In the later part of the interview she asks some questions about the recent discussions regarding minorities in the startup community. I thought his response was insightful:
"Well, I think the problem is upstream from us. I don't think it's that, like, huge numbers of women and minorities want to start startups and that we're filtering them out because they're not white or men or something like that. I think that the applicant pool, the applicant pool has the same problem that people see in our output, right? The same problem is in our input that people see in our output. The problem is further upstream. The problem is that the pool of startup founders is the people who are messing about with computers at age 13. If you want to fix the problem, that's what you have to change."
It took him some time to come up with the right response and you could feel he was "computing" the right answer in his head. When he finally found it, it just felt a very natural conclusion. Paul defused the race bomb question with great calm and clear thinking.
After the incident involving Arrington, it'd be surprised if PG had not thought about the issue in that context and had reached some conclusion at that point --even if that conclusion was simply an affirmation of his experience and what amounts to his anecdotal evidence in the applicant pool.
That's how he responds to all q's. He is so smart and so well thought out that you can literally see him thinking through the question to be able to give a clear thoughtful response. Whenever I see him doing an interview I always think he is operating on an another level.
Some other incubators fund more women and minorities than YC, so this seems like it's only part of the story. A couple questions this leaves in my mind:
- what are they doing to broaden the applicant pool?
- have they considered that something about the process or their reputation might lead to women and minorities steering away from YC?
Probably the biggest things we've done to broaden the applicant pool are Hacker News, Startup School, and all the essays I've written about startups.
I don't think there's anything about our process or reputation that directly discourages people of any gender or race. But we do prefer founders who are hackers, which presumably thus causes fewer members of groups that are underrepresented among hackers to apply to YC.
It's definitely a complicated issue with no clear overriding cause. As a minority who did take apart his mothers computer at 13 and compressed the windows partition to install linux (woo hoo slackware) and has gone on to found a startup I think some of Paul's commentary is spot on. Though I was representative of some of the larger stats (single parent home, lower middle income specturm) I was also one of two kids on my block with access to a computer in the home, and my single parent was a teacher who made education a priority. If you looked at the cohort of kids I grew up with you'd find myself and 2 others out of 11 or so that are not either a. dead or b. in jail, or have been in jail.
Why is that? I don't know and don't claim to know, but the road to changing it starts very young.
what are they doing to broaden the applicant pool?
And why you want them to do so? I personally want that YC or any other program select applicants based on their skills and abilities and not any other criteria.
I think broadening the applicant pool here means getting more underrepresented types applying rather than changing how you select them after they've applied. Just because you get additional, different people to apply does not mean you have to relax your admissions criteria; if anything, it's the opposite: you have more to choose from.
If this is not what was meant by that phrase, then it's what should have been meant and I stand by my point anyhow.
I think everyone wants YC to pick the best candidates (thus making the best startups who make the best products for us to use), but I also want them to not systematically overlook qualified candidates due to external factors. If there's a way to tweak early-childhood education to turn more women into potential startup founders, hackers, etc., that increases the number of startups and hackers available for startups, which is awesome.
It would be reasonable for YC to use successful women/minority applicants (Jessica Mah; Leah Culver; I'm sure there are others, plus of course Jessica Livingston's role as a partner) to demonstrate that women can be successful in startups. Other than that, it's really something for parents, K12 educators, etc. to address.
Women do seem to be a lot better represented in bio/biotech than in computers/ee, so if over time more startups happen in the bio space, that should help to address the imbalance too.
I am sorry but PG is wrong and blindsided by experiences rather than knowledge.
Look at developing countries, their shares of minorities doing start-ups is at an all-time high.
No, in fact the USA problem is directly tied to housing debt and EDU debt being too high in a specific demographic age group that happens to be the talent and the co-founder pool for start-ups in the USA. Whereas if is compared to small business starts per year you see a shift towards more minority business starts bu not a huge increase.
* The share of minorities doing startups being at an all-time high in developing countries -- are you referring to people who would be minorities here in the US but are majority in their own countries? How does that refute PG's point about the problem in the US being "upstream" of YC?
* I don't get at all what you're trying to say in your third paragraph, about the USA problem being directly tied to housing debt, etc. -- again, how does that refute PG's point?
I really enjoyed the interview but I found this exchange puzzling:
Emily Chang (quoting Max Levchin): ‘Entrepreneurs aren’t taking big enough risks … ’
Paul Graham: ‘Well, that probably isn’t true because it’s gotten cheaper and cheaper and cheaper to start a startup and when the cost of failing gets lower usually people can do riskier things’
Risk is the potential for loss. If the cost of starting a startup is lower then the potential for loss is lower which means the risk is also lower. More people are starting startups because it is less risky, not more. That means the number of companies being formed is going up; in turn, that means that more people are taking the risk of starting a company, whether they are risking much or not. This might seem a subtle distinction but I think it is an important one.
A week ago I sat down opposite an entrepreneur who built a 24-hour transaction layer on top of the UK’s banking system which runs on nightly batch-processing. It made the transition from 9-3 banking to 24-hour online banking possible. He started out with a really big vision and had to take really big risks and the result was a really big company that created a lot of value economically and socially. It took him two decades and nearly bankrupted him on dozens of occasions. His vision sustained him. It kept him going through all the terrible lows and in the end he shook up one of the slowest-moving industries out there. That was a big risk. That was a big reward.
"it's gotten cheaper and cheaper and cheaper to start a startup"
"maybe it's just not the right time yet for a vacation on the moon"
There is a terrible amount of segment bias here. It is orders of magnitude cheaper than it used to be to start a software or virtual service startup. Other varieties of tech startups (chip fabs, space launches, flying cars, etc.) aren't nearly as cheap yet and represent far greater risk. It's a big world. Lots of markets. Lots of possibilities.
Agreed, and if anything this makes it harder to aim for something like holidays on the moon.
If it costs the same to try and build a spaceship as it does to create the new social fad then why not reach for the stars (literally) - but when it's so much cheaper to try and start the next Facebook, and you know that starting the next Facebook could have just as much potential to make you far less people will take the riskier option.
You know, it's always a visceral pleasure to hear/watch pg talk; at least for the way my brain is (soft-)wired, he's so charming and listenable.
But every time I see him interviewed, my "he's got an axe to grind" detector is going off so loudly that it's starting to drown out the charisma. It's understandable, in that high-tension soundbyte-oriented environment, that pg wants to ensure he stays on-message, but it's also disappointing from someone who we know is capable of being so much more informative.
He says, in response to several questions, "oh, that's not how a perfectly rational market would work, so it's probably not what's happening". I mean, I agree with the argument as far as it goes, generally reality roughly approximates an ideal market, but the approximation is always rough, and pg repeating this is sort of wasted air, isn't it? Examples:
at 2:08 in response to "big enough risks",
starting at 2:38 in response to valuations like Dropbox's,
at 5:25 in response to Aston Motes's very plausible theory about bias in "pattern-recognition"
Everyone's entitled to try to manipulate the media as best they can, and I'm not even doubting the sincerity of pg's claims about the market's effect on all those issues. But it sure would be nice to hear you dig a little deeper, pg.
I don't understand the women+minority problem. Different people have different interests, and I believe this is a good thing and not a problem. If more diversity were to come to the startup world that would be awesome, but I don't see this as a problem that needs fixing.
Startup people just have to keep doing what they do and if more people get interested in startups then things will all fall into place. What about the percentages of women and minorities in such places as politics? Is this a problem too? What about other industries that are dominated by women, is there a problem there?
Eric Ries claims to have had "shocking" results when having recruiters remove all demographics from resumes. His suggestion for YC:
I suggest the following experiment: for your next batch of admissions, have half of your reviewers use a blind screening technique and the other half use your standard technique, on your first screen (before you’ve met any applicants). Compare the outputs of both selection processes. I predict they will show different demographics.
We see them at the application stage too, because the application includes a video. The video tells us a lot more than their race and sex. So if half the people reading applications didn't see the video, we'd do a lot worse at reading applications.
I do not think YC wants to do an experiment for the sake of experimenting because YC is an experiment in itself. Seems like many commenting are missing the his point. The pool of applicants seems to come from those who gain access to computers at an early age. Unfortunately, minorities don't get this early access to computing, therefore, very few applicants are minorities. This has little to do with race and color and more to do with the level of interest of the person, but again, this tails back to how early a person comes in contact with a computer.
They spent a long time talking about this, at the expense of more interesting questions she could have asked PG. Especially more interesting to people interested in startups.
It is a serious issue though, 96/4 men to women is a really bad ratio. And my guess is that the ratio of white and asian to hispanic and black is even worse. So startups miss out on a huge potential talent pool and the perspective of people who might approach problems in a different way. And at the same time women and minorities are largely failing to take advantage of a really exciting opportunity.
PG is probably right, the bulk of the problem is that thirteen year old girls / black kids are not playing with computers. So maybe this is a question that would be better directed at parents or teachers, but it definitely should be on the radar. It is a problem that can be solved, that we should solve, and is a good thing for people watching Bloomberg / people doing startups to think about.
"Well, I think the problem is upstream from us. I don't think it's that, like, huge numbers of women and minorities want to start startups and that we're filtering them out because they're not white or men or something like that. I think that the applicant pool, the applicant pool has the same problem that people see in our output, right? The same problem is in our input that people see in our output. The problem is further upstream. The problem is that the pool of startup founders is the people who are messing about with computers at age 13. If you want to fix the problem, that's what you have to change."