Intel (you seem impressed with them) are right in the heart.
Cisco, same.
Google seems to be unimpressive to you, but they are producing some pretty solid (and new) technology at a pretty high rate.
There are plenty of mediocre folks here as well, but there are also plenty of sharp ones.
(My background, btw, was DoE and NSF funded research. I came to work on California startups when things got much more interesting technically in the private sector)
Mine was US national security, computing, applied math, and associated research around DC. When I saw commercial technology, I upchucked. As Betty Davis said, "What a dump."
> much more interesting technically in the private sector
Hope so. But I don't have much hope, as I explained, due to the backgrounds of the VC partners and CEOs. In blunt terms, around DC, I could always get a competent technical review of my technical work; in the commercial world, I've never been able to.
Another explanation for the sick-o climate of innovation in the US commercial sector is that the decision makers, i.e., VC partners and CEOs, picked some "high payoff" projects, tried them, saw the projects fail, and gave up. So, Silicon Valley has adopted some rules: Regard anything beyond middle school level material as blue sky, academic game playing, nonsense and give up.
Of course the problem was that the VC partners and CEOs didn't give competent technical evaluations of the technical parts of the projects, and that because they were incompetent technically, and, thus, they picked poor projects. Simple as that. Again, those decision makers just did not know how to evaluate technical projects; plenty of people in the US do know how do such evaluations well.
So, now the norm is, don't evaluate technical projects. Just don't do it. Instead, evaluate projects much like an accountant would, that is, based on assets. The main difference is that accountants count in dollars, and the Silicon Valley decision makers are also willing to count data from ComScore, Google Analytics, etc. about traction. The crucial, core technical internals, if any, they don't care and don't look.
Intel? I picked them because their work at 14- nm and their path down from 1 micron has been so easy to notice. Maybe much of the credit goes to their vendors for, say, UV sources, optics, photosensitive materials, positioning systems, etc.
I used to be impressed with Cisco, but now I wonder.
Google? For their search engine, they have written some quite nice code. I don't see much serious innovation, e.g., like a drag-free satellite, the P&W engines for the SR-71, the Bertsekas polynomial algorithm for an important assignment problem, what I mentioned that Breiman did, but a lot of really nice code.
For their self-driving cars: My guess is that mostly they are in effect using electronic tracks and that people will give up when the rates of accidents and/or traffic jams are too high. Then the hope will be to change the roads to make Google's efforts with self-driving cars practical, and my view is that the roads won't change and self-driving cars will go the way of turbine engine cars. At times, driving as it is done now by humans takes some actual intelligence, and I don't believe that anyone knows how to program that.
To me the self-driving cars look like mostly a publicity stunt, a head fake with a lot less to it than meets the eye.
For Silicon Valley's innovation, I would ask: What is new, correct, significant, useful, powerful, and valuable and well beyond a bright, self-educated high school student and, hopefully, well beyond essentially any efforts outside the US.
With that, an example was the SR-71, especially its engines. Why? The Soviets also had the MIG 25 that could go that high and fast but only for a few minutes before its ordinary turbojet engines overheated. Meanwhile the SR-71 kept going along, 80,000+ feet, Mach 3+, for another 1500 miles or so, no problem, never got shot down. Why? The turbojet engines converted to ram jet engines starting near Mach 2 or so and, thus, didn't overheat. I'm not sure the USSR ever figured that out. Nice.
In the commercial world, we get innovation drip, drip instead of wild, raging rivers like we should.
Actually take away what came from high quality work for US national security, and there would be not much left of the computer industry or Silicon Valley. No such thing is true for medical innovation.
Yes, US national security and progress in medicine (the guys in Congress who vote for the NIH budgets are old and want better medical care!) are the main reasons NSF and NIH research grants fund ballpark 60% of the budgets of Stanford, Harvard, Princeton, Johns Hopkins, Cal Tech, etc. Okay. Fine. Terrific.
But for the rest, there is still too much dependence on US national security. US national security is a great goal, but the commercial world also needs to take the results of the research and create wild, raging rivers of powerful, valuable innovation instead of the present drip, drip.
As it is, Silicon Valley is doing next to nothing that a determined effort in England, France, Germany, Italy, Poland, Russia, China, South Korea, etc. can't do. Bummer. Really big bummer. Big time bummer. Threat to the US economy bummer. Might put US workers in direct competition with the 17 cent an hour workers in China. Did I mention bummer?
So, how to get ahead? Really, there's only one way -- great, leading edge exploitations of great, original research. Sorry 'bout that. To do that, need the decision makers able to judge the research efforts, the research results, and promising exploitations and make the possibilities real. For that, the usual MBAs, LLBs, business devs, marketers, self-taught C++ coders, UI designers, unicorn CEOs, need not apply.
Or did we notice that Microsoft, China, and Russia were all able, apparently quite routinely, to duplicate Google search? So, where's the unique, world-class leadership, which the US very much needs? We don't have it and are becoming just one more of the pack of 10 or so countries that can do nearly anything the US does. Bummer. We need to get going. For that, we need project decision makers who know what the heck they are doing in innovation.
I feel like I'm trying to explain the importance of medical research and innovation to a country doctor in 1850, the importance of a drag-free satellite to an auto mechanic in 1910, the importance of quantum mechanics to a radio engineer in 1930, etc.
Okay, Intel is now at, what, 14, 10, 7 nm?
Intel can put how many thousand cores in a single processor?
There was the Human Genome Project and Craig Venter.
The US did TCP/IP, IP and BGP routers, and the Internet.
James Simons did a lot, he won't talk about, but that resulted in a lot of money.
Can look at what is going on in fusion at Princeton, Argonne, Fermi, and LLNL.
Can see what L. Breiman did in classification and regression trees and, then, random forests.
Of course the USAF did GPS, which is accurate to one inch, and now farm tractors can plow very accurately due to GPS.
The US has done well with carbon fiber materials.
But, IMHO, relatively well informed, is that the US has powerful, valuable innovation coming like drips of a leaky faucet when it should be raging like a massive, wild river.
Why the difference? The people managing the money don't have even as much as a weak little hollow hint of a tiny clue about how to do innovative projects successfully. As soon as they see something beyond what they knew in middle school, they roll their eyes, throw up their arms, scream about blue sky, intellectual fun, games, and self-abuse, etc. That's grotesque incompetence.
The main exception is the innovation for US national security. A second exception is medical innovation, especially funded by the NIH and the research hospitals.
Silicon Valley? It has history majors with LLBs and/or MBAs who are 99 44/100% non technical looking for routine software for yet another social, mobile, sharing, membership, consumer thingy, throwing their line into their backyard above ground pool hoping to hook another Facebook.
We're talking totally clueless. Hopeless. Their ability to identify, formulate, evaluate, and pursue powerful, valuable technical innovation? Zip, zilch, or zero. Their technical competence? Even worse. Did I mention hopeless?
Instead, just got to have people running projects and allocating funds who actually understand how to innovate. E.g., need an Andrew Viterbi, Gordon Moore, Craig Venter, James Simons.
For Bill Gates, Mark Zuckerberg, go back to Harvard, complete a good education, and then try to allocate some funds effectively. Page/Brin -- back to Stanford, guys.
I've seen a lot in innovation, and what I've seen in the decision makers in Silicon Valley is the lowest, lower than I would ever have suspected could exist. And their ROI? Actually, on average, poor.
Blunt fact: Could hold a convention in a commercial airplane wash room of all the Silicon Valley venture partners and CEOs qualified for a problem sponsor position at NSF. Sorry 'bout that.
We're talking business development people? Marketing guys? Hold on while I go for a big swig of the pink stuff, Pepto Bismol. Darn, too late -- upchucked, on the sheared 100% wool carpet. Those people believe that the crucial core of the innovation they are looking for is some C++ code on Linux done by self-taught hackers.
At least when the US DoE and NSF fund fusion research, they don't look for really big breakthroughs from 10 year old naughty boys out back playing with matches.
In the past there has been innovation, also, when there was money enough involved and people ready to do something: E.g., clocks. Why? Well, in 1800 a wooden ship good for open ocean sailing was big time expensive. But on a successful voyage, such a ship could make a lot of money. But, the ship needed to know its longitude -- latitude was much easier. For longitude, ..., right, need a really accurate clock. Newton knew that, too. So, at whatever effort, ships got some quite accurate clocks. Else the ships too often ended up on the rocks.
Innovation can be done; it has been done; here in the US now we should have a raging, wild river of innovation; instead, we have a drip, drip.
Net, the people running and funding the projects are technically incompetent. E.g., they just didn't study the right stuff, long enough, hard enough in school, and, gotta tell you, now, away from graduate school, no way will they reinvent that material.