"Palo Alto-based Stanford University ... and MIT have historically been the primary talent pools for the tech industry"
This is far from true. If you look at the total workforce in tech, the majority is by far from much larger, public schools, like Berkeley, Michigan, Illinois, Washington, Texas, Wisconsin, etc., etc. MIT and Stanford have had amazing great people, but they are far from the "primary" source of talent.
Yep. A lot of the Big Ten schools along with universities like Washington, Texas, Florida, etc. have been the drivers in overall production of engineers.
I think what they probably meant talent pools for some companies in the tech industry, or something similar.
I came to the comments say the same thing. I’ve hired and worked with many engineers and the overwhelming majority did not attend MIT or Stanford. Some of the best didn’t attend college at all!
I filter based on experience. Previous jobs, roles, projects. Degree is a weak predictor and I consider it mostly for candidates with less than 3 years of experience.
I guess my question transitions over to: If a candidate has less than 3 years experience and no degree, where does that fall on your radar? Genuinely curious as it falls into the "need experience to get experience" job meme many people find initially.
Speaking from my perspective as a prolific FAANG interviewer: Listing whatever experience you can (high-school jobs, sites created for friends, college jobs) alongside interesting personal projects is a huge leg up for getting through recruiting. If you're early in your career, find a few small-time jobs somewhere in tech that'll hire you, and learn as much as you can as fast as you can. MSPs are desperate for people with a little coding background. They'll chew you up and spit you out with long hours and terrible pay, but they're a great place to cut your teeth on getting things done.
Once a candidate gets to me as an interviewer, I'm mostly looking at a resume for what topics I should ask about, both things that you're comfortable in, and things that you're not. If you have personal projects, I'll ask pertinent questions about them, and often I've learned some very interesting insights from the personal projects of candidates.
In my time interviewing, I've hired people without degrees as senior engineers and I've rejected people with doctorates in computer science as junior support engineers. Turns out, degrees are a mediocre predictor at best of ability to do a job, and practical experience is 99% of the time more impactful than a piece of paper.
> Speaking from my perspective as a prolific FAANG interviewer
That's interesting, and somewhat rankling - as both times I interviewed at the G in FAANG, not a single interviewer gave a nickel about any of my experience, or projects I had worked on.
It's such a joke that the keys to a golden line on your resume and a giant salary are gated by the luck of which 7 people get picked to grill you that day, how busy they currently are, and what their pet question is.
If a candidate has less than 3 years of experience AND he has nothing to show AND he has no degree... he will not pass the filter.
Moreover, I'm not really interested in "years of experience". Previous titles don't really matter, too.
What does matter for junior and middle developers: personal projects, specific products or features developed by them or their team, accounts on Upwork, GitHub or StackOverflow, recommendations from other people, a personal website or a blog.
We are talking about the initial screening filter, so _anything_ proving you are a real developer is fine.
All the DC area computer science programs feed right into the NSA and defense.
Which is funny because Silicon Valley companies won't think candidates from those schools are that interesting, while simultaneously thinking that the intelligence and defense has better programmers. People are people.
Very odd statement indeed. I have never heard anything even remotely corroborating this. In fact (highly anecdotally, take it as entertainment at best), I have read that public opinion holds NSA's technical capabilities in high esteem but in reality, the talent is vastly below that of the top consumer tech companies because most tech employees would never contribute to the type of work the government does.
-- The rather enormous intelligence budget that allows all kinds of creativity
The above indicates that there must be some corners of these agencies that are extremely technically skilled, and able to do real stuff with breathtaking ambition. Given the budget, these "corners" might be pretty big, and attract some excellent talent.
Those that know, aren't talking, and those talking (ahem!), don't know.
I can assure you that there are programs with insane skill requirements, many of which have fed the executives and senior Security and IT staff at FAANG. They're not likely being overt about their previous jobs, choosing to instead make them sound less sexy so to speak, but they're 100% there. I would also note that you're over estimating 2nd party contributions to all the above.
You're right on the money for math though, crazy smart mathematicians.
How does the fed pay these programmers more than the normal federal compensation schedule? Is the compensation uncapped if they work through a contractor? If they dont, why do these programmers choose to work there?
If you got a Ph.D. in some field and wanted to do R&D in it without trying to go tenure-track, then the national labs, and occasionally the contractors, used to be a decent compromise. Bell Labs was gone, Google wasn't yet running its moonshot factory, and Softbank had yet to open its pocketbook. Depending where in the country you want to be it still might make a little sense.
But if you're a programmer without a scientific speciality, then there's less reason.
The contractors have some freedom to pay as they see fit -- they pay more than the national labs in my experience -- but they still have to justify their rates to the government during contract negotiations, and the government can push back. The pay is "ok", but has been falling increasingly behind "Tech". So now there's brain drain, particularly among programmers.
I have limited first-hand experience, but yes, contractors can be paid independently of government pay scales as far as I know, and far more than you'd think (in terms of breadth of jobs) is done by contractors. I remember reading that the people who do security clearance investigations largely are, or were, contractors.
My impression is that if you do work directly for federal or state government, you also are likely to be alongside a substantial number of contractors.
That was the common transition where I worked. Do gov stint, get into a great office, establish a name for yourself and get the right access. Then transition to contract, bounce between them as they contracts are merged / transfered etc. but still make multiples greater than as a Civilian. Some billets were GS 9-13. Only difference was education background, same work same tasks. Same contract could be making 135k base plus profit share or guaranteed bonuses. The sky was the limit for some specific people (though not everyone got these offers, just like most won't get a FAANG job).
Multiple programs were built to fix the compensation for military members whom were paid even less than civilian counterparts in the same offices. It's why the churn rate is so high. Even then, the majority of cleared jobs still aren't commensurate with external salaries.
Can't do anything but agree. The intelligence budget, as you said, is enormous so there will certainly be things that come out of it that would be shaky business decisions to even attempt in the private sector.
For one, GPS was created by the DoD and I am still blown away that it is so good, accessible, and free for consumer use.
You should have ended your post with "<mike drop>".
Pretty hard to argue with that list.
A friend of mine (who went to NMSU) worked for NASA. He wrote the code that determined what angle the solar panels need to be at to get the right amount of sun to power the ISS. Crazy number of edge cases, and people die if you have a bug.
Was pretty impressive to hear the things he had to do and the processes they had for minimizing the risk of someone dying. I'd put that on par with many things, too (although Stuxnet has its own place for genuflection).
> most tech employees would never contribute to the type of work the government does
I don't even think it's ideological. The reason the defense industry lacks top tier talent is because it's quite literally the most stifling and frustrating place to work possible. You're mired in endless bureaucracy, incompetent management, arcane rules and regulations surrounding secrecy, and some of the world's least flexible working arrangements. It ends up being a great place for people who want to punch a clock and get a reliable paycheck, but a deadend for anyone trying to make their mark.
I thought the dominant public opinion is government agencies are staffed with mediocre engineers. Not even because of any moral concerns, just because of bureaucracy and low pay.
Agreed, but we still think of the CIA and NSA on a technical level as being capable of unparalleled surveillance, crypto cracking, and other general hackery and engineering to maintain its power. Perhaps that's part of the propaganda. With that said, I imagine much of that is possible with backdoors to mass consumer data like Facebook which is actually collected via ingenious (but unethical) talent in Silicon Valley.
Silicon Valley really loves some formal education it seems. Both of my Si Valley interviews, where they were willing to fly me out and accommodate me, involved an older gentleman having a sit-down with me about university, and how I should have gone.
Allow me to rephrase, people believe the NSA has superior technology with mystical capabilities. Who do you think that develops that stuff, where do those people come from?
I think the more accurate perception is they have relatively unlimited budget and interest in attacking computer security, along with special powers like doing basically secret illegal shit that make them difficult to beat in security matters.
Not that any specific set of their employees are super geniuses.
While SV still has to run a profitable business while trying to care about security.
Laughable, because a large number of tech workers in SV are from India, and a large percentage get their undergrad degrees in India, most come here for Masters degrees, but a majority go to engineering schools all over the US.
This. Anecdotally one Indian Institute of Technology (IIT) senior Professor told me that normally the best achievers of their undergraduate students go to US for work or further studies after graduation, the middle achievers work in India and the bottom achievers mostly continue their study at IIT with India govt's sponsorship. According to him even the bottom achievers are actually still good students since they graduated from IIT.
Since there are currently 23 IITs, the numbers that go to US (and SV) every year can be very significant. The long term effect of this phenomenon can be observed from the increasing number of US (and SV) tech companies CEOs are from India education background.
> the majority is by far from much larger, public schools, like Berkeley, Michigan, Illinois, Washington, Texas, Wisconsin, etc.
These schools are very good btw. Berkeley is tied with Stanford for example, UI #5, Washington #6, UT is #10.
Conversely Harvard is ranked #15, but that's the things about rankings: they're often used to prune out people who don't "know someone", and when the metrics do not match someone's perceived "elite" status, they are ignored.
All of the schools you mention are attended by elites and commoners, produce talented people and untalented people, jerks and kind folks, and all the permutations of that.
I just wish we'd focus more on competent and kind, and less on connections or wealth - companies that do this will prosper IMHO, and companies that rant about "meritocracy" will wither as they drive away half the population.
Women are just as capable of coding well - creating a toxic culture is reducing your talent pool by half, something people obsessed with "facts and logic" should think more about. Insisting on "elite" schools when the good ones are public will also harm you long term if you're doing more cutting edge stuff.
These are graduate school enrollment numbers. Without breaking down CS/EE versus most other engineering disciplines, are MS/PhDs the "primary" talent source for the tech industry?
For comparison, GATech claims 61% of their students get engineering degrees and an enrollment of 14,142, thus ~8,626 undergraduate engineering students.
I've always wondered why mega billionaires from this era such as Bill Gates, Jeff Bezos, Warren Buffet, Mark Zuckerberg etc. haven't built universities like Stanford, Duke, Cornell etc. - all started by monopolists from the previous era. I guess Eric Schmidt would be the first one
A big part of that was that there were simply far less universities in the mid-late 1800s than there are now, so there was a demand - this is particularly true for the West Coast. Likewise, both Cornell and MIT, though private and philanthropist-funded, were also land grant universities, and part of a deliberate policy (starting with Lincoln) to increase the number of US universities, along with providing a more "practical" baccalaureate. In general the post-Civil War era in the US was a period of significant educational reform and expansion at all levels.
don't you think the demand still exists today ? the economy seems to be in ever increasing need of a highly skilled workforce, in addition to reskilling existing ones.
You could argue that we could use better schools or schools that have more of a focus on subjects needed by a modern workforce. But there's no real shortage of colleges and universities. Yes, existing universities come with lots of legacy entanglements (which can be good as well as bad) but for the most part you're probably better off investing in those schools than tearing down and starting from scratch on a prairie somewhere.
No, because school prestige is a positional good. This is like saying you can address the shortage of Olympic gold medals by just manufacturing tons of them.
It’s not obvious this smallness serves any functional purpose. Two (U of T, Mcgill) Canadian schools have extremely good international reputations and serve more than the entire Ivy League. U of T alone educates more than the whole Ivy League.
But the point of school prestige, at least in tech, is to assess skill. If the available pool of raw talent is larger than can be funneled through prestigious schools, then there is a need for more schools.
But the point of Olympic medals is to assess skill. If the available pool of talented athletes is larger than the number of gold medals, then there is a need for more gold medals.
I think the point people are missing is that you don't need more gold medals. Rather, the people "hiring" athletes need to recognize that for many purposes the difference between gold and silver and bronze and first and second runner up to bronze is trivial and possibly arbitrary, especially in situations where outcomes are not nearly as easily quantified as in an athletic competition.
People don't go to these universities for prestige's sake, but for what prestige gets them. It's not really about the gold medal (for the vast majority), it's about being an elite athlete (i.e., being generally healthy and physically capable of impressive feats) and having the security to practice the discipline you love. Knowing that, you would build a network of schools that meet the full need of potential elite athletes, rather than rely on a handful of clubs and coaches. Which is exactly what we (and China, and the Soviet Union) did.
"That kind of school" is (call it) a Top 10 school in a given field. And there is always going to be a shortage of slots at Top 10 schools whether you have 1,000 schools or 10,000. What potentially makes sense is investments in relevant subject areas in schools that aren't in the Top 10 but are in the Top 100 so they're functionally less distinguishable--except in prestige perhaps--though arguably that's already the case. In the US, cost aside, a qualified applicant can already get into a "good" school.
There may be a problem once you look at the social and economic structures that encourage students to apply to these schools.
For schools, like Harvard and Yale, isn't a large part of the value they provide to students derived from the school's network of alumnus and the prestige the school's name holds? If so, don't these schools derive a large part of the value they offer from their exclusivity?
> the economy seems to be in ever increasing need of a highly skilled workforce
This is a claim I've seen repeated often but have never seen any real evidence to back up. It seems more like "we need skilled workers but don't want to pay them skilled wages." Regardless the problem (in the US) is not an absence of universities like it was in the 19th century - financial barriers to bachelors degrees are significant but that is not a symptom of low supply and high demand.
So no, I don't think the demand exists (in the US) today, at all.
There may be a subtle distinction between a demand for university-trained workers and a demand for universities. The latter produce the former. However, more universities may not be the as efficient a way to train more people as expanding existing universities.
It depends what percentage of a college education you ascribe to "vocational skills" and what amount is "signalling".
The Morrill Land-Grant Act was, in part, a response for the need for new skills of an industrialized economy.
However, if you believe schools are mostly about credentials and signalling, online learning institutes often have a large accreditation hurdle. E.g., it's generally preferred to have an MIT degree rather than a list of MIT MOOCs taken because there is a certain amount of elite signalling that comes with the degree even if the skills learned are the exact same
> E.g., it's generally preferred to have an MIT degree rather than a list of MIT MOOCs taken because there is a certain amount of elite signalling that comes with the degree even if the skills learned are the exact same
From my perspective, it is also much preferable to have an actual degree from Plymouth State University, or even an online degree from Southern New Hampshire University, than a MOOC "degree" from MIT, and I strongly disagree that the "skills are the same" in practice, evne if that might be true in principle. It's not just about elite signaling, there are huge differences in rigor, workload, and personal responsibility with an actual university education versus a MOOC.
MOOCs are fine to supplement a bachelor's education, but if I saw a resume that only had edX/etc, they'd better have good work experience or a great project portfolio.
You are right and thank you for clarifying. Skills may only be equivalent “in theory”.
In my experience, this even holds true for the same degree within the same institution across different timelines. There’s other pressures (financial, political, etc.) that can drive rigor and curriculum to change. That is at least part of why accreditation is important; it’s supposed to provide some third part vetting that the product is meeting minimum standards. (Again, in theory).
I think it will get much more interesting if online courses can fix the quality/accreditation disparity. With colleges no longer having a monopoly on education that’s still their strongest chip to play
>haven't built universities like Stanford, Duke, Cornell etc.
I assume the answer is that they generally find that it makes a lot more sense to give oodles of cash to their alma mater or some favored school to get your name on a building and/or really beef up a department than to try to found a new school which would likely take decades to establish any sort of reputation.
You're underestimating the egos of billionaires and how they love winning compared to other billionaires. A name on a building is nice. A name on a university that's mentioned with Duke and MIT...now that's something you can brag about. The owners of NBA teams - all billionaires, most willing to lose money year-to-year in order to win and beat the other billionaires. The women's basketball league in Russia - teams owned by oligarchs who all lose money on the league but spend lavishly to beat each other.
>A name on a university that's mentioned with Duke and MIT...now that's something you can brag about.
That's the rub though, isn't it. I'm not sure even Bezos has money that would be needed to pull that off over, say, a 10 year horizon. MIT has almost a $20 billion endowment to give some idea of the sorts of money flows we're talking about here. To say nothing of the fact that they already have the faculty, physical plant, reputation, etc.
That’s why I think they should donate (ie buy) an existing university and transform it. I’d go for Pepperdine, right in Malibu, near the beach, perfect weather, great architecture.
But I actually agree. Why try to build a university from scratch even starting with an existing "tier 2" (whatever that means) school when you can get your name attached to a significant new initiative at an institution which already has a lot of prestige?
I'm not sure that's literally true. Given sufficient money:
- Desired faculty: We'll triple your salary and hire your spouse too. Oh, and give you unprecedented freedom.
- Desired students: Free tuition and board and a great campus environment.
At the least you could jump-start a pretty darned interesting educational institution at least within a certain focus. But I'm not sure the money is realistic, at least at a scaled undergraduate level.
I'm surprised Olin college hasn't come up in this discussion yet (http://www.olin.edu/). They're small, but essentially went with this model, at least for a while. They're more undergraduate-focused though, so I don't know if they'll ever garner the "reputation" of MIT and Stanford's impactful research programs.
Look at this way - if you're going to send your son/daughter to a university and you're trying to decide between MIT and Bezos University (all things else considered equal). Which one is more risky? It's a no brainer. Not to mention many of these tech "gods" notoriously scrap projects when they don't work out in a short term timeframe. There's just very little reason not to send your child to a prestigious university if you have the means.
I'm sure there is no shortage of universities which would rename themselves to Bezos U for enough money. It's not even a particularly radical notion. I don't know any examples of entire universities but Cornell's Business School renamed itself after a very large donation and MIT's new computing school is named after a very large donor.
Hasso Plattner, one of the founders of SAP founded the Hasso Plattner Institut in Potsdam, Germany in 1998. It’s not a full fledged university, but a part of the university of Potsdam. A lot of the alumni work for SAP.
True. But yes, it’s not an university. I also doesn’t deliver any kind of state recognized degree (IIRC), which can be a problem for visa applications.
I agree with you big-time on this. Many of these guys are concerned about "their legacy" and also philanthropy. Why not create a top university named after themselves that seeks to be at the forefront of technology/medicine/biotech (whatever they choose). Better yet, as some have suggested, why not donate enough to a university to rename it after themselves and take it over, adjust the curriculum to the desired goal, hire the experts in the fields to teach and study, and see results 5-10-15 years down the line.
As a nice bonus, why not turn it into a basketball powerhouse school too, something to help brand the school, like Duke or Stanford. Something MIT is missing.
In Turkey, the best private universities (that are also up there as best universities in the country) are still held by the corresponding company holdings, i.e. Koc and Sabanci.
By contrast, in Germany there are no universally notable private universities at all (save for some business schools).
Not to point out the obvious, but 2 of your examples never bothered to stay in undergrad university long enough to get a degree and they managed to succeed beyond all of their contemporaries.
Perhaps it says something about our current university system/paradigm that the current wealthy have chosen not to support this route. Or maybe it says more about the Gilded Age robber barons that they chose this route (in addition to the large inheritance fortunes they left).
Marc Andreessen's "Build Stuff" article[1] pointed at the lack of university capacity in the USA (I doubt this is accurate[2]), but leveling up the average university to be closer to the top tier universities would certainly help or at least justify the current cost of university.
Overall I liked the Build Stuff article, but the idea that there's a lack of capacity is incredibly silly. It's not difficult to get into a Cal State school, UC Merced or my alma mater for instance.
Ever hear of Trump University? Granted he isn't a "mega billionaire"...and turns out it wasn't a University at all, in fact it was shut down and resulted in lawsuits Trump settled for $25M.
This will fail not because it's a bad idea, but because the government will reject it.
Contrary to popular belief, established universities do a very effective job of preparing tech talent for government jobs. The issues, which Schmidt seems to miss, are twofold:
(1) the government, with few exceptions, is a TERRIBLE place to build effective tech skills. Slow, heavily outsourced, bureaucratic beyond belief, and low paid, there is very little creativity and variety in which to build unique technology, and no reward for the BS.
(2) Current universities already prepare tech workers well. In some cases, there are even joint CS-Public Policy degrees (e.g., MSCAPP at UofC).
Those engineers with policy expertise just don't want to work for the government. I know; I was one of them, and I could not for the life of me find a policy tech job that paid even HALF what tech companies would pay me.
Schmidt should focus on building "the department of innovative technology" and pay engineers appropriately.
I mean... "The Department of Innovative Technology" is kind of what the USDS is. It's ran by very competent people, and do short "tours of duty" with competent SV engineers instead of long-term, safe, bureaucratic jobs. Here's a podcast:
I think you're misinformed on the topic, but that's OK. I assumed it was a typical government agency until I got further information, and then I changed my mind. Will you change your mind?
It is not a lack of talent, it is a lack of good governance/administration that causes most software/IT failures for the government. USDS is actually pretty cool, and they seem to be doing really great work. But they're a small part of the system. In order to actually improve software/IT within the government we need to fix a lot of small problems.
Here are some more:
Just the universal waste within the way government treats software acquisition. If you ran a private company and you had small teams (each less than 10) going out and individually buying a software system, say half of them end up buying small-team licenses for Visual Studio as an example, you'd want to murder someone. They could save 20-40% by combining all purchases under a common contract.
Or instead of hiring an actual system administrator, the new guy gets tasked with managing in-house servers. Oh, but it's not one person for an office of 1k, it's every small team of 5-15 people dedicating one person to manage 1-5 servers. Servers that could, mostly, be shared across teams and so instead of 100 admins you could probably get by with 10-20 full-time and trained admins, along with a fraction of the servers and a fraction of the software cost.
The US government does this all the time because there is little-to-no communication involved. What's hilarious(ly sad) is that I've seen this where the purchases went through the same purchasing officer. A bunch of teams all buying expensive (because of scale) licenses when they could've combined them into one joint contract. Millions wasted a year doing crap like this even in smaller branches.
Fix the crap like that and then USDS's model can become effective and universal. Until then, there is so much waste abounding within the way US government does software/IT that it will always be a joke.
This sort of thing is exactly what USDS works against. However, it won't be solved until federal agencies hire on people with the skillset to effectively manage technology.
I'd tack onto the first point: obsolete technology everywhere. Worked in the public sector for a short while and the amount of outdated tech being used is astounding. Some of the solutions being discussed as "new" have been outdated for decades.
Having to deal with overwhelming bureaucracy in exchange for working on what everywhere else would be legacy systems really killed any desire to stay.
The most important aspect of any technology is that it works. In that regard, functioning "legacy" systems are ideal systems. The legacy bureaucracies are the problem.
Kind of like SpaceX popularizing rocket reuse, we should think of older code, languages, and frameworks as "battle tested", not "legacy". They should only be replaced if it's clear that essential maintenance is not possible, if a critical requirement isn't being met by the old system and will confidently be met by a new one, or if it can be demonstrated that a new system will save more money than improving the old one.
Absolutely... in the situation where the systems are actually legacy systems and not new installations or adoptions of obsolete technology. What I was trying to get at, and realize I expressed poorly, was that they are sometimes looking at implementing (today) things which are obsolete. That's what I meant by 'would be legacy anywhere else'.
This, I guess, would be a consequence of legacy bureaucracy, given that the people making the calls are not very aware of the new technologies and their advantages. A reluctance to adopt Python for example out of a distrust of FOSS comes to mind.
But we should refactor it into a functional programming language with microservices running in containers because... resume-driven development.
I do agree that many organizations accumulate so much cruft they have trouble getting out of their own way. But the flip side of rewriting things for reasons isn't necessarily better.
> Slow, heavily outsourced, bureaucratic beyond belief, and low paid, there is very little creativity and variety in which to build unique technology, and no reward for the BS.
Currently a public sector tech worker, although in Canada. We can recruit some great people as although the salaries are not FAANG, they are pretty good for people who don't live in cities with that.
Problem is, they don't give raises or if they do, the raises are tiny. I earned 10,000 more choosing a government job over other offers where I am. I have been in the role 11 months. Recruiters are consistently offering 10,000 more than what I make now. Guess how long the average person stays? Just over one year.
If I quit my current department and went to a different government, I could get that 10,000 dollar raise and if I quit after another 1-2 years there to another government department, I would probably hit near the maximum bracket for tech work in the government unless I took a lead position.
Government also doesn't recruit people out of school for tech unless they have a full slate of internships. You need at least one year of experience to join. So there is this narrow 2-3 year in your career window where government work can make sense.
So not only are you kind of forced out fairly early in your career, but they incentivize spreading your time among many different departments as that is the only time you can get a big salary jump.
Plus all the other things you say are true. Decisions take months for anything.
I wonder if you may be discounting the value of a pipeline that's presumably going to be substantially less expensive than going to a private university.
Perhaps one of the goals here is to flood the government tech sector with new personnel who have been indoctrinated to expect a certain culture, in such a way that the existing structure won't be able to resist changing. If you and your entire cohort have an n-year commitment, and all of you are saying, "No, this is too slow, don't outsource that we need to do it ourselves, that's not the right tech we need to use this," eventually something gives.
This is a very good point. A university to transition Business Analysts into junior data scientists or HR reps into HR analysts, train managers in tech fundamentals, and upskill SWEs in security or new tech would be SUPER valuable.
Its called the internet. This new tech even enables you to learn things as you need to, often for free, instead of packing in 4 years of stuff then trying to remember it later.
Classic modern HN comment. The hard working person with a history of success who is trying a risky ambitious project will fail, because they didn't consider the few things a commentor instantly spotted as problems.
Classic modern HN comment response. You act like these criticisms are somehow inherently negative...
Just because people love to point out the HN discussion where Drew Houston got scrutiny for Dropbox, doesn't mean that discussion isn't valuable. I'm sure as hell Drew learned a ton from the criticisms.
Maybe you're right. I felt the over-confident claim that it would fail to make it cross the line.
If we discourage people from trying 1/100 or even 1/1000 chances because it's overwhelmingly likely they fail, we miss out on the massive upside offered by these risks.
Yea, that's fair. I should - I do! - commend the effort.
Some good things here:
(1) It's good to try to solve problems.
(2) There is a shortage of tech talent in the country
(3) Eric Schmidt is very smart. Sure as heck smarter than me.
Still - from my admittedly-limited vantage point, I can name a four people with degrees in public policy and CS (or similar) that tried and failed to get good jobs in government, so they took $150k+ tech jobs in the private sector instead. So seeing someone saying "the problem with government tech is one of pipeline" is jarring to say the least.
Except in the world of education there's been a substantial amount of people without their own direct experience in the field who've put out claims that they can materially change things when they have been completely unable to do so time and time again.
I do see value for an institution described as such in this post, but much more on the research side of things with a secondary focus on should a new branch of the military be established to focus specifically on cyber issues, such an institution could become valuable in the same vain as something like west point.
You could flip it around and say that the commenter is the exact type of person that Schmidt needs to sell on this idea for it to be a success (people who would study tech and then go work for the government) and they aren't buying it.
Schmidt failed to consider the perspective of one of the key stakeholders in the area he's trying to change. His idea would get laughed out of the room if he wasn't already a billionaire.
"(1) the government, with few exceptions, is a TERRIBLE place to build effective tech skills. Slow, heavily outsourced, bureaucratic beyond belief, and low paid, there is very little creativity and variety in which to build unique technology, and no reward for the BS."
Strong claim; links or citations to back it up? (Not necessarily doubting you, but it sounds a bit like a right-wing talking point, and hence I ask)
You're right about pay problems, though. With public sector jobs, you have to do it for the calling, not the bottom-line cash (imho).
I think the null hypothesis is more useful here; assume that NOWHERE builds tech skills until otherwise demonstrated.
What you see: a bunch of tech startups founded by ex-Google, ex-FB, ex-Amazon employees.
What you never see: hot tech startups founded by ex-government bureaucrats
At very best/worst, you get beltway contractors who have "skills" at extracting money from the government. Those contractor firms NEVER end up selling back to private industry, because they have nothing to offer.
I think that this is one of those cases where "citation needed" is a way to avoid the blindingly obvious truth of the statement. No, we haven't done a controlled experiment on the government vs private enterprise in skill creation. Sometimes you have to use common sense and obvious experience.
Although I wouldn't be surprised if there's data to support what you stated, I also thin it's possible there are other confounding attributes.
Many government employees are motivated by different factors. Somebody who is motivated by pay would likely gravitate towards the private sector and starting their own business. Somebody motivated by service may be more apt to apply to government jobs. For instance, I would seriously doubt mathematicians who work for intelligence agencies are lower-skilled; this is probably an occupation that selects for other traits.
I am also aware of a LOT of former government employees who start their own companies but it's not clear how much of their success is due to skills vs. network effects.
i don't know that I would use "founded a tech startup" as a good proxy for skill creation. perhaps the type of people who acquire tech skills in government are aren't necessarily primed to found a startup
To elaborate on this, it's conceivable that startups come from tech giants not because of tech skill, but rather because of confidence and connections.
It reads to me more like Schmidt is proposing the government set up its own "US Digital Academy" so that they don't have to compete with MIT and Stanford. (I assume the model would be something like free or low tuition in exchange for a requirement to work for the government for x years upon graduation.)
I think you're probably closer to the truth there. Even if it isn't an ISA university, my first guess is that it's a way of laundering government funding to create an institution under private control. (This is how most great universities get their start anyway, it's nothing new).
I'm sure ES also has a vision for a modern university that has merit, but I'd be skeptical if it was about more free inquiry and public service - probably more alignment with industry.
Google execs often couch their plans with claims about how it's going to be good for the world, especially partnerships with the government. For instance, the takeover of Moffett field to provide a place for the founders to park their jets, as well as space for the new Google HQ.
>I'm sure ES also has a vision for a modern university that has merit,
I don't disagree with any of your points. Cornell Tech in NYC is arguably an example of reimagining things with at least a somewhat clean sheet of paper but I haven't studied it in any detail. One thing that is probably be clear is that there's too much stovepiping in universities in general. MIT's new interdisciplinary Stephen A. Schwarzman College of Computing is another example of trying to bring more disciplines under one roof.
If so they can build on established offerings like officer training programs that many countries have: get a free education at a military or civillian university in exchange for x years of service. The future of military engagements is likely to be heavily related to cyberspace, so you can enlist someone without the threat of field duty, which is likely far more attractive to many young people.
The biggest challenge in recruiting technical talent to the government is the GS payscale. This program is being recommended to Congress, so I assume it's an attempt to indirectly increase compensation by offering free tuition to people who commit to government employment- it also is likely attempting to target college age kids, who could accept government payscales more easily than someone who already works in the private sector and has gotten used to higher compensation. Schmidt is no dummy, and I expect that this would work, but it's still a hack around the fundamental problem of IT compensation scales that are roughly 50% of private sector levels.
A more straightforward solution would be to just create a new payscale (say, the Executive Technical Service) for IT professionals and other technical domains where federal compensation lags the private sector. The Senior Executive Service (the payscale for senior leadership like department and agency executives, including most political appointees) has about 8000 members- if you gave this system twice as many people and paid them all $100k more than the GS system would, you'd increase the federal budget by about $1.6 billion a year, or 0.04% of the 2019 budget, almost all of which you'd recover anyway by replacing IT contractors who generally cost the government between $250k and $350k each a year. Unfortunately, this solution would really piss off the big federal contractors, who would heavily lobby against it, which is why it doesn't happen even though more reasonable payscales have been found for federal doctors and lawyers.
You must ask yourself why the government is able to retain word-class managers and jurists despite stiff competition from the private sector. James Comey was an executive at the world's largest hedge fund, why did he quit for a government job where he was making 100x less? Any federal judge could've made a fortune in corporate law, why didn't they choose to? Because there is a sense of prestige and national duty about their posts. Creating a similar status for technologists in government would go a long way towards recruiting and retaining top talent.
If we want to convince people that being a government engineer is worthwhile and respected, starting with students seems reasonable.
I think it's more complicated than that, all of these roles come with substantial amounts of near unilateral executive power that can impact the way the country is run for decades to come.
Even all of the existing US government tech posts like the USDS and 18F are ultimately there purely to service the needs of a bureaucratic agency and their particular management.
For a technology specific role to have the same level of prestige as someone on the federal bench or leading the FBI, they'd need to be giving them unilateral regulatory powers over the broader technology industry in the country.
I'd just push back - technology is a tool to accomplish the missions of government.
The positions that run those missions - enforcing the law, ruling on the law, running Medicare, etc etc - should be more prestigious. That said, if the government wants to have prestigious technical work, it should be more open to taking big swings (in an agile way) on hard problems, but that requires investment and taking risks.
I think there would be comparable prestige if the government did something like roll out public broadband or made some sort of publicly owned social media.
Why? For some the reasons you list are certainly it. For many more, it's the revolving door. You make money in private industry, change the legal or regulatory structure and then go back to private industry to make even more.
My understanding is that you can sell stock (that might pose a conflict of interest) tax-free when you take a government role. I presume that a hedge fund manager such as James Comey would have significant unrealized capital gains that he could sell prior to taking his government position.
Gary Cohn, Trump's chief economic advisor did exactly this[0].
"The chief economic adviser was required to sell his Goldman Sachs shares -- and he did so tax-free and with perfect timing"
Maybe this article is missing some important details but I don't see how having a special university is going to help the problem of getting more technical talent into government work.
There are already tons of government scholarship programs with strings attached in the form of post graduation service requirements and other hiring funnels that bring in new graduates. The government's problem isn't hiring new grads it's keeping those people once they get a few years experience. The biggest drain is probably government contracting companies who often offer literally twice the salary for the same work. Combine that with the soul sucking aspects that you would find at any large company and you have the perfect recipe for not being able to recruit and retain qualified mid career people.
It's hard to imagine how any amount of indoctrination in the ideals of public service at a special school will overcome those things without something else changing.
I remember a story that Leland Stanford asked what is needed to found a university and the answer was 200 million dollars. What they didn't tell him, that it also takes 200 hundred years.
Universities are a longterm project, if a university is good or not, is something you don't really realise while you are there but when you are finished and try to use your knowledge.
I'll give you an example. MIT was founded in 1861. OK. Sorta bad timing. (Two days after it was chartered, the Civil War broke out.)
But even after it got going and moved to its Cambridge campus in 1916 (significantly funded by George Eastman--as in Eastman Kodak), it didn't really become known as a premier institution until WWII.
Not everything in the world can be measured by pay.
Sure, an FBI Criminal Profiler earns a tiny fraction of what a google dev earns. But one can take serial killers off the street, and the other just makes ads more annoying.
A Social Worker earns almost nothing, but they get to help the most needy of people get the help they need. A FB engineer earns 10x the pay, but just makes ads more annoying.
A teacher earns almost nothing, but they get to teach the next generation of society. An Amazon engineer earns 10x the pay, but just finds ways to sell you more stuff you don't need.
No one goes into government work b/c the pay, they do it for reasons they find more important.
Not to mention the fact that the government is one of the only funding sources for blue-sky research nowadays. Pretty much the only surviving research institutions that rival Bell Labs are the national laboratories. Getting to work with insanely bright people at the frontiers of your field (without the funding problems of academia) should be considered a benefit on its own. Also, many of the national labs pay quite well. They don’t use the USG pay scale which is a misconception I see very often in these discussions.
This is so true, but it's also an issue. If you can't make a living in government, then you're more likely to leave or turn to rent-seeking opportunities (i.e. becoming a lobbyist after a few years working on Capitol Hill). I worked in public policy for years but had 5 roommates even 5 years in and was saving zero money. In order to build a stable future for myself, I had to leave government. If I could've stayed working on the issues I was passionate about and been able to buy a house or start a family, then I may have stayed. But I couldn't do that, so I left...
I completely agree. Just the headline I didn't agree with. Universities for teaching, social work, fbi profilers will never be as prestigious as those where you can get trained to earn millions.
Except that isn't explicitly true. West point and the navel academy are some of the most prestigious universities in the entire country.
The fundamental difference between these things lies in the fact that a lot of government work comes with higher levels of responsibility and ownership in terms of the way in which you want to do your job.
There's tremendous demand to join things like the US foreign service officer program because as these people move though their career, there's basically an unwritten promise that they will have a legitimate impact on US state department policy for the countries they work in. Something that will only really ever be true in the world of government enterprise software work if you're the PM.
I'm not surprised. If you grow up as a computer nerd all your life with few recognizable achievements, and suddenly a Big Tech company dangles in front of you $100k or more, very few people can say no, especially if you also grew up from a relatively poor background.
I think it's actually something unique about Americans. When I started my company, I wanted our slogan to be "Helping each other, together". I very quickly realized that no one in the US actually responds to a saying like that. The vast majority simply doesn't "get" it.
Sadly it makes it harder to find a job as a software engineer in which I can work for a company or in a position making a positive impact on the world (as amorphous as that may sound).
Any interaction that I've had with a recruiter when considering looking.
Me: "I'm not interested in working on ad tech or a bank, etc. I'm interested in privacy companies, green technology, conservation, etc."
Recruiter: "I have a job with Investment Bank X that I think may be interested in."
Only if you assume that "the common good" in this country isn't a mirage (which tech workers are more likely to see, IMHO).
That phrase is roughly equivalent to how large corporations thank "essential workers" during the pandemic. It's marketing. It's an excuse to underpay and leave workers without sufficient support.
The meme that never dies. Advertising funded tons of tons of businesses on apps and on web, employed hundreds of thousands directly and created a large economy.
Disclaimer: Former google, current facebook employee, working in ads.
well, did adtech change the available income or why is it that it's so much bigger than the old way of doing advertisement?
I think the only point, why you are seeing this "growth" is that you have an oligopolic structure which is just skimming money (mainly to the benefit of the 0.1 percenters) without creating any additional value compared to the traditional way of doing advertising? Or how again are you increasing available income with your smokescreen of targeted ads?
EDIT: actually I think wikipedia captured it perfectly https://en.m.wikipedia.org/wiki/Rent-seeking - and no, financing some productive webbusiness (which there are?) with a minuscule fraction of the ad money or following personal spleens of the founders/crazy people is not altering the outcome.
Plenty of (say, gaming, entertainment etc) companies in developing countries employed plenty of people due to advertising revenue, which would otherwise not be viable.
Google display ads takes a fraction of revenue and passes the rest to publishers. Same with audience network.
Digital advertising lowered the barrier significantly to making profit for lots of digital businesses. It enabled lots of physical businesses to get recognition and discovery in one way or the other.
Not sure why this is not clear, but happy to discuss more.
Why not pay them directly in a functioning economy, if its worth it? Why do you need to skim from random (!) other ventures to finance this kind of business?
I'm also not sure, if I understand this correctly, but 32% (or 1 third) seems like quite a big "cut" for any service to me - but apparently ratios changed with the internet which allows placing these ads at such a low cost :)
https://support.google.com/adsense/answer/180195?hl=en
So, do these "digital businesses" create anything, or just skim money? Just because you've built a network of rent-seekers, this doesn't really change the concept.
Also could you explain how physical businesses actually benefit from the "new kind of ads" in a more concise way compared to the weasel-wording "one way or the other" - is more product sold because of ads in a certain area (doubtful, because ads won't change available income), does the percentage of advertising spend go down, when all business go "online" (doubtful as well, or how can you generate growth beyond the growth of the economy, which is the declared target of most adtechy-enterprises?!). So this is frankly neither clear to me nor apparently to you (nor anybody else I suppose), but these are the hard questions society has to answer, before accepting "adtech" as a productive sector of the economy.
Try not to pick on specific instances for the sake of argument. US military power and it's geopolitical position, combined with a bunch of other things made dollar the reserve currency. Fed is printing money yet there is no inflation except asset inflation. Noone can threaten us military, hence we will be for foreseeable future the country people will pour money on.
Surveillance capitalism was only impactful because there is no way to pay small amounts of money to view a digital good. No one is going to applaud you or your employer for setting up the greatest invasion of privacy the world has ever known.
I would go so far to say you have harmed the business models of tons and tons of apps by setting up an expectation that apps should be "free".
Yet there are still lots of people who work for various government and quasi-judicial organizations, and the military. Two big selling points:
(1) you will likely get lots of resources to do things that your typical private sector organization doesn't prioritize, like longer term research.
(2) Nobody thinks a job != your career anymore; you could start out with a few years in government service and get education & experience on their dime, then move into the private sector. As more opportunities to get started get harder to find or aquire, this will be a huge motivator.
The question isn't whether it will rival Stanford and MIT, because it never will with a one year program. The question is whether it could incrementally attract more talent into the industry.
The naming, "US Digital Service Academy", seems to suggest that it will be tied to the US Digital Service - which would be great. I think that might be the only government organization that I would actually be excited about working for.
Side note, I really hope the tour-of-duty model becomes more popular with techies, and more encouraged by employers. I can see it having numerous bidirectional benefits
The issue with the US government and software is not a lack of talent (within the country, or even within the government itself). But a poor structuring and assigning of responsibilities and a severely risk-averse culture, paired with the overreaction of a risk-accepting culture occasionally showing up and fucking it all.
There is, or will be, sufficient talent in the US to maintain and develop government software systems. Flat out, people keep going into CS and software engineering degree programs. Programming itself is ludicrously accessible compared to nearly any other technical discipline because the cost of entry is about $200 and time.
The problems that actually need to be resolved:
1. Government employees (contrary to popular opinion) are grossly underpaid for their expertise. This leads many of the best to leave and go to private industry and/or government contractors. Those who remain are truly civic minded, or zombies. Moving away from GS helps here (pay for time in grade), but the alternatives are also severely abused. NSPS paid people for changing positions, so there was a ton of job hopping before it was canceled. Expertise wasn't really cultivated, only the same stupid culture of bouncing between jobs you find in parts of private industry. AcqDemo is somewhat better, but in implementation lost all its potential value. Still better than GS, but only barely.
2. In-house development shops are typically made to compete for contracts in the same way as private companies. This creates some interesting conflicts within the organizations. And a ton of the in-house software development efforts are spent on maintenance, not developing new systems. So a different kind of expertise shows up, but not the expertise needed when truly novel projects start up.
3. The people in charge of the IT and software development efforts are usually inexperienced with IT and software development. They're either straight out of college EEs (CS people are at a disadvantage in many government shops) or career civil servant bureaucrats. The latter rely on those same inexperienced EEs. This leads to an incredibly advantage for contractors to push program offices around because they can come off as reliable experts, until you've been burned a few times. But hit a couple milestones and the civil servant in charge is already promoted out of that position and a new one is left holding the bag when shit hits the fan.
4. Risk averse IT severely restricts development environments even when software is made in-house. This leads to incredible slowdowns and often automation is impossible due to having to develop and test on multiple test/dev network enclaves that are separated even from each other. An example is the typical CCB (configuration control board) process which is a monthly (at best) meeting where, mostly, managers decide what new things can be installed on the network. You want an updated version of VS or GCC? Gotta wait a month. You want to use a common library? Gotta wait a month. And that's on top of the delays imposed by security scans (a manual process, that could be automated if it's necessary).
5. Risk accepting IT goes too far the other way as a reactionary response to the risk averse culture that held them back. Their fuck ups lead to the people in (4) being able to justify their positions and cement their culture.
You have to improve the approach to IT, the staffing of the acquisition offices to have competent individuals in the appropriate fields, not just successful (from a promotion perspective) bureaucrats, and probably increase pay across the board for technical people.
> The issue with the US government and software is not a lack of talent (within the country, or even within the government itself). But a poor structuring and assigning of responsibilities and a severely risk-averse culture, paired with the overreaction of a risk-accepting culture occasionally showing up and fucking it all.
Bureaucracy in a nutshell. I wonder if there is a single country the western world where it is actually efficient...
Bureaucracy will never be efficient. It is resilient and can survive. It is an organism. It presses itself into the cracks and crevices of society until it seems like the essential material composing our structure. But it's not, it's just slime mold.
While these issues exist, we should look to the countries that have successfully overcome them and shamelessly copy their successful strategies.
The US government already does really well at training people for the military, both leaders and fresh high school grads.
Personally, I think we need to stop passing 2000+ bills into law and instead give our technocrats some high level requirements that they can figure out themselves (hopefully in more of an agile than waterfall way). Assigning a technical person micromanaged requirements written by politicians and lobbyist lawyers will never be an appealing or successful formula.
Perhaps. So what countries overcame these issues by training new people and not addressing the internal systemic issues that lead to the drain or burnout of skilled workers?
How does training of new people address the established culture that encourages hitting milestones (internal) but not achieving the overall mission? How does training of new people address the fundamental problem of developing and testing using 3-20 separate networks with data sharing consisting of bobnet (sneakernet but Bob is the only one authorized to move data between systems)? How does training of new people prevent poaching by contractors or private companies paying 2-3x the salary? How does training of new people fix the problem of inexperienced (training isn't experience) novices running acquisition programs and getting bullied by contractors?
The fundamental model of IT acquisition and software development within the US government is broken. Training the next generation isn't enough, and universities already do that well enough. You have to break the existing system.
Yeah, my comment wasn't geared towards fixing those downstream issues before fixing the upstream ones. I'm not arguing "pay the existing employees more and expect output to magically change".
When I said we should look at successful country strategies, I expect those strategies will include things like "don't let legislators and lobbyists be apart of contract details -- only high level requirements" and "outlaw the bribery that you call donating to a legislator's political campaign". Perhaps even change the government worker model to encourage small enterprises within existing government (like USDS, 18F, etc) and empower them to help earlier in the design process -- like during requirements gathering.
The article doesn't mention it, but given the current crisis and the huge move to remote education underway currently it would be interesting if this had an approach that took that into account.
The elephant in the room when talking about making government work more like startups:
We don't allow governments to fail (easily) but startups fail frequently and we acknowledge that both bankruptcy and team failure leads to knowledge that is rare to gather without actually experiencing the failure.
I'm not advocating for anything in particular except to acknowledge this fact.
The federal government pay cap in 2019 was $166,500....
These folks are going to go to this university that is better than Stanford and MIT (let's assume this mammoth task happens) and then eschew working for big tech paychecks in favor of working for the government?
That is not entirely true, but the devil is in the details. As an example, the Federal government employs thousands of physicians at the VA and, depending on specialty, they are paid significantly more than this. Last time I checked, they capped out at $400k:
Anyway, the point is that the federal government can and does pay significantly more than the GS pay scale to people who are federal employees and enjoy federal pensions. Whether or not the feds will choose to pay that to those who work in technology remains to be seen.
I remember a couple departments (including the NSA) complaining about the federal cap when it came to some IT projects, so I'm not sure it is a sure deal to get past that cap.
The VA is a whole other land of bureaucracy that plays by its own rules.
The NSA and handfuls of other departments have figured out how to effectively use contractors (who are paid handsomely above that federal pay cap).
That's separate from the discussion of what these perverse incentives (eg. pay more to mercenaries than to lifelong government servants) do to morale, capabilities, and institutional knowledge. It gets far worse when the president casually says that we should just fire emergency specialists and suggest we can hire them back on short notice in the middle of an emergency.
Oh, I very much agree and they were most certainly capped at that organization. It's a major problem because and I very much agree with your original comment that there are better options now working for big tech.
I guess my point above is that feds can fix this if they truly wanted to. There's precedent for them hiring people for far more money than the regular GS rate when they need to. It would be impossible to hire a surgeon for $160k, but they need them, so they pay more. At the moment, the government doesn't feel they need to pay more, so they don't. I hope someday that will change and think it's important to educate and lobby the federal government to help them understand this and change their policy.
This isn't a decision that can be made within the executive branch (I think). Congress would have to pass laws that enable "bureaucrats to get paid more."
Peter Thiel was thinking about starting a university at some point too. He decided not to because it was too hard. Seems like it could be a great idea though. Government at all levels in the US has a lot of problems and getting smart, reform-minded people into it would hopefully help correcting that.
1. What guarantees does it offer for ideology neutrality?
2. What guarantees does it offer for the diversity of ideas?
3. What guarantees does it offer for excluding censorship of the expression of thought as a measurable true defense of freedom of thought and expression?
"Per the report, the Digital Service Academy would serve as a third reservoir for the best and brightest technology workers, but one that would ideally produce graduates imbued with a sense of government duty."
... because places like MIT and Stanford sure as hell don't!
It looks like Schmidt just wants the next generation of students to forgo their dreams of starting the next Google.
I think a better idea to help this country stay competitive is for companies like Google to stop offshoring their operations ... and pay their goddamn taxes like the rest of us.
I can't speak for Stanford but MIT is extremely closely tied with the federal government. The Lincoln Lab for instance gets ~$1B/year from the Defense Department. The only reason MIT became a well known institution at all is because of its ties to the military.
This is far from true. If you look at the total workforce in tech, the majority is by far from much larger, public schools, like Berkeley, Michigan, Illinois, Washington, Texas, Wisconsin, etc., etc. MIT and Stanford have had amazing great people, but they are far from the "primary" source of talent.
(yes, a bit of an aside, but it rankles)