What is innovation?
Mar 18, 2024
“Unions are a better invention than chatGPT”
-Joan Westenberg
I ran across this hand-wringing WSJ article recently which bemoans the European proclivity to enact regulations that protect people’s privacy, their health, and their way of life at the expense of being able to compete with China and the U.S. in technological ‘innovation’. Similarly, here’s the Economist congratulating San Francisco for having ‘staged a surprising comeback’ based on all the AI startups in the city, while the basic infrastructure of the city remains broken, allowing only the rich a decent quality of life.
I guess it’s not surprising that publications like the Wall Street Journal and the Economist consider technological innovation to be, ipso facto, good. I guess most people would not be surprised if I, a technology director, also thought of technological innovation as an ipso facto good. Most tech people I know are not luddites–they’re gearheads, gamers, IOT enthusiasts, AI early adopters. I am a bit of a luddite though–a luddite in the original sense who looks at technological innovations and asks: Is this going to make life better, people happier, the world a better place? Is it going to take away meaningful work? Leave people grasping for handholds in a society that has bypassed their skills? And, as the pace of innovation has increased (most recently with the AI revolution), I’ve found myself wondering more and more what it’s all good for (aside from putting more yacht money in the pockets of billionaires).
Isn’t innovation and progress supposed to make our lives better? What does it mean to have a better life? Does it mean having more money? More things? Higher economic output? Greater ‘productivity’ (code for putting more money in the owners’ pockets)? More electronic toys? More access to media? Or does it mean having more leisure, deeper relationships with others, a feeling that you are a useful member of your community, health, well-being, mental clarity…happiness?
In another recent news item, we have Jensen Huang, CEO of Nvidia, apparently saying that kids shouldn’t study computer science anymore because AI will be doing all the programming for us from now on.
This brand of technology triumphalism is, as my students say, ‘cringe’. Should we also stop teaching math? Computers can absolutely do all the math for us. I decided to teach an elective this year called Open FabLab. The idea is that students can come to our maker space and create whatever they want. We have laser cutters, 3D printers, sewing machines, cardboard and hot glue guns, and a full wood shop. The most popular equipment so far? Saws, drills, and impact drivers. These kids are hungry to create things with their hands.
Most kids, let alone adults, in industrialized societies don’t make stuff with their hands anymore. Has that desire disappeared? No, it’s still there. Its an essential human trait. We like to use tools to make stuff. A computer is also a tool for making stuff. Right now, I’m teaching a video game programming unit to my 5th graders. Will any of them grow up to become game developers? Maybe. But it doesn’t really matter. What matters is the joy they show when they conquer a difficult challenge and see their game work for the first time. The love it. It’s flexing their creative muscles while also teaching them how to think logically through a problem and adapt their thinking when their first idea doesn’t work out. Whether they become programmers or not, those are useful skills. And whether the skills are useful or not, the joy is. Joy is useful. Its a worthy goal.
Corporate leaders like Jensen Huang don’t have the best interests of people as their baseline goal. Their baseline goal is the same as the goal of the factory owners the original luddites fought against: remove humans from the production equation. Or, if you have to have humans involved, make their contribution as unskilled as possible so you don’t have to pay them very much.
When tech titans tell us that VR, or crypto, or AI is the next big thing that is going to change the world and solve all of our problems, let’s remember that they own the infrastructure that makes those things possible. Five corporations own over 70% of the world’s cloud infrastructure (servers, GPUs, data centers, fiber optic cables, etc.). Imagine if five companies owned 70% of all the roads in the world and could charge us for using them. What would they be trying to sell us? Flying cars? No. They would do everything possible to crush any company selling flying cars. They would be telling us that the next big thing was more cars and trucks, bigger cars and trucks. That’s exactly what they’re doing with AI. AI requires massive amounts of data processing. The owners of all that capacity don’t care if what they’re selling is bad for people and bad for society as long as they are able to leverage their monopoly on infrastructure to make money.
This is not to say that there are no legitimately beneficial uses for AI. Scientists have used AI to invent new medical treatments, develop mathematical proofs, and explore the workings of time and space. Those applications are cool. Using AI to generate images (instead of hiring a skilled artist), write a textbook or newspaper article (instead of hiring a skiller writer), or write code (instead of hiring skilled programmers) is less cool. This is especially problematic given that these AI models have largely been trained on the copyrighted works of the artists, writers, and programmers they are displacing.
I realize that I can’t bury my head in the sand and refuse to teach my students about AI tools. But, I’m definitely going to center these issues. I want my students to understand exactly what AI is, where the training data came from, what inherent biases it might have, and what the ecological and social impacts are. I want them to think hard about what the actual goal of ‘progress’ and ‘innovation’ ought to be.