Changing the Demographics of Innovation

September 16, 2010 – 2:56 pm

Text of notes for a talk given at the 50th Anniversary Conference of the New Zealand Computer Society in Rotorua, 17 September 2010. I will link to video when it’s posted by the conference organizers.

Hello everyone. Thank you for the kind introduction, and thank you to the New Zealand Computer Society for having me here.

Let’s get nostalgic for a bit. It’s the 50th anniversary, we can afford to be nostalgic a little.

Whose first computer was a mainframe? Whose was a mini? Whose first experience was through punched cards? Who had a microcomputer, like a BBC Micro or Spectrum? Whose first was a PC?

My first computer was a Commodore 64. 64K of RAM, not enough colours, great programmable audio, built-in BASIC …. The reset switch was a paperclip across two of the terminals in the exposed cartridge port. That was where I first played text adventure games …. Loved it.

You’ve probably got warm feelings about your first computer too. It’s like a first love, but you never break up. It won’t burn your t-shirts and cut your face out of photos. It might not boot up any more, but who here doesn’t have that problem sometimes?

My C64 lead to a PC and then into Victoria University’s Computer Science degree starting in 1990, where I got to explore a VMS box and then a shiny new SGI Unix machine, and to download Linux (my verdict was “what a pain to get installed!”).

In my time at Vic, on those VMS and Unix boxes, I got to play with the Internet at a time when most people hadn’t heard of it. You young folks here, there was a time when most people hadn’t heard of the Internet. We had text games to telnet into, programs to download, and Usenet to post messages to.

I was one of the first New Zealand users of the web when it was released, and I successfully convinced the IT department to let me build the shiny new “Campus Wide Information System” on the web instead of on gopher. Lots of fun, discovering all sorts of things along the way:
– the anonymity of the net is great–nobody knows you’re a nineteen year-old from New Zealand, not even Tim Berners-Lee and Marc Andreessen.
– you do not want to be the sole conduit for an entire organization’s web site
– but at the same time, it’s hellish to teach people how to put things onto the web (particularly if “export from Microsoft Word using hard-coded styles” is your markup technology)
– no matter how much effort you’re poured into your website, no matter how many new things you’ve done, it will always be described as “shit”

Fortunately this has all changed. *cough*

From there I went to America, worked for a startup, wrote a book (“The Perl Cookbook”), got married (on a MUD), helped along the Perl project, and worked for O’Reilly Media. They make those books with animals on the covers. Their basic business is talk to a lot of people, find the weak signals that show something big is coming, then have books and conferences ready when the masses discover the trend. They did this with the web, with open source, with web services, with web 2.0, and now with the Gov 2.0 open government work.

At O’Reilly, I was running some of their conferences. The bit I still love most about conferences is that moment when, for the first time, a group forms–people who didn’t realize they had something in common get together.

And this is where that “on the Internet, nobody knows you’re a dog” cartoon comes out. Yes, on the net nobody knows you’re a dog. But when you get everyone together in person at a conference, as you do here, … well, to lapse into lolcat: different dog is different.

Let me explicit about this: we’re at a sausagefest. And this is not a reflection of the fine people at NZCS. Our industry is a sausagefest, far more men than women. Women are between 10-30% of CS classes at university, depending on who you talk to and what level you’re measuring. But don’t worry, if you’re in the software business then you’re alright: in open source, it’s 1.5%.

Right now, this can go two ways. One way is to ask why this is so, who’s to blame. This is the path where we all get defensive, fold our arms, and say “stop attacking me, I’m not an asshole, this isn’t my fault, I didn’t set out to build or perpetuate a misogynist class structure, this is bullshit, women can choose and they choose not to work in computing, I don’t see there’s a problem at all, Torkington’s a jerk, what a disaster”.

I don’t like the path of blame. Blame is an attack, and attacks provoke defense, and defense short-circuits rational thought and the fight-or-flight reflex kicks in, and we’re not any better off.

So rather than have you all thinking about blame and looking for victims and oppressors, I want you to forget that crap. That’s solving a headache with an AK-47. Let’s agree not to see this as a finger-pointing battle.

I’m going to let you in on two related secrets: gender is just the tip of the iceberg, and this is a *huge* fucking opportunity.

Any of you know the Myers-Briggs tests? I think they’re science-infused astrology but there are plenty of people who swear by them. And if you go to your developers and you give them the Myers-Briggs test, you will find something odd: they’re almost all INTJ or INTP. Those two categories account for 5% of the general population, but nearly 100% of the computer programming crowd. And, it just happens, most INTJ and INTPs are men.

There are great efforts afoot to recruit more women into computer science classes at university. To do that, though, you really need to get girls into computing in high school. Given how grunty teenage boys can be, you almost have to get the girls into computing in primary school (back when they tend to have better textual and conceptual skills than boys) so they’ll do it at high school.

But there are some interesting signs that the things you have to do to get women into computing is how you get more *people* into computing. That is, the things that drive away women are driving away shiploads of men too. Let me say this clearly:

If we change our culture so we attract more *people*, we’ll get more men and more women coming in.

In open source software we like to pat ourselves on the back that we’ve built a meritocracy. We hide behind this, pretending it’s impersonal. The best coders thrive, the weak wither away, and it’s as Darwin intended. This is often treated as license to be an asshole. The culture this builds is poisonous to those who don’t look, sound, or act like the folks who are already programmers. I can see the ComputerWorld journalist scribbling away furiously, so I want to make it clear that I’m not saying all programmers, or even all open source programmers are assholes, I’m exaggerating for effect.

But many programmers and many projects and many teachers and many classes aren’t welcoming to beginner programmers. This isn’t universally the case–there are projects showing us how to do it.

One of my favorite examples of this is Dreamwidth. It’s an open source project for blogging software, a fork of LiveJournal, serving over 200,000 users. Thirty of their forty developers are women. They do this by being friendly: they encourage new contributors (“little devils”). They work hard at being welcoming, they don’t assume people will feel confident enough to contribute and take the usual curt feedback.

Deep down, I had always assumed coding required this kind of special aptitude, something that I just didn’t have and never would. It lost its forbidding mystique when I learned that people I had assumed to be super-coders (surely born with keyboard attached!) had only started training a year ago. People without any prior experience! Women! Like me! Jesus! It’s like a barrier broke down in my mind.

But very few of these Dreamwidth developers are typical programmers. By changing the culture, they attracted more *people* to being programmers. And this is our huge opportunity: to get more people into programming.

I volunteered in our local school, teaching 8 year old kids how to program. I got asked “why!?” Parents thought I was turning their kids into 8 year old Bill Gates, not the rich Bill Gates, just the dorky and deeply unsocial Bill Gates.

I think it’s important because the jobs most folks want to have are the jobs that work with information. It’s like driving a car these days–you gotta know how to change the oil, fill the tank, top up the fluids, change a tire. If you don’t, you’ll be hugely less efficient in how you use that car. Same with a computer: all very well to shuffle information in spreadsheets and reports, but if you spend a day doing something that would have taken 5m to code and a second to run … you’re wasting human potential.

How do you teach kids to program? Easy. Use a system called scratch. scratch.mit.edu — very simple to use, drag and drop actions and loops and so on, the pieces connect like Lego, you can’t have a syntax error. I love it, and the kids did too — they built games, stories, toys, demos, even a book review, and girls did better than boys.

So it’s not that we don’t have programming tools suitable for 8 year olds, it’s not that girls can’t do it, it’s not that they don’t want to do it. We need to do a few things:
– teach the teachers,
– change the culture so we don’t drive them off when they’re ready to program socially,
– make them love it.

Loving the field is more than knowing how program. It’s the difference between a day job and a passion. Some folks are motivated by abstract problems, intellectual challenges. These are the people who have historically populated our industry.

But we’re realizing now that these people aren’t the only people we need. Silicon Valley is littered with the smoldering wrecks of companies founded by technical people. For every Google, there’s a thousand anonymous failures.

Because we don’t distinguish clearly between technology and industry. We treat the two as the same, but they’re not. And this comes to the heart of the topic today, innovation.

Invention is not innovation. Invention is discovering something new, doing something for the first time. It’s the thrill of exploration, of research, of the unknown. It’s technical. It’s scientific. It’s often solitary.

It’s not innovation.

Innovation is building a new business or product line. It’s sales, marketing, channels, cash-flow management, interaction design, user experience, iteration on product, new markets, making the first sale, ….

Research is proving something can be done, once, no matter how long it takes. Business is doing something that people want, consistently and repeatedly, for profit. These are very different.

We need people with these skills in our business. In our industry.

Paul Graham, early Lisp guy, wrote the code that became Yahoo! Stores, now runs an incubator called the Y Combinator, he was asked the secret to a good career. He said you have two paths to greatness: you pick one topic and go deep, drilling down into the details until you’re the best; or you pick two topics and occupy the intersection. There are many more intersections, and there’s less competition so it’s easier to be great at an intersection.

How do we get people into the industry. There’s no single magic bullet–if there were, we’d have found it by now. I think we need to do two things better:

1) Talk about the exciting big possibilities in our field. Google Goggles, when everyone goes “wow!” … I saw a device for shipping audio lessons around famers in Africa, where you can record your own advice and share the lessons by sneakernet. Changing people’s lives, making science fiction real, that’s the exciting stuff.

2) Offer people the turning points that let them turn towards computers. We’ve built a great culture that turns them *away*, let’s get them in.

I reached out to some of my friends from my time in America, people who had become successful. What were the opportunities that they had that got them into the field and let them be successful?

There were themes. First, it was getting your first computer. Full-time immersion in the tool makes you better at it. Part-time, shared access … less so.

Craig Nevill-Manning clipped garlic when he was 14 to get his computer.

Chad Dickerson worked at a small regional newspaper by night, but his employer had an amazing scheme where every employee was allowed to buy a computer with the company discount, all interest-free. He was making $7/h and was able to buy a $3500+ computer.

Brenda Wallace, a great Wellington coder, got her start when her Dad went to Australia in 1980 and came back with a TRS80.

Caterina Fake, cofounder of Flickr, also had a TRS80 when she was 10. She didn’t start off going down the computer industry path, but when her life made that possible in 1995 (“I’m broke, I’m in San Francisco, I’ve seen the Internet and know my way around programming computers, it’s ridiculously easy to get a job as a web developer”) she taught herself, entered the industry, and was away.

This is why I love projects like OLPC that get laptops into the hands of kids. The good folks at Point England School are inspirational: they’re working on a project to get laptops and connectivity into the hands of their students and their community. They know it’s not just hardware now, it’s connectivity that matters too.

Second, it was someone taking a chance on them, believing in them.

Danah Boyd, who is now knows more about social use of computers by teens than anybody else, and has a PhD in the topic to boot, she had a great mentor at Brown University, someone who talked her off the quitting ledge, steered her into computer science and humanities.

Gina Bianchini, who was CEO of Ning and of other companies, acknowledged her two cofounders: Mark Kvamme and then Marc Andreessen.

Mitchell Baker, who runs Mozilla and was the “chief lizard wrangler” who got the source code to Netscape out of AOL and into the hands of developers who could eventually build Firefox … Jim Barksdale who took a risk. “I think he knew I had a bunch of capabilities, and Mozilla was a quirky thing that i would fit with”

Andy Baio, who sold Upcoming to Yahoo! and whose blog is hugely influential and who was CTO of Kickstarter,

2000: With *very* limited experience, a Slashdot-loving geek took a chance on me and I was hired to code Perl at a web design firm in L.A, changing the trajectory of my career forever.  In the next week, I taught myself a crash course with a copy of the Perl Cookbook and the Camel book, and fell in love.

Danese Cooper, the CTO for the Wikimedia Foundation which runs Wikipedia, gave props to the person who offered me a job at Apple back when you could get “on-the-job” training to transition from “French Major” to “Technologist”

This is giving someone a go because you see something in them. When I organized conferences, I organized sausagefests because I chose people whom I knew were good. Which meant my pool was limited to those whom I knew. Which meant, because we’re all like this, they were people like me. When I decided I wanted to change that, I brought women into the program committee, trusted their recommendations, and eventually turned the chair of the conference over to a woman.

There was talk yesterday about internships. I love the Summer of Tech they run in Wellington. The numbers are amazing: 2/3 of students go on to jobs at the company they interned with. The common thread between internships, computers for kids, talking about the good things of our industry, and giving people opportunities is to create on-ramps.

I challenge you to think about how you personally can create on-ramps.

Can you volunteer to teach programming or entrepreneurship at a local school? I’m happy to tell you what I did, it’s dead simple and worked well. There are lots of other successful programs you can emulate.

Can you set up a cheap hardware plan for your school, maybe identify the cheap netbooks, loan plan, etc.?

Can you change how you respond in your workplace to newbies and silly questions?

Can you be conscious of your human unconscious bias towards people like you, and you reach out to make change–ask someone who’s not like you for recommendations or advice?

Can you take a chance on someone?

Can you build a mentoring and supportive environment for people who would ordinarily be bounced off?

There are things we can do now to change the demographics of ICT innovation in New Zealand. Some of those things are the big national programs that are hard work and take a long time. But many of them start with individuals. We can give many more people the same experience of their first computer, and a rich and rewarding career in our industry.

Gandhi: be the change in the world you want to see.

[thanks to Brenda, Skud, Val, and Allison for their patient and selfless work with me over the years. Anything good in my talk will have come from them, anything bogus will be mine alone.]

  1. 4 Responses to “Changing the Demographics of Innovation”

  2. Sigh, first computer was an Apple IIe. Avatar was the game, but text adventures are the best.

    Nice talk!

    By Jeremy on Sep 17, 2010

  3. Fascinating talk Nat. I had no idea that the gender gap was as wide as that. & You got married on a MUD? This is something I will need to learn more about :)

    By althecat on Sep 17, 2010

  4. Hi Nat – this is my first time visiting your site (was pointed to this post by a friend), but I am so happy to have read this & had to let you know how much I appreciate you sharing it. There were several points while I was reading it where I found myself agreeing with you out loud. :)

    To your comments about being open to beginner programmers, I heard Kaliya Hamlin of She’s Geeky talk recently about how important it was that they reached out specifically to women who not only wanted to be programmers, but to those who didn’t – who might for example work *with* Ruby developers and just want to know enough to be able to communicate more effectively with them. Lowering those kinds of barriers to entry has a significant effect on the diversity of attendees, as you might imagine.

    Thanks for a very thoughtful (and funny) post!

    By Lauren on Sep 18, 2010

  5. Encouraging open thinking is an awesome to do. Thank you for sharing these examples. As the father of a 9 year old daughter who has been online since she was 3 I do think a great deal about how we use technology everyday.

    Ironically open source should lead to more open thinking and more diversity but there is a kind of social vacuum around much of the programming community.

    At my local school (Westmere Primary) I’m proud of the number of junior bloggers and thinkers who are showing up the teachers for the most part.

    I’m a big fan of Paul Graham but I do note that the ycombinator approach of locking yourself to a computer (around the clock) for a few months and eating only noodles is maybe not the kind of social experience that appeals to a wide group.

    Maybe that is a start-up dynamic but I do think we need to look at the social side of making great workplaces attractive to a more diverse group of people.

    One of my favourite recruiting ideas is check musical background and the ideal company would also have a great house band.

    Maybe not so practical but somehow the relationship between music , math and programming is actually very close.

    I should also say that I studied music as my 6th subject at high school and played in an orchestra for 5 years.

    Over the past 20 ish years I have worked with programmers of many shades starting on Dos and Perl way back when but my reference point to all of this was really musical.

    I’m sure there are other ways to encourage diversity in the programming community but personally like the musical route.

    Thanks again for your thoughtful post.

    By dialogcrm on Jan 26, 2011

You must be logged in to post a comment.