Google is also a very good learning opportunity; there is lots of training available to employees, you will be working with very talented people, and will get to see how the planet-scale dogfood is made first hand.
Google's CI system is proprietary, but learning how a CI system should work is valuable. Google's code review system is proprietary, but learning how code review should work is valuable. Repeat for every proprietary system Google has.
When I left Google for a startup, all the tools I took for granted didn't exist. But because I knew how things could and should work, I was able to apply open source and commercially available tools to improve software development practices at my new employer.
That is not true. No matter the shit Google gets on HN it's a tremendous place to learn and grow. All the large scale systems you see now were first invented at Google and now are replicated at Facebook etc.. The code reviews alone will make you a better programmer and working with senior people will give you an appreciation for designing complex systems. It's a great place to work IMO.
Put another way, you'll learn much more about how to identify and work with talented people on difficult projects at Google than at a startup. At a startup you're worried about how you'll pack a bologna sandwich for lunch tomorrow, at Google you're a sous chef for a 5 star banquet hall in NYC.
I've done both extremes, and you're absolutely right, but the point still stands no matter how many cases we can find that might hint at the opposite :)
This used to be the case, but these days many fundamental Google tools are open.
Borg is Kubernetes, flume is externally available as Google Cloud Dataflow (it's basically Apache Beam), Tensorflow is open source, many of Google's internal java libraries and tools are available externally (Guava, Guice, Dagger), and absl (c++ common libraries) are also available externally.
More importantly, assuming you land on a team with strong software engineering culture (this is likely, but admittedly not guaranteed, even at Google), the general skills you pick up for software design, planning, testing, etc. are transferable anywhere.
That said, all of this applies mostly to software engineers. If you're applying for an SRE position, then I've heard the experience is less transferable, but it's not something I could comment on first hand.
That's true about many of the specific tools you learn, but you also learn a lot of generic development skills that no one will take the time to teach you at a startup.
Google's in-house tools are generally stronger than the publicly available stuff across the board. Ex-Googlers are more often frustrated not by a lack of ability with those tools, but knowing what they can't do with them.
Also, Google's best tools are increasingly open-source.
Really, for the most part, the talent goes to the best places to work. That's subjective but the major factors for most people are mission (e.g. charity good, privacy violation bad), working environment and compensation.
Amazon isn't really particularly good at any of these, so you'll tend to find more talent at places that do better. Facebook and Google both have better compensation and from what I've heard from friends (who've also had experience with Amazon), both also have a better working environment.