Hacker News Reader

Comment by asciimov | Hacker News Reader

Does Facebook spend any sort of money on R&D? I just can't fathom a company that once had a near trillion dollar market cap, not spending any of their resources to try to find their next big revenue stream. They should have been able to pivot to a plan B, (or C,D,E... etc). If you point to their 3d ventures, well, I've never seen it in person, so they are doing a lousy job of marketing it.

View on HN

RealityLabs is a $10billion initiative with only about $2 in revenue (Quest 2 headsets come from here).

RealityLabs Research is a large component and indeed the major R&D investment. This was spun out during the 2021Q4 earnings call when the different sectors of the company were reported separately for the first time, following the name change to Meta.

Think the OP is say - over the long term not since the recent somewhat desperate pivot.

Similar to sibling comments: copying your competitors is not R&D, and neither is building a software library that’s an incremental improvement on an existing OSS library.

RL has been there since 2019 [1]. Aside from the previously-mentioned (and not necessarily incremental) React, CassandraDB, PyTorch, etc there is work in terms of egocentric data which is far larger than existing datasets [2], even more fundamental research related to mathematical optimization [3], and of course FAIR (used to be Facebook AI Research), which is now part of RL-R, who recently published work on language models [4].

It’s not all that well advertised but there is a lot of research that gets open-sourced so it’s a bit (understandably) naive to call it copied or incremental work.

[1] https://en.m.wikipedia.org/wiki/Reality_Labs

[2] https://about.meta.com/realitylabs/projectaria/

[3] https://github.com/facebookresearch/baspacho

[4] https://ai.facebook.com/blog/democratizing-access-to-large-s...

A bit off topic but it has to do with R&D, how twitter invest almost 2B in R&D and there are no visible new features with that investment (well this year it seems that they started to add new features, but they doesn't worth 2B, edit tweet, I saw something with gifs?). I'm not a twitter user, so I cannot 100% sure but to me it sees like the same twitter back en 2010 or 2012 (when I first used twitter.

So my questions, is the R&D money invested to keep engineers happy? I mean they can for sure experiment there without any worries.

> how twitter invest almost 2B in R&D and there are no visible new features with that investment

From a financial reporting perspective, nearly all engineering costs will be recorded as R&D.

Google has been trying to diversify for over a decade. They are still mostly an ads company.

And they are doing everything they can to remain such. They have established themselves a reputation of a company which is shutting down services left and right, so it makes very little sense now too try any of their new offerings.

To be fair, cash cows pay a lot

But many of the diversifications help cement the leading position of selling ads. Chrome, Android, Youtube are all critical to the continued success of ads.

It's even more direct. The diversifications are Google Ads, in a very real way. Google financial reporting splits Google ads revenue in an important way that hopefully allows people to see this. Total revenue of Google AdSense, on the non-Google internet, is about 7 billion. Nice, but Youtube ads is ~10% less than that ... and Google's own properties get 5 times more than that in ad income. Ads income from non-Google properties is called "Google network", ads income from Google properties is called "Google Search & other" in the financial reporting.

That's not even counting money someone actually paid to those diversifications, no ads involved. That's another 10-15% of the total ad revenue.

Google's internal websites, taken together, are just shy of 85% of their total ad revenue, and 90% of their total revenue. Google does not make that much money from people browsing the internet, it makes money from people browsing Google. Although not split out, it is widely assumed Google search is a whale here. But how much exactly? Not clear. People keep saying it's going down for a decade now. And of course, a business 1/10th as good as Google search is still a great business.

Yes, Google Ads receives the dollars, but it's the "diversifications" that "only cost money" that brings them in. Numbers like this make one thing clear, 7/8th of AdSense is a way to optimise ad income on Google's own websites, and only 1/8th of AdSense is an actual Ad Network. Google is not the biggest Ad Network, despite their reputation, by a very wide margin. Google is the biggest "TV Network".

So saving money on diversifications as a whole ... I'm going to say this is not what I'd like Alphabet to do. Now saving money on specific efforts, sure. But, frankly, the way to grow Google, even their ad business itself, it to spend more money on the diversifications.

What?! They built a phone, copied or bought every social media they could, invented lots of tech (DB, react-native…), they’re spending everything on the meta verse

That’s why they hired so many people

The tech like RN, GraphQL, doesn’t really bring them any money.

And what is the metaverse really? How is it monetized?

That’s OP’s point, aside from Facebook, everything they have is acquired - Whatsapp, Instagram, Occulus.

And even within Instagram, they constantly just copy other apps - Snapchat stories, Tiktok for reels, etc

Facebook can’t have an original idea to save their life

The meta verse will be monetized the same way Apple monetizes the App Store and anything else that goes on through their platform. A percentage cut of all transactions.

That’s why a lot of companies are trying to make the meta verse work. If they’re successful in owning the meta verse they now have a nice big walled garden on which to extract continual rent from.

It’s also very much not what the average person wants so we’ll see how it goes

Reply on news.ycombinator.com