway, followed by a brief foray into crypto and finally the eye-watering boondoggle that is the metaverse, the answer to the question no one asked.
None of that has anything to do with hiring engineering ICs with whiteboard coding interviews and it's crazy to suggest it is.
"Meta is product-driven not engineering-driven."
My point exactly. Divorcing product management from technical implementation is a fools errand in an extremely technical company.
When every engineer is focussed on execution, it adds up. Interesting engineering led ideas cannot lead the company in new directions.
you don't think dangling a massive salary behind an easily gamed interview format can lead to problems and attract the wrong type of people to a company? FAANG is now attracting people who would have been on Wall Street 20 years ago. They are money chasers, not the type of people who actually build things
>Meta is in the position they are because post-WhatsApp there has been no company strategy
>How engineers are hired doesn't really matter because it's generally not engineers at the helm
you don't see the dissonance between these 2 statements? Maybe if you had builders with opinions instead of coding robots who follow instructions they wouldn't be such a mess
Disclaimer: I'm a software engineer myself.
I think if a company considers product innovation as part of its software engineers duties, they have already failed.
Software engineers innovate on technical solutions, they are the bridge between innovative products and their implementation, and it happens that they can bring out innovative product, but innovation is the job of product people, not software people.
I know some on HN would like to believe that software engineers are the ultimate worker that can and should do everything and anything, but it's not the case.
The best engineers I’ve seen are doing far more than just solving coding and system design problems for existing issues.
Also a software engineer, though I’ve moved to the management side (I miss the engineering focus).
Well it's a good thing neither of those are particularly necessary or even going to be received well at them.
After all - this is why you have PMs. It's not engineers who are dictating the direction product is going. It's the PMs and execs.