This month, Paul Graham wrote an essay on reforming U.S. immigration. In a nutshell, he argues we must let more smart people into the country because our supply of smart people who can build software is tightly (almost absurdly) constrained. He notes that simply going by raw numbers, there will always be more smart brains born outside our country than inside it.
Graham is a thoughtful guy, and his essays often spur much discussion and critique. One critique of this essay regards the assertion, common in our industry, that there are a limited number of elite programmers worth dramatically more than "normal" programmers. Not everyone agrees that the so-called "10x" or "100x" programmers exist. I've never seen the "x" defined, for example, which makes it a rather hand-wavy kind of discussion. Placing a numerical multiplier on this undefined measurement may be more truthy than truthful.
The other critique of Graham's essay regards how it seems to gloss over the untapped talent already in the country. For those Americans fighting systemic barriers, it can feel like a slap in the face to call for more immigration without mentioning how much we'd benefit from dismantling the walls that block out the smart people who already live here.
When it comes to a question of priorities, I like the way Ellen Chisa put it:
Finally read this: http://t.co/Ktt2rAdIJ1. We should let more technical talent immigrate. We should also train more people. Not either/or.— Ellen Chisa (@ellenchisa) December 29, 2014
None of this, however, is what led me to write this post.
In Which I Get to my Actual Point
Paul Graham founded the famous Y Combinator, which funds and guides early stage startups. He recently handed the reigns of Y Combinator over to Sam Altman, who seems from my casual and occassional observation on Twitter to be a very nice, very smart, caring sort of fellow who wants to build things well and thoughtfully. Demographically, Sam's a lot like me: a young, highly educated, white guy running a technology company.
As the discussion of immigration bubbled around Twitter this month, Altman offered up the following observation:
also, when people say "we need to train more coders domestically": while that's true, anyone who wants can teach themselves to code online— Sam Altman (@sama) December 30, 2014
This is the one. This one makes me scrunch up my face and take a deep breath and not know quite where to start.
I don't say that with malice toward Sam. I don't say it as a prelude to an attack on him. I say it because I think I recognize the blind spot, because I've had a similar one myself, and because I don't doubt I remain blind to parts of myself that would make me cringe if I could see them. I say it because to the degree I can see the disconnect between that statement — from a smart, caring, and well-intentioned guy similar to me — and the people it leaves out, I see the subtlety and awkwardness of the work we have before us.
Many folks responded to Altman, with observations such as this one from Iheanyi Ekechukwu, which led to an elaboration on Altman's part:
@kwuchu what about free access at libraries? I taught someone to code there once, it worked pretty well.— Sam Altman (@sama) December 31, 2014
@kwuchu the kid I taught walked a few miles to meet me there. not saying it's not incredibly hard and unjust, just saying it's often doable.— Sam Altman (@sama) December 31, 2014
On the Definition of "Doable"
I'm a dad of two young girls, ages 3 and 1. I rarely get through a day in which I don't wonder, with great sincerity, how single parents get through their lives. It's a question I had prior to being a dad, but it was vague and formless. A sort of intellectual recognition of a hazy, distant thing, without the specifics that might give it color, taste, sound, or feeling.
My perception of how "doable" it is to raise kids as a single parent is, of course, very different now than it was before being a dad.
My experience living the privileged life of a middle class white man has limited the number of direct experiences like this I can draw on to understand what it means to live under other circumstances. It doesn't make me bad, but it does limit my understanding.
Living in Baltimore has, to a degree, provided a few more perspectives. When someone says "anyone who wants can teach themselves to code", I can step out my door and gut-check what that means for my neighbors. For example, I had about 30,000 homeless neighbors this year. Can they teach themselves to code? What about the 1 in 4 neighbors living in poverty ($11,500 for a single person and $23,000 for a family of four)? Baltimore's cost of living is no San Francisco, but you don't want to live off an income like that. What about the people just a smidge above the official poverty line? A few years ago the national average was 1 in 3 people living just 50 percent above that poverty line. It's unlikely my neighbors here in Baltimore brought that average down.
Can people spending their energy to scrape together a living spare enough energy to learn to code?
What about access to the Internet, as Iheanyi observed? In Baltimore, somewhere between 20 and 40 percent of our 600,000+ residents don't have access to the Internet in their homes. I love me the central branch of the Baltimore Public Library, but even they can't provide regular Internet access to hundreds of thousands of people.
And these are all the easiest examples. This is the low hanging fruit of the counter-example gut-check. The simplest ways to see that "it's incredibly hard" doesn't sufficiently describe the extraordinary barriers that hundreds of thousands of people in just one U.S. city face if they get the fancy to teach themselves to code online. The more subtle systemic barriers of education, or family stability, or living in food deserts, or any of a great many other factors make it even less feasible. Under these conditions, the challenge transcends "hard" and passes over into "functionally impossible".
I don't have answers
I don't speak from a position of moral superiority here. The company I run was, until recently, all white men. Even now, at the time of writing, it's still all white. Contrast this with folks like my friend Jess Gartner, CEO of a Baltimore startup that is the same size as us but much more diverse. I need to do better just like many of the people like me need to do better.
As such, I don't write this to proclaim Altman a Bad Guy. I write this because I think it's important for folks like myself to help aid the slow-moving realization of the limits of our perspective.
When Altman writes:
when i was starting my startup i lived on instant ramen and coffee ice cream. today had coffee ice cream for first time in 8 yrs, wretched.— Sam Altman (@sama) December 30, 2014
I recognize myself. In 2002-2003, I was an apprentice at Actor's Theatre of Louisville. For ten months, I worked without pay. I lived off food stamps. I regularly had less than ten dollars in my bank account. We worked startup-like hours: all day, 6 days a week. Our only guaranteed days off were Christmas Eve and Christmas day. As it turned out, I had to work those too.
Was it hard? Yes. Very hard. Am I proud of that effort? Yes.
Was I fundamentally at risk? No. Living off food stamps with no job for ten months was extremely hard. But it was still privileged, in a way that all of my creative life has been. The safety net of my family spread out under me, and while I was exhausted, I was never scared.
Similarly, I am privileged in my ability to write publicly on this topic, without fear of alienating myself from the technology power structure. The company I founded was bootstrapped, and we don't want or need an assist from Y Combinator or the venture capitalists of Silicon Valley. I know several people in town who could not do the same without evaluating the risk that it might count as a mark against them in, say, a future application to YC.
Even my pride in bootstrapping this company is built on privilege. It wasn't really bootstrapped. Very very few of us are forced to build up the entire foundation of our lives. I mean, I've got nothing on my fellow Baltimore entrepreneur Chris Wilson.
The positive thing about all this is that once we recognize it we can work to dismantle the barriers. But we have to recognize it first. To get to "anyone who wants", we have to start by understanding how many are in want, and what kinds of things they want, which is a list that includes a great many things before we can talk usefully about wanting and fulfilling the knowledge of code.
For those trying to get past the barriers in our own back yard, who are ready (or nearly-ready) with the skills to contribute today (or soon), it's important to recognize them and then include them in our barrier-busting efforts.