Why Computer Science Needs Human Science
To build better technologies, break less things and create more human value, I argue that computer science needs a little more of the human sciences.
For decades, the closest human sciences have come to the world of digital products has been Human Factors Design/Engineering (HFE) and Human Computer Interface (HCI) design. And a juicy squeeze of sour psychology to figure out how dopamine works.
In this article I argue that by bringing more of the human sciences into the technology industry, we will end up breaking less things and making more human-centric things. And being more, well, human. In a brief, summary sort of way.
The field of computer science has grown tremendously. And if it wasn’t for HFE and HCI, many of the lovely software and digital hardware products we use today, wouldn’t exist, or at least not be very nice looking or functional. It is perhaps the starting point for the use-case of why the technology industry needs more human sciences involved.
HFE and HCI helped make some products very useful and functional for humans, such as the brilliant way that smart devices from phones to watches work. They even made tablets embedded in fridge doors work well. It’s just that nobody wants them, at least, not enough of a share of market to justify the continuance of the product.
Then along came the toothbrush with Artificial Intelligence (AI) in it. The product was pulled with quiet abandon a few months later. Or the ovens that have AI in them too and all kinds of features that you can use with your smartphone, not on the oven itself. If it can connect to WiFi and if the phone responds. In three netnographic studies I did for brands, less than 10% of the time did the app even work with the oven. User satisfaction was, to say the least, quite dismal.
There are UX designers and product teams that agonize and fret over the colour, size and placement of one tiny button in an app or on a website. Computer scientists, coders, who spend endless hours figuring out how to solve a problem just so. Sometimes they spend as many hours figuring out how to solve a problem making the code even more complex. It’s an odd paradox of the coder.
Behind it all lies a common thread, a uniter of almost all of these products, designs and implementations. Data. And there we have the problem. And a solution.
Less data, more human. Even Google, who was hardcore data all the way, everyday, learnt the limits of being data driven. While it is still a very data heavy driven decision engine, they also learnt the value of design. It was about the time we saw them start to apply design to their products.
Eventually, we saw the rise of UX research in product design and management. Somehow, the concept of conducting ethnographic research from the world of anthropology crept in. It was well intentioned and helped, but it was a diet Coke version of ethnography. By early 2023, digital products, software mostly, cut back on UX research. Talking to customers is expensive and takes too long. Many product teams largely scrapped or severely cut back on their UX research teams.
Along came Generative AI with Large Language Models and and multimodal creation. And the computational theory that everything humanity creates can be reduced to simple algorithms. Culture on a diet.
The tech bros mantra of move fast and break things became move even faster and break even more things. Do what the data says, not what the humans do. Data is valuable and is proven to be effective in developing digital products. Up to a point. Data tells you what happened. It does not tell you why. It quantifies, it does not qualify.
And this is where we bring the human sciences into play, beyond stuffing anthropologists, sociologists and psychologists into dark basement offices. Occasionally to be dragged into sterile meeting rooms when the coders hear and then dismiss the soft stuff. To many, not all, coders, moving slower is human and that stalls innovation. To many a tech startup and the Tech Giants, being more human means less and slower innovation. It is exactly the opposite.
Time and again, it has been shown that tighter regulations (but not over-regulating) and being more attentive to the human, leads to not just more innovations, but better ones. Somehow, we have entered a bizarre, warped zone of only being faster, trusting only in the data and being unbounded can possibly propel humanity forward. This approach hasn’t and isn’t, working.
This is reflected in how culture is pushing back against social media platforms. Banning phones in school, suing the platforms for harms. Governments rushing in with privacy and data governance laws. It’s why there is growing frustration with the purveyors of mis/disinformation.
The laws and regulations we apply in the real world that (mostly), work, we have lagged in applying to the digital world. Arguably the digital world needs them more, because that is where we tell stories and story telling is how we make sense of our realities.
The human sciences, through anthropology tells us the stories of what it means to have been and how we may be, human. Sociology tells us the stories about how we come together, work, play and live together. Psychology tells us the stories of how we think, why we do things and why we feel the things we do.
Wrap these into the world of software development, computer engineering and what a marvel of creations we could come up with. Tech startups and existing tech companies, would end up with moving steady, breaking far less things, including people’s minds and hearts and probably be more successful.
Apple understands this approach well. It is one of the most successful and wealthy technology comppanies ever. To this day, it has on its product teams anthropologists, sociologists and psychologists. Apple’s innovation model is to watch inventions and startups to see where and how they fail. Then to perhaps acquire them for the technology, or reinvent it and launch. When and how humans want it. They’ve not always been right; Newton and perhaps the Vision Pro.
Computer scientists, coders, do great things. They’ve solved some hard problems. And created some. A way forward is in university curricula and even online certifications. Mix in some anthropology, sociology and psychology.
When computer scientists, typically a rather curious lot to begin with, learn more about the meaning of being human, they’ll find even more new, novel and interesting ways to solve for problems, unleashing a new wave of innovations. Perhaps fixing things more than breaking them.