The Disposable Academic: Why doing a PhD is often a waste of time (2010) | Hacker News Reader

The Disposable Academic: Why doing a PhD is often a waste of time (2010)

economist.com

47 pointsKopion57 comments

A computer science Ph.D. is nothing like what is described in this article. For various reasons (other opportunities, generous industry grants, minimal financial overhead to run a lab), students in computer science are comparatively well paid and free to work at their own pace. YMMV based on advisor.

It's been exhaustively shown that stopping after undergrad or masters degrees will be more financially lucrative, so some reasons to go on:

  - you want one of the specialized researchy positions that
      are typically filled by Ph.D. holders
  - you have a genuine interest in some specialized field and
      want to work purely on learning/investigating that field
  - you want to be a professor

Unfortunately the "generous grants" often conflict with "free to work at their own pace". It's getting more common to find arrangements where PhD students essentially have a part-time job working for their RA funding. In the best case this overlaps with their own research, but sometimes doesn't, and may involve considerable additional work that doesn't really "count" as research. Particularly the case if you're funded on a DARPA grant, since those have very specific timelines and deliverables, plus days where you have to be ready for a "site visit" to demo things. Some industry grants are fairly hands-on as well, expecting the lab to basically lend out a PhD student to work part-time on a project the company wants completed. Though others are more hands-off and just donate a pile of money without demanding specific things in return.

I would limit that to just #2 in your list. The other two carrots are so out of your reach or subject to factors so far beyond your control that a love for learning and a burning interest in your research is the only thing that you can count on sustaining you.

#1 is also pretty achievable. For example, if you get a Ph.D. studying how to implement compilers and virtual machines, you basically have your choice of which place that implements compiler or VMs to work at (source: I am about to defend my Ph.D. in implementing compilers and VMs).

But how many places are there that do that? You have google, Microsoft, maybe intel, a few companies in the valley. PhDs are very specialized, and so they really can be at the mercy of their hot skill.

And even google...might tell you to go work on ads with your hot VM/compiler skills. Disclosure: I'm a PL PhD who got out of the vm/compiler area to focus on PL design, which is even more esoteric.

I know of quite a few data analytics companies that hire mainly PhD's for their research positions. Many traditional companies do so too, to get on the big data trend.

Yes, but they are hiring PhDs in systems, PL (for some reason, we are good at MapReduce), database, and some data mining fields. Its not like you want to hire an HCI researcher to do big data.

I don't know about google - all the PL people I know who went to google no longer work on PL stuff, though I don't know if that's because of working at offices where PL work is not supported rather than a lack of company-wide freedom.

From conversations with hiring managers at each, Apple, Microsoft, Intel, NVidia, and Adobe all have far more positions than they can fill for people with PhD-level VM/compiler writing skills. Heck, we can't turn out experienced _undergrads_ fast enough for their tastes. I assume there would also be stuff in embedded systems, particularly FPGA compiler-related, though I don't personally know hiring managers at the big firms in that area.

Google has found PL PhDs very versatile I guess. Many of their top people are in PL, so it could also just be a top down hiring bias.

I'd be weary taking just any job in VM/compilers, much of it is quite tedious and boring for a PhD. Intel used to have compiler groups filled with PhDs, Sun also. Them there are smaller companies like coverity.

I think it depends on whose perspective you're talking about. I think for typical engineers, especially in EE/CS, the increase in earning power from the PhD doesn't make up for the six years of lost income and career progression. But I think from the point of view of society, it's absolutely worth it. The most brilliant engineers I know are PhDs. They have this incredible depth of theoretical understanding of a particular area that I think is hard to achieve without those years of dedicated study.

It's hard to know which way the causality goes here. It might just be that really exceptional undergrads happen to end up in grad school (because of pressure from professors or family, or simply because "it's just what smart people do"), even though they would have been exceptional engineers either way. I agree that PhDs are very knowledgable when it comes to their precise field of study, but this is often very narrow. How useful is this expert knowledge unless you're hired specifically for it (ie, as a researcher)?

Socially, there is a big value in getting a very bright person to intensely study a single area for many years. PhDs do usually go on to work in their field, and a vastly disproportionate amount of innovation comes from them. That's why I don't really buy the correlation argument. I don't think even very bright people are going to be able to make contributions like new speech recognition algorithms or new algorithms for wireless networking, robotics, etc, without the kind of dedicated and focused study that comes from a PhD (or an equivalent level of theoretical study outside a PhD program).

I think you're right about some classes of PhD and benefit for society. It may be a case of market failure; I'd certainly prefer to support PhD programs in CS through charity and/or taxes than a lot of other things.

I do sometimes wonder if this "PhD is a waste of time" is being pushed so heavily to get a glut of the best computer science grads, who would otherwise consider a PhD, onto the engineer job market. A cute little trick to drive down the salaries a bit, cut back on the perks...

I mean, people posting links like this, on this board, don't really have English lit grads in mind, do they?

Not that I disagree with the sentiment entirely, but it seems to be getting a lot of mileage the past few months.

I think this is just a continuation of the old "College is a waste of time" thought, but I have noticed a _lot_ of people -- especially from China and India -- enrolling in CS PhD programs the last few years. I talked to a few of them from my college, and an exceedingly common theme seems to be family and peer pressure, so personally, I'd really like to understand more about this topic.

An important, but rarely overtly stated reason for foreigners to attend grad school in the United States is immigration. A graduate degree (esp in STEM disciplines) is a relatively secure way of getting a student visa, getting a job while on OPT, getting an H1B and finally a Green Card. Of course it does not mean that immigration is the sole reason for asians to attend US grad school in hordes, but it is an important one.

Funnily enough, when you apply for your student visa, you're required to assert that you will return to your country after your education in the US is over :-)

I don't understand the motivation there at all - shouldn't the US be trying to hold on to the most educated foreigners as much as possible?

You made the mistake of assuming that the US immigration policy is rational.

One country that does it right is Canada--they try very hard to keep top talent in the country instead of going to their previous homes.

I used to ask that question and have been told that the US greatly benefits from having US-educated people in positions of power around the world. I don't know if this is true, but what is true that a surprising number of heads of states of countries that are considered political allies of the US have degrees from US universities. I first noticed this trend when reading about Pacific island nations.

I suppose, but the number of people leaving the country that that requires is fairly small, unless you're talking about spreading a cultural hegemony. I think it's likely just irrational.

If we follow this logic for the subset of foreigners attending grad school in the US, why wouldn't they opt for a Masters degree instead of a PhD?

I don't think CS grads, however good, will make much of a dent in mech or chem eng salaries and perks...

Well, I meant the software engineer job market...

I don't think mechanical engineers get any perks anyway.

> I don't think mechanical engineers get any perks anyway.

Not unless you consider going home at 5 pm a perk.

On HN I doubt it. I'd say we keep seeing the stories up here because people want to validate that they took the right choice. I'd personally say your better off doing that by convincing yourself you are working on something useful, not by shitting on other's choices. It would be the equivalent of justifying a PhD by saying "I don't think our best and brightest should be trying to get people to click on ads."

As a hiring manager of non-specialized software developers, I take people with 4-6 years of real world coding over a phd any day. We generally hire phd's at the same level/pay as a bachelors in cs with maybe only a nominal amount more in salary. the skills you learn as a phd are not really what the industry needs, unless you manage to find one of the small number of jobs that requires whatever specific thing you studied as a phd. like the article says, those are very few and far between compared to the # of students exiting with degrees.

I concur. My experience with CS PhDs is that they lack the production-oriented mindset needed for the real world. We hired a CS PhD to write reports for us. In six months he did not write one single line of code, but produced a very impressive whitepaper about the code he intended to write. He went on to write more whitepapers for a company specializing in vaporware somewhere else.

I've had similar experiences with PhDs in many other fields as well. I have a great deal of respect for the degree, the person who worked so hard for it, and I love to read their often brilliant research, but either the effort of getting the degree or the mindset it takes to want to obtain one makes these people nearly useless in the business world. Give me an entry-level BS degree or 5-6 years equivalent experience any day.

Meanwhile, the top tech firms aggressively recruit CS PhDs.

Well, more power to them. They have the resources to train them up to a level where they'll be productive.

Or perhaps this thread is full of obscene generalizations that are untrue. Hell, it even started off on the wrong foot by invoking "real world". All that term means is that you think your work is somehow relevant while the work of others, in this case PhDs, is not relevant.

Parts of this community have a real problem with ego and unsubstantiated generalizations.

No, real world = in production for potentially a long time, iterated with customer feedback. phd/ms work = functional hacks that maybe a professor or the student ever looks at/codes against and that are eventually thrown away.

except in very rare circumstances

Ya, google is sucking up many of them. But don't tell parent, he doesn't get it and doesn't need to get it.

In my experience, there are two kinds of CS PhDs, one which is more theory oriented and not very good at coding, and the other which is more systems oriented and tends to attract crack coders too bored to stay in industry.

I find your experience with PhDs fascinating because it's been the exact opposite of my experience with graduate school. My graduate program has been far more production oriented than any of the private sector work I've ever done.

In the private sector, I've been given a list of specifications and told to report back in a week on what progress I'd made. In research, on the other hand, a professor would hand me the requirements and then stand there until I delivered the code. If the professor was on a looser deadline, he might take a chair while he waited me. Most deadlines are measured in hours, instead of days, so the thought of a PhD not having his code finished in a month is baffling to me.

Of course, I'll also be the first to admit that this still produces some poor production habits. When you have a twenty minute deadline to write code in a language you've never seen before on during your first ever experience on a VAX, you tend to skip the unit tests. And the source control. And the comments. And the white space. But, I'd like to think that, if I had an extra hour, I'd go back for all of that.

Perhaps PhDs just need to be paired with real world coders to be productive?

Having worked with a number of PhD and non-PhD people over the years I've not observed any correlation between "skill" and "years of education."

That said, its always to pair up new hires with real world coders to help them become productive quickly.

I am in the process of finishing up a PhD (Robotics, Cornell) and spent some time looking for jobs outside academia. What you write sounds very familiar because I heard it almost everywhere I looked. There seems to be a widespread misconception that graduate school consists of people focusing on one (usually obscure) topic without ever doing anything that "normal people" would do in a "normal job".

Here's how grad school really works.

At least in the US, completing a PhD requires you to take classes. I took classes in my field (Mechanical Engineering) and in my elected minor field (CS). None of them had anything to do with my research are (modular robotics): Vibrations, Linear Systems, Feedback Control, Intermediate Dynamics, Intro to AI, 3D Perception, Evolutionary Methods, Autonomous Mobile Robots (you get the idea, there are more). I also took a management class and a course on technology commercialization, but that's just me. Basically, they are the same classes Master level students take (and companies pay them a premium for doing so), just more of them.

Very few grad students work on only one project during their PhD program. Many students "bounce around" for a while before selecting an adviser. Sometimes projects don't work out and you switch to something else (happened to me). Sometimes the availability of grants dictates what you are working on and when. And often you just start a new project because the old one is completed. What causes every PhD to have that brief, often obscure sounding, description for their work, is that the thesis describing all those different projects needs one title (not five). When you open the cover of the thesis you might find sections on a wide range of topics covering many areas of expertise that the author had to master before being able to produce original work in their obscure topic area.

There might be disciplines where you can work exclusively on one subject area without using any external tools. I don't know any. As a PhD student you rarely have someone to delegate work to. And because what you do in a PhD is new, you often have to build your own equipment for doing it. As a result of having to do everything yourself, you often end up accumulating experience of varying depth in all kinds of languages and technologies during grad school. Dealing with external suppliers and contract manufacturers is also quite common, at least in engineering. My actual everyday work in grad school is definitely more varied than any job in any of the companies I have seen.

Soft skills are something frequently cited as an area that PhDs are lacking in. I don't agree with this assessment. Some skills are less important in academia (negotiation skills, leadership), but others are more important (endurance, teaching/presentation skills, ability to learn quickly). Only very few scientists are the socially awkward eggheads that popular media portrait them as.

Finally, it's worth pointing out that PhDs have spent several years working in the future. The technology that in development at universities today includes things you might consider cutting edge a few years from now. Your freshly graduate PhD job applicant just spent five years interacting with this stuff and the people building it on a daily basis. If your company can't make use of having somebody from the future work for you, I'd consider that a problem.

Please do yourself (and people like me) a favor and kindly stop pigeon-holing anyone with a PhD as "overspecialized". Have a look at what kind of work they were doing during their five year time in the ivory tower, and see what useful skills they needed to get their maybe-not-so-useful research done. You might be surprised at the breadth. And don't worry if your PhD job applicant's qualifications are a little strange and don't quite fit between any of the department walls in your company. The PhD is used to working interdisciplinary and learning new skills quickly.

you know what I didn't see in your reply? All of the large scale coding projects you implemented and put into production for a while. The big problem is that PhD/MS in CS is just an extension of BS, except you now have 4+ more years of bad coding practice under your belt. From my experience, coming out of school people need to relearn how to code in the real world. It is no more assignments you get functional, or in the case of PhD's mostly throw-away software/hacks that does something for your thesis. This is what I mean by skills that I need. Taking a bunch of random classes is nice for you, but none of them teach you how to write code or design systems in the business world. At least bachelor degrees only have ~4 years of bad coding behind them, and they have all of the fundamentals at that point necessary to be great software developers. Adding another 4+ years of bad coding on top of that does not really help you, this is why we pay PhD/MS candidates the same as a BS CS, because from our perspective they are the same!

I was writing about grad school. Talking about my hobby projects that have been "in production for a while" would have been out of context. We can talk about those elsewhere if you want. Also note that this was about Mechanical Engineering. We Mechanical Engineers are indeed not very strong at writing "large scale" enterprise level software, even after years of industry experience. The only CS PhD graduate so far from our lab wrote [1] and this seems to have been not too bad in terms of quality because it had thousands of happy users and is now a successful commercial product.

The point of my comment was that you, and many others here and elsewhere, seem to assume that completing a PhD requires very specialized work only, which is wrong in my experience. I did not make any comment about whether PhD graduates have the skills that match the requirements of the positions that you are trying to fill.

Your focus is exclusively on coding habits and you do not seem to assign any value on other skills. It appears to me that you are seeking highly specialized people whose only expertise is the production of high quality code. In that context it seems rather amusing that you are suggesting that others are too specialized to be useful. Maybe PhD graduates are in fact too generalist to be a good match for your positions?

[1] http://creativemachines.cornell.edu/eureqa

And you are in the vast minority in paying PhD/MS candidates the same as BS grads. I can say with confidence that you've hired the absolute worst most desperate PhD/MS candidates. This isn't something I'm making up, it's widely available salary statistics which as a "hiring manager" you should be intimately familiar with.

Academics work on real world problems, a great many solutions which you use daily. They care about accuracy, otherwise the science is bad. The only thing they don't design for, and this is because of the nature of research, not bad practice, is sustainable code that you can build on. Why try to build a great design when your requirements can change weekly? All you will do is slow down your research work and end up with a jumbled mess.

The right thing to do when exploring is write quick correct code with little thought to sustainable design. If you find an idea that works, rewrite it into something you can maintain.

All you've demonstrated here is your ignorance both of your own domain as well as that of others. You don't understand what PhD/MS grads are going for, are unable to hire good ones, and don't understand what it takes to research new ideas. That latter point probably means that your company is not at all innovative.

This article was pretty good at identifying drawbacks of higher education. (unlike the recent article by the literature PhD)

There's so many assumptions surrounding a PhD degree. "I'll make more money!" Maybe, but it takes a lot of years to make up for making $0 for several years. "the only jobs for a PhD is in academia" depends on the subject, last I checked google hires lots of PhDs. "I'll only work 5 hours a week!" just because you only saw your professor when you showed up for class does not mean those were the only hours your prof was working. "PhDs are super smart!" maybe, or they're just good at jumping through hoops.

The article pointed out that there is a glut of PhDs compared to available academic jobs. But then laments that only 57% of PhD students finish the degree within 10 years. The author seemingly argues that more people should graduate sooner, and increase the glut?

It's not about smarts, self-motivation and leadership are required to finish a PhD. If someone expects to have their hand held and be told what to do, they might never finish.

The article also says that these students are the smartest of their class. If that's true, then they are smart enough to view the job market and weigh the pros and cons of higher education themselves. (if they can't do that, then I'd argue they are not the brightest in their class after all)

I wonder when the first article was written claiming that doing a PhD was a waste of time? I'll bet it was in the 19th century. And we'll no doubt keep seeing such articles until such time as the PhD fades from history.

(Nothing against the article in question, which is a better-than-usual discussion of the issue.)

Getting a Ph.D. is "often a waste of time". Sometimes it's much worse: I've seen too many lives seriously hurt or even destroyed, as in they died, e.g., from suicide from clinical depression, from the stresses of a Ph.D. program. Some of the students hurt or killed were brilliant, some of the best undergraduate students in the world, PBK, 'summa cum laude', Woodrow Wilson, NSF, etc.

Still the fact that a Ph.D. is "often a waste of time" should not be regarded as very important. Why? Because that statement is essentially a probability. So, if W is the 'event' of a waste of time and P is a probability measure, then the claim is that the probability of W, P(W), is too large. I'd agree P(W) is too large.

But a given student usually doesn't have to 'face' only P(W) and, instead, has information about some more 'events', say, A, B, and C, that are relevant, and then gets to observe an estimate of the 'conditional probability of W given A, B, and C', that is, P(W|A, B, C). Then, even if P(W) is too high, P(W|A, B, C) can be quite low in which case that student might go ahead.

We know that a Ph.D. is the 'union card' for professors in college and universities. While there can be a lot to discuss about a career as a professor, I set that aside as, at least for HN readers, not very interesting.

But, what about a Ph.D. for careers other than being a professor? Okay:

I limit my discussion to Ph.D. degrees in 'technical' fields. I omit biomedical fields due to my limited knowledge of those fields.

There are pros and cons:

Cons. In nearly every line of work other than being a professor, having a Ph.D. is 'unusual'. So, in particular, if a Ph.D. takes a a job, say, job J, at company UGE, then somewhere in the path in the tree of the organization chart of company UGE from job J to the CEO likely there will be a person S without a Ph.D. where person S has a Ph.D. reporting to them. This is an uncomfortable situation; lawyers know this, and supposedly their profession tries to arrange that any working lawyer reports only to a lawyer. This unfortunate situation is a special case of the larger situation that Ph.D. holders in technical fields do not have a strong 'profession' to protect them.

A Ph.D. in electronic engineering can at times want to swap their degree for an electrician's license -- good union card, key to a sole proprietorship that commonly pays well enough to buy a house and support a family, a rock solidly stable industry, a great geographical barrier to entry.

Why "awkward"? Because person S will be afraid of two cases: (1) The Ph.D. training is irrelevant so that the Ph.D. holder will have their degree influencing their job performance in ways irrelevant to the job. (2) The Ph.D. training is relevant in which case the subordinate may conclude that their supervisor S is low on competence and, thus, be insubordinate and, with their extra competence, may be a threat to the career of person S.

So, if get a technical Ph.D. and pursue a non-academic job, then likely will be in an awkward position without help of a recognized, respected profession.

Pro. One aid in making 'the big bucks' is to do some work that meets three criteria: -- powerful, valuable, and new. If everyone over at UGE is determined to hire people only like themselves, keep down any additional competence, and, net, just keep doing things they way they long have been doing them, then a person with work meeting the three criteria might have a golden opportunity. Call it 'innovation', 'creative destruction', 'survival of the fittest', or just blowing the doors off those other guys, getting the big house, winter vacation house, summer vacation house, sports cars, good, old Stradivarius violin for a talented daughter, help getting into Princeton for another child, yacht, for wine for dinner parties, shopping, for a few dozen cases, between Beaune and Dijon, seat at the Round Table of the university president, for a child help getting through graduate school and seed funding for a startup, etc. Or just quite a lot of additional financial security.

Of course, in this case will be 'different'; maybe in a good way different but definitely not the same. Often will not 'fit in'. E.g., would not be hired at UGE.

But being different is commonly what goes with success, which in itself is 'different'.

So, what does a Ph.D. have to do with doing some work that has the three criteria of powerful, valuable, and new? Answer, a lot: Those three criteria are essentially those of some successful original research -- the usual statement is "new, correct, and significant" but for non-academic work we can substitute the three criteria I gave. So, the crucial work is necessarily about 'research'.

So, how to do 'research'? Well, history shows that commonly it's not easy, but there's education for that, called Ph.D. programs at the best 20 or so US research universities -- many of the names are well known, Harvard, MIT, Cornell, Brown, Yale, Columbia, NYU (and Courant), Princeton, Johns Hopkins, UNC, Georgia Tech, CMU (see, Jerry, I didn't leave you out!), University of Chicago, University of Washington, Berkeley, Stanford, Cal Tech.

Nearly always the main difficulty in a Ph.D. program is the research part. For a good, succinct description of much of the challenge, buried in at least the machine readable copy of D. Knuth's 'The TeXBook' is

     "The traditional way is to
     put off all creative
     aspects until the last part
     of graduate school.  For
     seventeen or more years, a
     student is taught
     examsmanship, then suddenly
     after passing enough exams
     in graduate school he's
     told to do something
     original."

And there can be other challenges.

But as difficult as the original work is via a Ph.D. program, commonly such work is more difficult otherwise. So, for someone in business trying to do work with the three criteria I listed, a Ph.D. program should be a big help.

In my Web 2.0 startup, the computing is often a pain in the back side due to poorly written documentation, system management headaches, security holes, ..., bugs but is all just routine. The key to my startup, if it is successful, will be some work that meets the three criteria, and for that work I have to admit that my Ph.D. program helped.

So, right, glad you noticed: If get a technical Ph.D. and want to use it outside of academics, then are heavily betting on doing some work that meets the three criteria and being a founder of a successful startup.

Maybe there are too few examples successful Web 2.0 companies founded by a Ph.D. who did some research crucial to the success of the company. Then can take this situation as evidence that (1) a Ph.D. and some research are not powerful or (2) there is a wide open field full of opportunities. My guess is (2).

I know; I know: But why can't there be just a nice, easy, secure, 20-30 hour a week job that will provide good financial security for life? Well, Obama already has that job (that's intended to be a joke!).

I've been "All But Dissertation" for two years now for a PhD in sociology. I haven't touched my dissertation after having my proposal approved. Instead, I co-founded an organization that addresses the problem that I discovered in my preliminary research. Right now, my decision feels right.

Are you sure you aren't just procrastinating? I'm joking, it sounds like you are doing the right thing.

Hehehehe, this is what my advisor thinks :) But honestly, right now, it just feels so good to reject the idea that a doctorate is everything. I've been in school for 12 years!

I was in school for 13 years (5+8). Procrastination is a common grad student problem.

"How should we make it attractive for them [young people] to spend 5,6,7 years in our field, be satisfied, learn about excitement, but finally be qualified to find other possibilities?" -- H. Schopper

old news

The decision to enter a PhD program only makes sense for one of the two reasons: either you think you're going to win a Nobel Prize one day, or, you don't mind spending five years in a way which doesn't advance any goal (literal quote from a conversation I had, "so you had an offer from Google but you decided to do a PhD instead--why??", "Oh, I just want to do nothing for a while longer".)

The two kinds of PhD students may look the same but they're very different. If you belong to the first category, even if you realize you won't get that Nobel Prize, I doubt you're going to regret trying. If you're of the second kind, why yes, you have wasted five years of your life, which is exactly what you signed up for so why do you complain?

I did a PhD to learn and to contribute some small bit to our collective knowledge. I didn't do it to be rich nor famous. As far as "waste of time" goes, there is nothing to me more important than contributing to the worlds knowledge. Pop stars come and go and the rich are forgotten and irrelevant just as fast. Knowledge is the only thing that accumulates and continually helps future generations.

This is a very ignorant and short-sighted response. While I agree that getting a PhD is not for everyone, you can't make a blanket statement regarding its uselessness.

As a relevant data point. I did a PhD because I wanted to learn a lot more about Computer Science. I did most of it while working at pretty good software development firms. Most of the people I work with have PhDs. They use some of what they studied at work, and I doubt undergraduates would be nearly as effective.

There are a lot of advantages for PhDs in CS, if you like learning and studying. If you don't, there's really no reason to do a PhD, you're not going to get an outcome that you'd like. The types of jobs that work well for PhDs require (at least) some continuing research.

Look at the interesting things Google is doing. Driverless cars is headed up by Sabastian Thrun (PhD). The technical lead on Google Glass is Thad Starner (PhD and professor at my alma mater). Heck, Page and Brin were halfway through their PhD program at Stanford (both have an MS from there) before they founded Google. PageRank was exactly the kind of thing that comes from PhD projects (extension and application of theoretical principles to new areas). And Dr. Starner has been working on wearable computing for two decades.

All these things didn't just appear out of nowhere because Google hired a bunch of bright fresh college graduates.

Google's early success was much about the supportive environment for founding a company, not the academics: networking with investors and early hires, getting recommendations from influential professors, than it was about the "PhD project" of PageRank. It was a standard graph algorithm executed on the Internet, not a ground-breaking academic project. (Obviously -- they left school before they did much any work on PageRank)

If PageRank were a new idea now, Google.com could have come up through YC.

It's not that simple and they did a prototype and published a paper before they left Stanford (three years in). See: http://en.wikipedia.org/wiki/PageRank. It was conceived as a "citation graph of the web" based on theoretical work in that area: http://web.archive.org/web/20020506051802/www-diglib.stanfor....

You say that today they could've come up through YC. I'm not sure that's true. Would they have the necessary grounding in those "standard graph algorithms" if they had just come straight to YC (or bearing in mind Stanford undergraduates benefit tremendously from the graduate program--skipped college entirely?) I think you're downplaying the interaction of study and invention here.

Reply on news.ycombinator.com