Theory on facebook failure/Google failure to innovate:
They have the highest leetcode standards out of FAANG. As does Snap. They hire people that basically memorized computer science algorithms. And so their employees are bookish, not innovative people. Apple, Microsoft, and Amazon do not have the same issues with creating new products.
Keith Rabois (top VC), mentioned that FB has always prioritized optimization over new products because of the Ad business. That also could explain their inability to do anything but copy other business ideas.
The opposing point of view would be that Ads on FB/Google are so profitable that working on anything else pales in comparision (in terms of profit generation). And so they dont innovate, but iterate and optimize.
My vote is that excessive leetcode testing selects for employees with no creativity. And while these companies are massive, lower level engineers hired ten years ago do rise to the top and run whole divisions
EDIT:
Counterpoint on how good early Google employees were:
the above maxim does not apply because that was before leetcode became huge, people passing those interviews were actually brilliant, they didnt spend six months studying in a basement
> They have the highest leetcode standards out of FAANG. As does Snap. They hire people that basically memorized computer science algorithms. And so their employees are bookish, not innovative people. Apple, Microsoft, and Amazon do not have the same issues with creating new products.
I don't think this is necessarily true because the people designing products are not the same people creating/implementing them. Even the most innovative company of the planet has some programmers and sysadmins doing "bookish" jobs as you say.
Don't fall in the trap of thinking a smart and well trained tech can't be creative; there's some code out there that is a work of art and can inspire awe in everyone. Don't think that if their work is not evident or easy to comprehend it is not creative.
Theres a difference between creating beautiful creative code and creating an innovative product that people want to use. Its a different type of intelligence, and why you dont see Math professors starting ecommerce/product/clothing businesses. The classic dichotomy between steve jobs and wozniak.
I have worked at a few big tech companies, I have regularly seen engineers drive huge innovation in how something is done. Product people cant always do that, they dont have enough understanding of large scale systems.
Admittedly I'm over a decade out of university so things might have changed, but when I was getting my masters in mathematics several of the math professors had various 'side hustles' doing cool and advanced stuff at startups and in industry or finance.
universities are dominated by anti-capitalist sentiment
I know this is trendy to say and I'm sure there are individual universities where it is true, but I see little evidence for it at large. If nothing else every hedge fund and investment bank in the world basically only hire university graduates, and they can hardly be accused of being anti-capitalist. Every economics course I ever took was based, almost axiomatically, on the premise that capitalism is awesome.
University graduates are not the same as professors.
The econ class I took at Caltech was taught by an award-winning leftist anti-capitalist professor, and if you didn't regurgitate his ideology on the tests, you got a bad grade.
Hedge funds only hire a particular subset of graduates. A lot of left-wingers set aside their ideology to get those high salaries.
My father was a finance professor at two colleges. He was the only faculty member who believed in free markets. He was constantly challenged in this by students and faculty. One time the faculty even challenged him to an open debate - the faculty against him. He took them apart. Faculty and students came up to him later saying they had no idea there was even a case for the free market, and thanked him for opening their eyes.
I wonder where you suppose Sanders' constituency comes from?
>there's some code out there that is a work of art and can inspire awe in everyone.
and the people who wrote it didn't spend thousands of hours memorizing the same stupid leetcode exercises to pass lame interviews for 6 figures job, they worked on real projects
tell me the fast inverse square root from Carmack would have been praised in an interview...
Disagree with the thousands of hours. The tasks are actually pretty easy and take about maybe 5 hours for any competent coder to brush up on.
But that's the problem. If a few hours of exercise can dramatically change someone's results then it's not an acid test and it speaks poorly to the quality of whatever team the candidate would be joining.
All interviews are two way streets and the problem is everyone knows what kind of results pass through the leet code filter and so if it's a bad process, perhaps your best candidates will decline to proceed or accept.
You want people whose time is more valuable than your money. That's kinda the whole point of employee/employer social arrangements.
Leetcode isn't an alternative to real world projects, it's an alternative to textbooks and lectures for beginners.
Zuckerberg was able to build Facebook because he got a good education and understood graph theory enough to apply it to social networks.
In the past, the only people who really understood concepts like graphs were people who endured rigorous four-year computer science curriculums. And only top students at elite schools would learn this. The average CS graduate has a very poor understanding of fundamental data structures and algorithms.
But with Leetcode, anyone can solve 5-10 graph problems and obtain the same knowledge. It's extremely efficient. In a couple of days, you can learn what used to take years.
Zuckerburg stole the idea of the product and knew php. I don’t think he needed graph theory to build it and his education was really only Exeter at that point.
The ones coming up with products are usually executives looking to advance their career. e.g. everyone I know who worked on Google Stadia who cared readily pointed out all the problems from idea to implementation. Management & product design waved off the concerns, said they were in for the long haul, and that the product would be a huge success regardless. And then they got cold feet after realizing how much of a flop it was, and how money they needed to spend to make a real entry into games.
I am no fan of leetcode. I think Google's engineering interview process is mostly RNG. But have you seen mock interviews for the product designers? They're even more inscrutable: https://www.youtube.com/watch?v=qTcXXGJiunA
I was actually impressed at how well Stadia worked. My impression is that they did an excellent job at the technical implementation. A lot of games don't require millisecond response times or amazing graphics and Stadia would have had one of the best anti-cheat systems for multiplayer games, basically for free.
What sank Stadia before it even launched was their asinine pricing strategy. It should have been clear to anyone with a basic understanding of computers that playing a game over the network will be inferior to playing local, but they could have made up the difference with pay as you go pricing and an emphasis on multiplayer games where the graphics don't matter as much. Of course they did the exact opposite, and then decided to double down on bad ideas by ignoring indie developers.
I still think the technology has a lot of potential (well in Stadia's case had a lot of potential), but with that kind of management, even the Covid lockdowns and the GPU shortage weren't enough to save a product that was designed to fail.
You're writing like Stadia's already dead, which it isn't and far from it. I've tried all the cloud gaming platforms and none touch the ease of use and pick-up-play quality of Stadia.
People today are even playing Stadia without knowing it thanks to Google's new white-labeling strategy.
I've had a couple young people complain to me about leetcode interviews at Google. I replied that they could spend a couple weeks going over the leetcode books, and if that was what it took to get a $250,000 job, then that will be the best investment of time they'd ever make.
Besides, one cannot help but learn programming things from examining leetcode solutions.
Google probably also wants to hire motivated people, and people who will study up on leetcode will be a better hire than a lazy one who doesn't want to put in the work.
Nevertheless, it's an empirical question as to whether this method of recruiting produces the employees that maximize long-term profits for the company.
I don't doubt that a major reason for asking these types of questions, as opposed to an IQ test or something, is that it is basically illegal in the US to do so.
It's an empirical question about every method of recruiting. Making a bad hire can cost a company dearly. If there was some obviously better way of doing it, wouldn't companies do that?
Avoiding expensive lawsuits is likely another reason for leetcode gates.
I agree, in the sense that it's less about leetcode skills and more about company politics. I can tell you for a fact that the programming geniuses don't make product decisions, much less so strategy decisions. These products would look VERY different if that was the case.
Instead, the important strategic decisions are left for various opportunistic short termers, like product managers, directors etc. They are incentivized on shorter time scales and are motivated by prestige, visibility and career advancement. What looks good in a slide deck will be rewarded, basically. These people are much more in touch with internal company politics first and the industry at large second. They are shockingly oblivious to their own user base – (YouTube rewind is the best irrefutable proof for this).
Big redesigns and product changes are rewarded for way too early, imo. Common tenure for a PM is 1-1.5 years, enough for a promotion, and short enough to weasel out before they have to deal with the backlash.
> The opposing point of view would be that Ads on FB/Google are so profitable that working on anything else pales in comparision (in terms of profit generation). And so they dont innovate, but iterate and optimize.
This argument is basically the same posed in the Innovator's Dilemma, and IMO it's more on the money at least for Google. The high hiring bar for Google was arguably much higher in its earlier years, when they were still innovating like crazy. Becoming a larger company, having a larger executive team instead of Larry and Sergey, and optimizing for profits rather than moonshots and "creative" projects like Loon, Wing, and even projects with a lot of internal support like Inbox all have an effect trending towards "less entrepreneurial".
People also seem to generally have a pretty big bias towards Cloud not being considered "entrepreneurial", while still having a huge engineering opportunity cost. You're comparing Facebook and Google but Google has spun out a #3 Cloud provider in the last ~7 years that people largely ignore despite that making up almost half of Google's employee count at the moment.
This seems like wishful thinking. Why assume that academic intelligence makes you less creative? Do you think that people who graduate with high grades out of Harvard/MIT/Stanford are less creative than people who graduate from second-tier universities because they are too bookish?
I think the reality is probably that it's hard for any large company to be truly innovative. Once you get to a certain size, red tape and size make you slower. You also have an existing golden goose that you don't want to kill (for these companies, ad revenue), so you are more constrained. Furthermore, you only care to take risks that have massive payoffs... new ideas at Google need to be $1b+ to make a dent and $50m ideas don't really matter much anymore. In any case, are there any large companies that don't do leetcode that you would consider innovative? I'd argue maybe Amazon (1 day delivery) and Apple (M1 and M2 chips), but not so much in their software divisions where the leetcode happens.
> Why assume that academic intelligence makes you less creative?
Perhaps it’s not the academic intelligence that makes you uncreative, but the lack of willingness to play those games means that anyone who does endure the process is bound to be both intelligent and uncreative.
The most intelligent and creative people I know are beyond uninterested in big tech (one even went as far as foregoing millions of dollars after their company got acquired by a FAANG because they couldn’t motivate themselves to turn up after their induction and quit). They have plenty of career and lifestyle options, and most of them are way less driven by money than the average.
You won’t find any of them at Meta.
I don’t think this is just about the environment you find once you get there. I suspect these companies are unwittingly selecting for dullness and failure to innovate from the very start of the process.
I posted this elsewhere, creativity with regards to creating and intuiting what will be a good product is completely different than creating a beautiful algorithm. Math professors arent known for their sense of style. And I agree, size hampers it, but here we have Amazon, Microsoft, and Apple all creating new products that people like. And who you hire as engineers permeates through the culture.
> here we have Amazon, Microsoft, and Apple all creating new products that people like
And you think that those companies don't test proficiency of working with algorithms when hiring? What?
Two of them (arguably all three) make operating systems. That's just one of many parts of what they do where proficiency in algorithms and data structures is a strong prerequisite.
Im not saying they dont test, but not anywhere near to the same degree. You absolutely need to be able to solve some of these problems, but FB takes it to another level. Its almost not even comparable in terms of the rigor.
This does not make sense just in the light that every division, every team probably, tests differently according to their needs. I have a suspicion that you are trying to extrapolate some small anecdotical experiences to whole massive companies.
I think OP, and others like myself, are questioning something else. It's not that being creative is different from being "bookish," it's that the two are not mutually exclusive.
> This seems like wishful thinking. Why assume that academic intelligence makes you less creative?
Unfortunately - it's common on HN - I've seen all sorts of hobby-horses tied to creativity. Everything from the economic system (capitalism vs "socialism"[1]), to "freedom", wealth and race (or something close to it as one can make on HN without being flagged - which is close indeed).
In all these scenarios, the commenter's favored group are creative and predestined for success because of their affinity, and the others should fail and therefore will fail and the author will resort to all forms of justifications, despite not having any evidence (or while ignoring the existence of contrary evidence)
1. Once, I couldn't get an answer why DJI was kicking all American drone companies asses (back,when they existed in comparable form), despite all the alleged inherent creativity from living in a country overflowing with freedom.
Having spent a good number of years in these companies, I can say that there is no shortage of creative and skillful employees there. What kills innovation is the lack of real ownership and the feedback-driven politics. Employees with great ideas and ability to execute can't go ask for funding and work on their ideas - those ideas would be either stolen (politics), or priorities would change (no real ownership), they would still have a manager to answer to, and they'd still have to be careful to not cross anyone's feelings for the inevitable complaint will wreck their careers (no real competition). So everyone plays it safe, wears a fake smile and cheers the team even when that team is surely going to sink the ship. The bigger a corporarion gets, the more pronounced this risk-averse safetism becomes. The terminal stage of such corporations is a gov-like bureaucracy where safety is the idol and risk is a slur word.
You really think Google have a failure to innovate? Obviously search was their first, core innovation, but over the years they’ve delivered so many other successful products (many acquired, true, but even those they’ve evolved really well). Off the top of my head:
- Maps
- AdWords
- Gmail
- YouTube
- Analytics
- Translate
- Calendar
- Sheets
- Docs
- Drive
- Meet
- Slides
- Drawings
- Forms
- Trends
- Alerts
I’m sure I’m missing a bunch too. IMO they’ve pushed the envelope in a massive number of spaces, and continue to do so. That’s an incredible number of insanely successful software products for a 23 year old company.
Edit: I was thinking web/mobile apps there, missing obvious big ones like Android, Chrome, self-driving cars, a bajillion Google Cloud products, Chromebooks, Go, Dart/Flutter, AMP, Bazel, Protobuf, gRPC, Quik/HTTP2/3-related innovations, etc.
Many of these started as acquisitions, but they massively transformed them. For example, they acquired some products in the mapping space, but none of them remotely resembled what Google Maps have become.
Taking some starting point, hugely evolving it/transforming it, and growing it into a dominant market leading product, I personally consider that innovation.
I agree with you on how much innovation occurred at google, but I think the timeline of when these innovations occurred vs when leetcode interviews really took hold is what the commenters are addressing. In the old days they had crazy hiring processes where riding a unicycle helped your case. This ushered in a group of very brilliant creatives that innovated. Now, most hires are required to memorize data structures and algorithms- things that a lot of the smart creative people loathe to do, and as a result the pace of innovation has slowed to almost a halt.
My personal opinion is that there were too many chefs in the kitchen and google is now trying to optimize the “creative leader” to “can pump out rote code” ratio.
Google Maps, the web application, was not an acquisition. Google bought two companies that made desktop mapping tools (One of them lived on as Google Earth Pro and was a commercial product they kept selling for years) and used that as a basis for developing the web based Google Maps.
Also based on everything I can find, Gmail started as an in house project. reMail was acquired 5+ years after Gmail launched.
Aren't they also mostly just web2.0 versions of Microsoft Office products? Like Docs and Sheets are just Word and Excel right? I feel like I'm missing something. Certainly putting them in browser plus some cloud storage is innovative, but doesn't it seem kind of obvious?
The other big thing is that Zuckerberg is really good at seeing the business value in other people's ideas. He chases financial value by copying what other companies do.
Instead companies like Amazon and Apple are obsessed with generating real world value for the customers and let financial value follow.
Without a doubt Zuck sets the business culture, people that think like him and solve the problems he sees get rewarded. This has a massive trickle down component to the entire org.
> Instead companies like Amazon and Apple are obsessed with generating real world value for the customers and let financial value follow.
While I would argue that even Apple is mostly about execution not innovation (many of Apple innovations had been around for a while, but apple took execution to a different level), but Amazon?! What are the big Amazon innovations? The one click checkout? Having pushed selling books online (arguably failing to become profitable for a long time)? I would say amazon is of all these companies the example which shows that you don't need to innovate anything as long as you have enough finances to push massively into any market you want.
Time-sharing was the prominent model of computing in the 1970s, it's really not that different from cloud computing. Seems like a stretch to say that folks at Amazon invented it. I do take your point though, Amazon was the first to market with what we now view as "modern" cloud infrastructure.
While amazon may not be as innovative as other FAANG companies, they do have execution down to a science. Supply chain, AWS, etc. all work WAY better than competitors in the market
That is a hopeful idea but I feel it might be our naivety that propagates it.
To my understanding Google's interviews have gotten easier and more arbitrary perhaps over time. The competition is high and thus there is the expectation of having studying to pass. Google is famous in the end.
Consider there are the following cases (kind of complete):
a) New employees hired are not brilliant
b) Google doesn't want to hire brilliant employees thus the leetcode interview
c) Leetcode has little effect on "they are brilliant" check, so a different trait is selected and indirectly brilliance distribution inside of Google is different but not the main one affected
d) Even if Google was full of brilliant people it wouldn't matter; they have no power over what happens by design (or inadvertently)
Google is huge. The main job of everyone there, I claim, is to not screw up the money maker: ad auctioning. After that they can do as they wish. The job of every manager there is to make sure that doesn't happen. Innovation is next. And they do innovate but not on things that make enough money to sustain. There is no direction thus.
As evidence I provide the following. They have been throwing ideas at the wall for years trying to see what else sticks even as close to the ad auctioning and data collection scheme and they kill what doesn't appear to. Thus, their reputation of killing things.
I have a few friends at Google, all of whom got in many years ago. They all claim the interview has gotten a lot harder, and that there's no way in hell they'd pass the current level of scrutiny.
Although FWIW, they also say they also would not have passed the interview for the time period they got hired if they had to do it again a second time.
I have no opinion on the quality of any particular employees, but the thing about Google interviews is that they seem to optimize for a certain kind of personality. Lots of people are "bookish" but Google seems to somehow further optimize for people with high IQ but low emotional intelligence. That lack of empathy is further reinforced by the needs of the advertising business, which is actually Google's primary business.
a) Meta can be accused of so many things but failing to build new, ambitious products is not one of them. They are literally pivoting the company to AR/VR and betting the future of the company on it. Quest2 is a huge success and their demoed R&D work on new headsets looks genuinely exciting and innovative.
b) You are clearly ignorant of how product management works in a large company. Bookish engineers which you weirdly use as a slur are not solely responsible for inventing and releasing new products. And so if they hired people who had less understanding of how to write scalable code you wouldn't suddenly get better products.
Small caveat there… they didn’t build AR/VR … they bought it then poured tons of money onto the smaller and more innovative company to get growth before trying to put a bullet though it’s customer base by forcing Facebook accounts in order to get greater synergies with the rest of Facebook only to backflip because the gaming market hates it so much it was hurting sales.
I disagree and, in fact, there's a whole website dedicated to its failed innovations[0].
Google's problem is not a technical one. It's a social and cultural one. Google is way ahead of the pack in terms of innovation but it's stuck in the past (and by past I mean present). Remember Google glasses? Remember how that turned into the Glasshole meme? (And apparently it's coming back[1].)
The rate of innovation is greater than people's willingness to adopt them. And also let's not forget that Google has a very specific modus operandi.
Google in particular is great at tech but grew from a very simple product (simple as in product surface area - a search box – the supporting infra is a different story). Their biggest issue is inability to deliver complex products that are coherent (1) with each other and (2) with itself, over time. I think mainly this is because they're not respecting the user (ironic, since that's officially their core tenet). Why? Because they are evil? No, mostly because they don't understand them. Employees live in a different world, demographically, socially, economically, everything. They've also never had a strong culture of truly understanding and embracing their user base. So they decide everything upfront without validating it externally first, and are genuinely surprised when they miss the mark. "That's so weird, all the execs & directors thought this was a good idea. Well anyway, I got my promo, better move on to the next team."
> Theory on facebook failure/Google failure to innovate:
> They have the highest leetcode standards out of FAANG. As does Snap. They hire people that basically memorized computer science algorithms.
These are mature Big Tech companies. While they do empower engineers to innovate, they still task Product Managers with managing product, just like every other big company.
The product direction isn’t very related to the LeetCode interviews given to dev candidates. I think you might be missing the context for the entire product side of a company and their interview processes and incentives.
I have commonly seen Senior engineers drive completely new product innovations at big tech. This hard split between product and engineering is commonly not real. Especially when the innovation requires understanding large scale systems deeply.
Same with Google, where Marissa Mayer would allegedly run A/B tests to find the best shade of blue
Pro-tip: that's not how you build a system. The best shade of blue is the least of your concerns
"Oh but Google/FB have actual designers, product people etc", yes but don't underestimate management's (those who got promoted, usually from the engineering ranks) ability to mismanage it
I saw an ad on LinkedIn a few months ago for Coinbase and applied out of curiosity. Was very intimidating as a British 'commoner' who went to a very average university to be being grilled on Leetcode/Hackerrank questions by fresh male-Asian Stanford/MIT/top-US-university graduates (not being stereotypical - this was true!). Didn't stand a chance really as I don't do well with algorithmic questions although the interviewers were friendly enough and I did somehow manage to stumble to the last stage of 1 recruiter interview and 3 coding interviews. One answer I was expected to answer the complexity of my code to which the answer was something involving logarithms which I genuinely don't have a clue. I consider myself to be a pretty good developer and constantly get praise/compliments from my colleagues.
Dodged a bullet really after all the news I've seen recently about staff having their applications revoked due to hiring stoppages.
> One answer I was expected to answer the complexity of my code to which the answer was something involving logarithms which I genuinely don't have a clue.
Being asked the logarithmic time complexity of some code you just wrote is literally one of the most common interview questions in Software Engineering, no? I'm assuming at least this is what you are referring to. All the well known interview help books in the industry ("Cracking the Coding Interview" etc etc) and websites like Leetcode spend a lot of time teaching it for a reason!
I've never had a software engineering interview where I wasn't asked to discuss something related to time complexity of a given bit of code, and given the regularity in which it will crop up its worth just spending the time to learn.
Yes, it was about the time complexity. I have a very primitive understanding of big-O notation from back in sixth form but I never did logarithms in maths (never did maths to a high enough level really).
My fault for not dedicating a bunch of time to read loads of books to pass an interview and instead spending my time making a living at a full time software engineering job.
The heuristic for this, FWIW, is to think about reducing the number of things you're messing with (such as a search set) by a factor (usually 2, not always) on every iteration. A binary search is a logarithmic algorithm because you chop off half of the (sorted) search set on every pass. That's all you really need to know about it to identify one (and thus realize you have a sublinear-time algorithm).
You don't need to read a ton of books or know any math at all really - the basic answers they are typically looking for are patterns you can spot with no knowledge of logarithms what so ever.
Spend a few days and you will just recognise, "oh hey I've a nested loop here, it's likely O(n2)!" That's really all it takes, and there are only 4-5 main classifications most interview code falls into, all of which can be spotted without knowing what a logarithm is. If I had interviewed you, I would not be impressed you hadn't understood this at such a basic level either, personally. The rest of us making a full time living at software engineering have to learn it to pass interviews at top companies too!
I'm comfortable enough with time complexity, and I like to think I'll sniff out a logarithmic-class solution when I've written one, but I don't think I've ever had a problem with a candidate bucketing a problem into sublinear/linear/superlinear.
I don't know if that is helpful but logarithms are quite an easy concept that helps with thinking about code complexity.
It's an inverse of exponent which translates to " grows very slowly with input size. 10 for a thousand, 20 for a million etc.". It helps to quickly ballpark feasability the same way exponential complexity usually means "doesn't scale above 30"
Log(n) is a
- number of digits required to write a number
- depth of a balanced tree of size n
- running time of any algorithm working on input of size n which does fixed cost operation to cut input in half (e.g binary search). If your code has cost function f(n) = 1 + f(n/2) then it has logarithmic performance.
I wonder how many of the people echoing this sentiment have tried and failed to clear the Leetcode bar at these top companies. It's hard to explain this level of consensus as anything other than motivated reasoning. As an earlier commenter stated, "I can believe that Leetcode is uncorrelated with creativity, but I have a hard time believing it's inversely correlated."
The anti-Leetcode leaning on this site has the same flavor to me as the anti-crypto leaning. If you missed out on the easiest money in history (as I did), you're going to have a strong psychological block on accepting that crypto can be interesting or positive, and you're going to be drawn to arguments saying it's doomed to disappear. I have to explicitly set aside my emotions about that topic in order to see it for what it is.
This may be true about what Google employees have become ...
But the changes at Facebook and Google have a clear cost-benefit equation. When a company is in an expanding market, improving it's offering to capture more of the market is very worthwhile. When a company is in a static market where they aren't likely to lose too much market share, the company has incentive to squeeze as much profits out of their market share. Usually that's by raising prices but in this instance it's by stuffing in as much ads and behavior-control shenanigans as possible.
You can say this is because of monopoly but similar dynamics happen with "oligopolies" - cellphone providers or pre-foreign competition US automakers.
I don’t think the two are connected. It’s more likely that it’s because feature decisions at these companies are made by thousands of product people who have their own performance objectives, most of which are misaligned with the wishes of users.
I think this is an underrated comment. It's hard to innovate/apply divergent thinking if you're so far from the user.
Adtech as an industry would also be a factor, since it's not only not aligned with value/user goals, but working actively against them. I'm saying that as someone who worked in the domain.
I interviewed at Amazon in 2008ish and it was very leetcode-ish in its interview process. I remember whiteboarding a garbage collector.
In any event, few companies have product direction set by engineering. There might be influence, but product tends to be a different set of employees than engineering.
Every company I've worked for has had some form of annual or bi-annual "hack week," where employees can form ad-hoc teams and work on something they think would be cool to exist. Sometimes it's a feature, sometimes it's a new product line, sometimes it's just something ridiculous (e.g. an eight foot wide NES controller).
Some of these projects did turn into shipped features, though I'm not sure I ever saw a new product line come out of it.
One of the places I worked had an internal incubator. An engineer or PM could pitch an idea, and get paid their salary to go work on it if accepted. That was still early stage when I left, so unclear if anything came of it.
The point though is that many companies do open avenues for engineers to innovate directly.
I don't think you're necessarily disagreeing with the poster you're replying to.
I've interviewed at some of the FAANGs and like you said, I didn't find the Amazon interview format any less leetcode-ish than Google or Facebook. However I found that the level of rigor and was noticeably less than what Google or Facebook seemed to expect. Maybe this has changed recently, since the last time I interviewed at these companies was in 2019.
Come on. A decent engineer can pass the programming portion of FAANG interviews in 30-40 hours of study and they take at most 40% of the interview time.
How much leetcoding do their PMs do? I understand your overall point, but I don't think having a high algorithmic/data structures whiteboarding bar for their engineers correlates to Facebook's weakening product abilities. Ditto for Google.
I think it does. As much as companies want to believe that product runs the product, an engineer that can create new things that influences the product is massively more valuable. Amazon commonly has engineers drive new innovations, product people nowhere to be seen
There are without a doubt a sea of creative people at Meta, the question is whether the organizational structure disincentivizes creative collaboration with engineering.
I've worked for Apple and Product Management absolutely runs the product.
Engineers of course have entry points into this process but ultimately they aren't responsible for the overall process of what, how and when things get build.
You act like you know better than all of these companies so maybe you can try building a multi-billion business with just engineers. Good luck with that.
TBH this not scientific at all but judging from my linkedin feed the worst 'product people' I've worked with have ended up at facebook/meta, amazon and google -- apple and a bunch of smaller companies ended up with the best
it's a mixed bag when it comes to engineers though, most of my highest performing coworkers are pretty evenly distributed between the batch with the exception of amazon which seems like a catch-all
Unlikely: developers do not decide what features to implement, at least in companies of that size, so whether they interviewed with leetcode or not shouldn't matter
Isn't snap known as being one of the most innovative social media platforms? e.g. creating stories before anyone else, bitmojis being popular, innovative filters, etc. I don't buy this tenuous connection between leetcode and innovation lol
> Apple, Microsoft, and Amazon do not have the same issues with creating new products.
Microsoft? Really? Is Azure more innovative than Cloud? Where is MSFT innovating? Acquiring Xbox and Minecraft?
And for that matter, where is Amazon innovating outside of AWS (their primary profit engine)?
FB & Google innovate in Ad Tech - that's their core business.
Apple is just more in your face about innovation because their core business is convincing people they need to endlessly spend $2000 every year to get the latest and greatest gadgets.
Even Exxon is innovative in gas exploration... All decent companies are innovative in their core business. Some core businesses aren't as sexy as AirPods.
Maybe the problem is that these companies just need to stop trying to innovate? All successful social media platforms start out as fun places to be. Then they become feature complete, but the laws of SV say they have to continue to grow for eternity, so what do they do? They mess up the timeline, throw increasingly irrelevant content at everyone, add and sunset new features every month, try to copy other platforms, screw around with people brains (oh, I mean increase engagement). Couldn't they just try for profitability instead of infinite growth?
They keep the talent from doing their own thing, basically innovate. That's how you choke innovators. Give a fat paycheck, a sense of accomplishment, while bleeding them out. That's all.
I see where you're coming from but I disagree. your posts assumes that people who are good at leetcode (or more accurately, get good at leetcode by spending time on it) are not creative. but I'd like to point out that creativity isn't an intrinsic ability, but largely a skill. you get more creative the more you practice. innovation is not just about thinking differently for the sake of it, but more about the cross-pollination of ideas from different domains. that's what mathematicians do a lot of the time- see what unsolved puzzle can be re-shaped to be like a solved puzzle, and that's what a good computer scientist should be able to do with tech. i belive, at the same time, that leetcode is pretty fucking useless as a measure of ability and a way of learning new algos because. my opinion on the matter is that big tech like Google and fb suck because each manager has to justify their bullshit job by doing some bullshit change. there's no coherent vision anymore and the company has fragmented into a thousand teams, each run like a bad startup with infinite funding just until the boss gets a promotion or smth. and then the team does some other bullshit changes. apple has always had a very coherent product vision, and so this doesn't happen. i belive my conjecture is simpler and more plausible than saying that people who work hard aren't creative.
You can also take the life out of any willingness to innovate with enough compliance policies, controls, code standards, and bureaucracy. That's why large companies fossilize.
It sounds like you have an irrational grudge against Leetcode-style interviews.
> Theory on facebook failure/Google failure to innovate: They have the highest leetcode standards out of FAANG. As does Snap.
As does Tik Tok, perhaps even more so.
> lower level engineers hired ten years ago do rise to the top and run whole divisions
Engineers have little say into product decisions, even if they run whole divisions. Nobody is criticizing the engineering decisions Facebook and Google are making. Just their products.
> Current valley hiring practices filter out the mischiefs needed to take things from sustaining to innovative. It's why tech is coasting and deteriorating.
> Passionate weirdos like JWZ, Stallman, TBL, Carmack, Cutler, Torvalds, Woz, Pike, Wall, Jobs and Bellard are required
> "The quality of a commercial software product is inversely proportional to the difficulty of the problem it solves, how long it's been around for and how successful it's been"
Essentially as a solution reaches stability and success, there's margins of profit that keep a staff of engineers and managers employed to work on the system. As a result they mutate the system - drifting it away from being a good solution to real problems to an unfocused solution to theoretical problems. It's a core organizational feature how we don't treat software as infrastructure where putting something in maintenance mode is the majority of the lifecycle of a successful product (for instance, whatever building you're in right now).
So as the generations of engineers cycle in and out you get this discordance between the two: earnest new people getting paid very well to change a product that usually has no pressing need for change. They do it anyway and thus successful simple commercial software products decline in relevance and quality over time as an inadvertent result of mismanaged good intentions. Furthermore when a change is warranted, nobody in this model has the agency, authority, political capital, or willpower to do the proper reframing (think Apple Newton -> iPod -> iPhone) without excessive compromises.
All of the people who you listed are unquestionably talented.
But they are unsuited to working in very large companies which requires collaboration, patience and tolerance of other views. And pretty sure they know it too.
You can make the tired hiring practices are the problem argument but actually it's the nature of large companies themselves that self-select for these types of people.
The passionate grumps don't lack collaboration, patience and tolerance. Instead they have more commitment to the product and company then to power relationships and hierarchies
They may also be impossible to work with but that's independent (alright, it's likely correlated).
Anyway, the hard-nosed ones are closer in spirit to scientists who get mischaracterized as uncooperative inflexible foot-draggers because their advocacies and endorsements aren't a functional of social relationships and ideally can't be changed by them.
I think following the long tail will lead you to what I witnessed with my kids in grade school; mischief is now akin to a disease - it's treated, counselled, made an example of, and sometimes medicated. I had to work hard with my kids to overcome this 'vilification' of anything against the norm.
> They have the highest leetcode standards out of FAANG. As does Snap. They hire people that basically memorized computer science algorithms. And so their employees are bookish, not innovative people. Apple, Microsoft, and Amazon do not have the same issues with creating new products.
I call bs on this. All of these companies have pretty similar interview processes.
> They hire people that basically memorized computer science algorithms. And so their employees are bookish, not innovative people
That is as far from truth as it can get. The problem in companies that fail to innovate almost always has the same problem: management. How companies are run has a significant impact on how much they can innovate. Even if you are willing to take massive bets as the management team, if the bets aren’t good enough, it’s going to be hard to innovate.
Apple has a top down innovation chain where goals and direction are decided by higher ups, while Amazon has bottom up innovation where people are encouraged to build. But the common theme in both is the ability to identify and nurture big ideas by the management.
> They have the highest leetcode standards out of FAANG. As does Snap. They hire people that basically memorized computer science algorithms.
I find this naive, as if coders have a say on the product. The jobs in big companies tend to be very boring and actually simple since their core product is already there and all these people simply maintain and steer it based on the data.
IMHO the difference is simply in their situation. Google&Meta are absolute monopolies, nothing new had to come out of them as they were making more and more money as they optimize the same old product. They suffered the Galapagos syndrome, similar tp the Japanese smartphone market.
Apple is not in that spot, they need to aggressively innovate as the competition is strong.
Regardless of the limited impact, its hard to argue that there isnt a distinct difference in innovation/releasing good products between Google/FB and Amazon/MSFT/Apple. The latter simply innovate more and create more things that people want. So the question is why
I tried to answer why :) Unlike FB/Google, AMZN/MSFT/AAPL does have strong competition, therefore they can't simply milk the current offerings. They have to innovate.
> their employees are bookish, not innovative people
This is false. However, there is probably a misalignment between what the employees want to innovate on, and what Instagram users want.
My guess is that deep down it’s hard for a talented adult to truly give a fuck about Instagram.
So they do what they have to to keep themselves feeling like their work is meaningful. And that diverges from what consumers of tween influencer / fake-moment-creators think they want.
Zuckerberg is not an innovator we all know where Facebook came from. He's a do-er and copier and he will lose the AR Glasses race as it requires someone innovative behind that ship.
He keeps throwing all this money at VR which has been around since the 90s. The general public doesnt want it.. want something strapped to their face/isolating themselves from real life... it is not the next iPhone!
Sort of. There is not a direct correlation between leetcode, engineering skill, and productivity.
I also know some boneheads who made it into google somehow. I think hiring is more of a crapshoot these days than it was in the early 2000s when the tech industry was not cool, was fairly homogenous and full of nerds.
I think you mention the real reason as a footnote. Ads are SO lucrative that it isn't worth it to be innovative or take huge risks. They do a little here and there but why invest $10B in an idea that is unlikely to have higher margins or larger volume.
Also it's generally hard to innovate in large teams because you need to convince every one. Personally I gave up innovation at work consciously because of this. Most innovation comes from creative people dictating.
High G factor correlates with high creativity. I see a direct correlation to coming up with creative solutions and being able to solve a battery of tests.
A lot of the leetcode hate mostly exists because it's hard. And a lot of coding interview problems actually require a lot of creativity to solve.
And imagine thinking that facebook isn't innovating. Facebook is leading the future of virtual reality, and leading AI breakthroughs. What do you think facebook marketplace usage is like? Just because you don't like getting tested doesn't mean people who test well can't innovate. No matter how much you trick yourself into thinking that Facebook/Meta won't stand the test of time, it has and it will.
Theres a difference between creating beautiful creative code and creating an innovative product that people want to use. Its a different type of intelligence, and why you dont see Math professors starting ecommerce/product businesses.
Apple was famous for taking forever to design “their copy” of popular thing.
Then Facebook and the rest became staples of iDevices, Apple focused on stupid slim design rather than pushing the product envelope, and risk their yearly mobile cash cow.
Their EarPods were not exactly the first BT earbuds, HomePod was well after Alexa. Subscription video and games?
For the most part they are doing the same with hardware; changing up housing but emphasizing their own chips.
MS aped Apple laptop quality with Surface, bombed mightily with Windows Phone, is doing fine with Azure; again all products that came as a response to others.
We’re beyond the bootstrap phase of technology. Something truly innovative to the user is still in some lab.
My money is on bio-tech mutant; custom drugs and high res simulation to embed an experience so real you think it happened. I dunno something that’s focused on more than b2b apps/ads deals and phone update cycles.
I mean there is no point to any of this. There’s no higher calling for people. One generation being addicted to computers as we know them is not necessarily a forever trend for the species. It’s just math. shrug
They have the highest leetcode standards out of FAANG. As does Snap. They hire people that basically memorized computer science algorithms. And so their employees are bookish, not innovative people. Apple, Microsoft, and Amazon do not have the same issues with creating new products.
Keith Rabois (top VC), mentioned that FB has always prioritized optimization over new products because of the Ad business. That also could explain their inability to do anything but copy other business ideas.
The opposing point of view would be that Ads on FB/Google are so profitable that working on anything else pales in comparision (in terms of profit generation). And so they dont innovate, but iterate and optimize.
My vote is that excessive leetcode testing selects for employees with no creativity. And while these companies are massive, lower level engineers hired ten years ago do rise to the top and run whole divisions
EDIT:
Counterpoint on how good early Google employees were:
the above maxim does not apply because that was before leetcode became huge, people passing those interviews were actually brilliant, they didnt spend six months studying in a basement