"A PhD is safer. If you don’t do so well, you are going to get the title anyways. A Startup is riskier. If you don’t do very well (90% of the time), you can end broke."

Plenty of people drop out without finishing their PhD and generally they are broke afterwards. I'd say a startup is safer as you are closer to industry in case you need to find a job.

I also disagree with the author's point that a Ph.D. is about changing the world. My seven years of hard work culminated in a thesis that had a negligible impact on the universe. Many of my friends justifiably felt the same way. I set out with the intention of making the world a better place, but reality had other ideas. It's hard to see how spending those seven years at a startup company could have been worse in that respect.

I was working for a big tech firm in Silicon Valley in late 2000. I was looking for a career change so I applied to a few startups including Google. The Google recruiter called me repeatedly to come in for an interview, but I decided to quit the job search and pursue a PhD instead.

Studying for a PhD was fun, but I couldn't adjust to the lower income. I'd been working in industry for the previous five years so it was a 80% drop in income at least. Plus I got married and had a kid in those two years. That complicated things considerably.

I believe the author is suggesting that the person starts their own startup, not going to work for one. I would suggest in that case theere would be a considerable drop in income also.

Startups are for pragmatic (as opposed to idealistic) self-starters.

I think most PhD students identify with "I want to think deeply and learn a lot about one thing" much more than they identify with "I am a self-starter who loves to work in an unstructured environment."

There are things that a startup can never replace. Most notably, access to domain experts whose goal is obtaining knowledge. In startup land, you get massive amounts of anecdote, but little knowledge. Almost all of that anecdote is geared towards earning money to the exclusion of all else. In academia, you get essentially the opposite.

The reason I say that academia can be more constraining is that often you end up focussing on a single problem to the exclusion of all else. This can, but certainly does not always, harm your breadth of knowledge. In transition to a startup, the biggest difference to me was dispensing with "publish or perish."

I've reproduced and experimented with more existing work than I ever did during my PhD. That's mostly because reproducing existing work doesn't earn you anything in academia, despite being a cornerstone of science. You spend as much time, and only as much time, reading background work as is necessary to do something new. Without having to publish, I've been free to really expand my breadth and understand which methods work and which don't in my new domain.

All that said, what I'm doing in a startup is largely still applied research. This is a whole different beast than I see in most startups. Their goals appear to be chasing populism and money. The antithesis of academia.

In my graduate work I have spent a great deal of time reproducing the work of others. This is because there have yet to be well defined benchmarks for the performance of various algorithms. In order to show that a new algorithm is superior to the current state of the art, it is necessary to implement the other algorithms.

This has resulted in a open source program that implements various published algorithms, where the authors did not release any program to implement their schemes.

Chiming in as a counter-example! Having an unstructured environment with few constraints is incredibly rewarding. If you pressed me, it might actually be my favorite part about being in academia.

(My experience with other PhD students agrees with your generalization, though.)

[1] http://shebang.ws/startup-or-research.html

on one side: the majority of engineers working at startups are hired and paid.

on the other side:

The vast majority of CS phds can't get an academic job unless they don't have any constraints on where they wish to live. The moment you have any location preferences and want an academic job, you have to be total super star, which is hard, and by definition few people will be. Theres also the fact that the typical phd recipient will never earn back the lost income from the extra years of education over their master degree only clone.

being a startup founder or a phd student, both can suck, they're fundamentally different experiences, and when one or another works out well for someone, its an incredibly unique and personal situation that will not be replicable by anyone else.

[1] http://www9.georgetown.edu/grad/gppi/hpi/cew/pdfs/collegepay...

You are right, the income element is orthogonal to phd or not. But if one cares about where one lives AND enjoying ones work, more than one cares about teaching at a university somewhere on the planet, they'll be much happier not doing a phd program.

Research is fun, learning is fun, engineering great new tech is fun. A phd is a university teaching certificante, not a magic "i can do good research and build great tech" certificate (though they can be highly correlated.).

And comparing with science long-dead science luminaries is unfair - it used to be possible to make a (scientific) breakthrough working alone. But, say, for 50 years or so it is no longer true (the transition was smooth). When a field is fresh you can make a great in your garage; latter - not so.

Also, academia is a strictly top-down organization with all its consequences for creativity (in particular, having a great idea and skills at 17 yo will allow you to startup, put not - to assemble a research group). So if there is a great impact to the world by a PhD student, I can bet a lot, that it is because it was the PhD student's side project.

And this is where I stopped taking the author seriously - they've got a seriously rose colored view of the job market.

As a PhD student, I tend to think the opposite.

Entrepreneurs who are successful tend to change the world, or at least improve it. Even if you have a small business, you are helping fuel the economy.

It seems that most academics tend not to improve the world or fuel the economy.

[1] : http://matt.might.net/articles/phd-school-in-pictures/

Also, I think Google came out of Page's & Brin's PhD work, but they decided that building a company around it was a better idea than getting PhDs for it.

This is a nit pick, but all the examples you gave are of people who actually changed the world, not just some sliver of academia.

That said, changing your sliver of academia is likely to imply changing the world if you're in certain areas of math or science or engineering.

The only time when you can change your sliver of academia but not change the real world, is when your sliver of academia does not contribute to the real world, which is (unfortunately) common, particularly in the humanities. (Which is sad, because we actually really need good humanities study/education in order to build a good world.)

I think that's going way too far in the other direction.

I certainly know of PhD theses in my research area that have made contributions to the area, and these contributions (or derivative ones) are very likely to make it into real products eventually.

A lot of PhD computer science students have the benefit of taking 5 years or so to really work hard on something consistently that would be very hard to accomplish with the same level of rigour in industry.

So at the end of 5 years, what have you got? With the Ph.D you've got something recognizable and real academic cred. With the startup it'd be a bit of a mixed bag. If you knew the startup would fail, would you still choose it over the Ph.D?

It's not that safe.

Yes there is a possibility to make it rich without all of the above, but there is an extremely high chance of failing. A lot of people who are making such claims about education are making them from a dishonest position.

Europeans do it in 3 or 4.

> In a PhD you have a scholarship and a tutor that tells you what to do.

Salary is usually contingent on applications, work, etc.

After the first half, the PhD should be directing the project, not the tutor.

> A PhD is safer. If you don’t do so well, you are going to get the title anyways.

A PhD is only a PhD if you make it to the endgame. You can fail quickly or slowly in an infinite cornucopia of ways.

> In a PhD you learn a lot about an specific subject.

If you have two brain cells to rub together, you learn just as much about yourself, people's motivations, etc. as you do about the domain.

> If what you want to do is to change the world, go with the PhD.

A PhD is at best learning to do research, not much else, and at worst, a poorly paid programmer or lab tech. Changing the world is thing altogether.

I agree that PhD's and startups have similar elements of risk. In both spaces, it's possible to stick to safe ideas or ideas that don't meet approval of your colleagues.

Startups teach you to be responsible for your own ideas, in a way the PhD system does not.

Not every PhD student has a scholarship. The ones who don't get a fellowship, tend not to accept. The ones that do get money still don't really understand where that money comes from.

The final point is also very very wrong. PhD's don't change the world, they push the boundary of what we know a tiny tiny bit forward. And it's widely known that a start up is not the most effective way to earn lots of money, investment banking or working your way up the corporate ladder is. The main difference is that a start up is probably more fun and fulfilling.

I think the post is meant to be about an American PhD program, not a European one. 3-year PhDs in the US are pretty rare.

I came across a joke once that resonated since it felt like truth being spoken in the clothes of humour -- "When you finish a bachelor's degree, you think you know something about something. When you finish your Masters, you realize you don't know anything about anything and are in awe of other minds around you. When you finish your PhD, you realize that nobody really knows anything about anything, including you."

The offshoot of that last realization is true humility that shows as a deep appreciation for the steady progress being made at the boundary of knowledge, and the wisdom in finding the right question to ask.

To put it another way, both endeavours have tremendous impact on the person. If "finding out about stuff" is your thing, getting a PhD would be worth it. If "doing stuff" is your thing, a startup would be worth it.

PS: This one just submitted his thesis and is considering starting a company.