Code is a new form of government: Clive Thompson

I wrote Coders to give people a glimpse into a hidden world that’s incredibly important to their lives. We all use software all day long — it shapes the way we learn about news, shop, play, and talk to our friends and family. So programmers, particularly those who found new software firms, have a huge impact on the way we live. Their priorities become our priorities. If they decide to make it easier to do something — like share photos — then they create Instagram, and pretty soon photo-sharing is a massive new everyday activity.

So, I wanted to help people understand who coders are, and what makes them tick.

Great question — no one’s ever asked me this. I think the most important turning point in the evolution of coding was probably the release of Netscape, the first browser aimed at everyday people. Netscape, more than any other piece of code, made the Internet seem exciting, accessible and kind of cool to the mainstream. Before that, the Internet was nerdy, weird, difficult to access, and didn’t have much on it except academic data. After? It transformed into the culture and politics and business of everyday life.

Another great question that no one’s asked me yet! I think the next phase is going to be the rise of “no code” and “low code” programming — tools for creating little apps and programs for yourself that don’t require someone to know much coding. I’m seeing this all over the place. In the world of entrepreneurship, I’m seeing people setting up businesses where they create their website and billing and invoicing systems entirely by themselves, using tools like Zapier or Bubble. In the industrial world, you’re seeing a new generation of robotics where employees program the robot using simple, visual drag-and-drop commands. It’s very interesting.

This won’t replace traditional coding. We’ll always need people who program complex, new ways of doing stuff. But it’ll greatly expand the number of people who do little bits of programming — automating things in their everyday life.

This is probably impossible to say. So many things we do — playing games, buying stuff online, understanding news — are now brokered by code that it’s tricky to say what the most important code is, because it requires you to say: What’s the most important human activity?

One possible candidate, I’d argue, is Facebook’s News Feed. When they turned it on in 2006, it propelled a huge shift in how we learn about the world, and the lives of our friends and families. Their algorithm for determining what we see has enormous, world-shaping power.

That’s why I made that story the first one in my book. There wasn’t a single coder responsible for it; it was a team of four or five. But I’d say it’s one of the most consequential pieces of software in the last 20 years.

In certain parts of the industry, it certainly is. In any of the arenas where venture capital is handed out to kick-start companies, there’s often still a strong reverence for the idea of the “10X” coder, the brusque obsessive who is rude to everyone because they’re constantly mentally pondering their code. It’s a stereotype that tends to favour only the most obvious, straight-out-of-central-casting young-guy nerds. And certain new fields, like AI and blockchain, are mostly a sea of guys.

But in other parts of the industry, that’s less true — the world of front-end engineering, for example, which is getting more complex and important all the time, appears to employ a much more demographically-broad base of people. The same goes for data science.

Are the self-declared alpha-male coders harder workers or smarter than the others? No, not at all. They might even be less hard-working and dimmer, frankly, because some of the mystique they cultivate is about being prickly and unwilling to work well with others. Yet, nearly every piece of serious code these days requires many programmers and designers and product leads working all together. If you can’t work really well with a team, you’re going to produce lousier code that’s harder to maintain.

No, it isn’t. It’s not like basic literacy, or numeracy. You can have a rich and productive life without knowing anything about coding.

That said, it’s enormously useful — and quite fun — to know a little bit of coding. If you dabble a bit with some of these low-code/no-code tools, or spend a few evenings poking around with tutorials on a relatively easy-to-learn language like Python, you can often hit upon a way to automate some of your slow, repetitive tasks at work. I know tons of people who aren’t coders, but who have created simple little scripts on their computers to help autoformat documents, or clean up spreadsheets — saving themselves hours of work a week. It’s incredibly satisfying, and makes you feel like you have superpowers!

We’re not paranoid! In fact, we’re probably not worrying enough.

That said, there’s been more and more scrutiny on the role that automated systems and algorithms have on our daily lives. We’re seeing a bit more discussion of that in everyday news and people’s conversations.

There are layers of problems we have to deal with here. Some of it is that algorithms are used to make split-second decisions in ways that take humans out of the loop, and by removing human judgment, remove some possibilities for fairness (“can’t you reconsider this decision?”) and accountability (“explain to us why you made that decision”). Sometimes the problem is that the algorithm is inscrutable; it’s the product of a long process of training a neural net, and nobody really knows how it works any more. Sometimes it’s not even really about the algorithm — it’s just that the company deploying it has a monopoly position, and has too much power, so any fillip of its algorithmic judgment has far too much impact on society. Facebook’s News Feed is inscrutable and lacks an appeal to fairness, but this wouldn’t be so much of a problem if Facebook were one of, say, seven major social networks competing with one another. Then, if people didn’t like the effects of Facebook’s sorting mechanisms, they could go somewhere else — and it’d put pressure on Facebook to change. So sometimes the problem is just monopolistic power.

Not really; it’s a new form of government, though. Regular government — either formal (an elected democracy) or informal (the protocols of a self-governing group of friends who play football on the weekends) — is all about rules. Code is just rules. When you use it to manage how people socialise with one another, it becomes very much like the government.

I think the identity of people is less important to know than, say, the code itself. In “open source” software, people openly publish the code, so that anyone can scrutinise it and see how it works. When done well, it can build trust in a wonderful way. I’m very much in favor of far more of our everyday code being open-source.

One interesting lever we, as citizens, have here is the purchasing power of governments. Governments buy a ton of important software, and we should require that as much of it as possible be open source. It would help tip the entire industry towards embracing that as a norm.

It’s changing a little bit for the better, but the progress is uneven. There are areas of coding where many more women work — for example, in front-end engineering and data science. But there are other emergent areas that look like the all-male self-styled priesthoods of the past — AI and blockchain, for example.

Part of what makes it hard for women in coding jobs is that so many things, over a period of decades, went wrong. Coding started out as a much more genuine meritocracy in the 1950s and 60s, and so women were — even back when society was much more openly sexist than it is today — suprisingly welcome in programming jobs. Then over the 1960s and 70s, coding became more central to big companies, something where they wanted to hire coders who might ascend to management; and since back then they virtually never let women ascend to managment, they stopped hiring women as actively, and rarely promoted them.

In the early 1980s, coding became something kids could do at home, but family norms were — not universally, but most often — that parents urged boys to embrace the mechanical, tech stuff, and frowned on girls who did the same. By the 1990s computer-science programs at universities were actively selecting for the teenage herds with home coding experience, who were, by that point, mostly boys. And by the early 2000s you’ve got companies that have seen two or three generations of management where women in coding jobs were a rarity.

So that’s like five or six huge things that all went awry over multiple decades. Changing and fixing the situation requires us to try and fix five or six big areas. There’s no simple, quick fix.

Technology writing has, I think, gotten much better in the last five or six years! From the 1980s to the early 2010s, a majority of technology writing — too much, I’d say — was basically just product reviews of software and gadgets. Comparatively, few writers considered the big social, cultural and political effects of technology — which to me, is the most interesting part of technology!

Today, that’s changed. There are many, many more writers diving and producing terrific, thoughtful examinations of how technology shapes society. I think it’s great.

They’re quite diverse politically, culturally, and socially.

But most of them do hold one value very strongly — they love efficiency and automation. They love taking some human activity that’s repetitive and writing code to do it automatically. They love it so much that many of them recoil, in an almost aesthetic way, from inefficiency.

This is part of why code has given us so many gifts, because coders are constantly creating ways to save us from dull, boring, repetitive toil. But it’s also a value that can go awry if we — and they — don’t watch out. Some of the civic problems we’re currently seeing stem from code speeding up the pace of economic or social activity to such a degree that it produces weird, hard-to-handle side-effects.

COMMENT NOW