“I’m against innovation”: Dr Daniel Glaser explains why innovation doesn’t always equate to progress

The founding director of London’s Science Gallery on what he’s learnt from his recent work

One of my favorite satirical newspaper headlines reads: “Scientists announce breakthrough that partly solves problem created by previously announced breakthrough.” In our headlong rush towards innovation, the pursuit of novelty for its own sake produces as many new difficulties as it resolves, and it can distract from existing, more important problems.

One approach that can guard against too much innovation is the idea of releasing projects in ‘beta’, meaning they are not complete and may be changed substantially in the future. If you don’t need to ‘finish’ something immediately you don’t have to force all the new things in at the beginning. This also has the benefit of allowing what people actually like to shape the outcome. For instance, Google’s email service, Gmail was in beta for five years and ended up with plenty of new features, most of which had arisen organically.

In fact, this approach to making new work goes back over half a millennium to the history of oil painting. Although a wide variety of pigments had been employed since prehistoric times, painters in the renaissance started to focus on oil-based paints because of a particular property: the fact they dry slowly. This meant the artist could refashion the painting over a period of time, sometimes after it had been seen in public. The piece could be fine-tuned rather than being forced to make its mark on day one. Resisting forced innovation is not a new idea.

Saint George and the Dragon by Renaissance painter Raphael (From the collection of National Gallery of Art, Washington DC)

The very notion of a sudden flash of inspiration, so often the driver of innovation, is an illusion. For millennia, human creativity has been understood as the muse or the external spirit downloading the latest idea to the human. Samuel Taylor Coleridge tried to write down "Kubla Khan; or, A Vision in a Dream: A Fragment" after he awoke from a vivid dream and then forgot it when he was interrupted. And August Kekulé claimed to have understood the structure of benzene from a dream-like vision of a snake swallowing its own tail. But we are fooled by the illusion that we can be aware of our own thinking. In fact, conscious thought is a story we make up about ourselves and most of what goes on in our heads is invisible to us.

This can be turned to our advantage. A number of studies have tested the effect of giving people irrelevant activities to complete once they have fully understood a complex task. Compared to the group that simply plugged away, those who were forced to do an irrelevant bit of maths did better when they came back to the original problem. Constantly focussing on the new makes us narrow.

Of course, technology always offers the possibility of improvement. These days machine learning algorithms (many of which have in fact been around for decades) are being unleashed on the vast datasets generated from internet searches and uploads. There is a widespread hope for an innovative approach to how we judge what people are good for, whether it’s filling a vacancy or selecting an artist for a show. Rather than messy decisions based on fallible human opinions we can use objective criteria to decide who’s right for the job. Unfortunately, it is becoming clear that the unconscious biases which underpin all of our thinking, despite our blissful ignorance, are in fact baked into the systems that we build.

Samuel Taylor Coleridge (From the collection of LIFE Photo Collection)

Study of human biases suggest that the most reliable approach is to find ways to blind the reviewer or curator to the identity of the applicant or artist by excluding tell-tale signs like name, address, date of birth, and obviously their photo. But the new algorithms ruthlessly hone in on any information that correlates with the previous decisions that they were trained on. The history of prejudice, conscious or not, is perpetuated by the very systems we trust to overcome it. In the end we may need to apply old fashioned political choices about what we want the people we select to look like. Where you put your advertisement or which existing networks you work with can be more effective than rushing towards an innovative fix. We always risk being tempted by novelty into ignoring the underlying problem.

This becomes more and more prevalent as the capabilities of our smartphones get better and better. We fool ourselves into thinking that we can do everything on them. But actually looking at photographs, watching videos, or reading emails is a very impoverished experience on a handheld device. Our biological visual systems have evolved to scan around images and text by moving our eyes, guided by the low-resolution information at the edges. But small screens are more focus and less periphery. We can’t see what we’re reading in context so we don’t assimilate it properly.

This changes our diet of information consumption from something much more wholesome and properly paced into a continual snacking on content, which leaves us hungry but also leaves the information poorly digested. If we didn't have the new illusion that we could respond to messages while walking along the street, we wouldn't get into loops of partial misunderstandings and incomplete responses, and our relationships and collaborations would not be disrupted by these rapid short brittle spirals.

Google Nexus One Smartphone, 2010 (From the collection of The Henry Ford)

Sometimes the people you want to work with don’t want your innovation even if you developed it with them in mind. When we were setting up Science Gallery London a few years ago we were interested in finding ways for young people from non-dominant groups to work with us at the cutting edge of culture. First we interviewed and got to know our young leaders group. Then, instead of dreaming up an innovative list of projects for them, we asked them to tell us what skills and achievements they wanted to say they had done by the end of their year with us and then tried to find a way to help them do that. The innovation was not to cook up something new but instead to listen to the young people and help them along the road they’d already chosen. If you’re interested in boosting diversity, in order to make your project better, using conventional career trajectories may end up being more inclusive.

Even when you’ve found an effective intervention, you might be tempted to find innovative ways to measure what you’ve achieved. This is another common mistake. When I was still in the lab there was always a tendency to use a sexy new brain imaging analysis to demonstrate your killer new result. But actually, if you think you’ve found evidence for telepathic transfer between humans, you’re probably better demonstrating it using a bog-standard technique to analyse your neural scans rather than a brilliant but relatively untested approach. Too much innovation means you can’t compare your results with what’s gone before. And that leaves you vulnerable to believing your own hype.

From these disparate domains we learn to listen to the audience, stay focused on old problems, and not to change too many things at once. Now that’s a novel idea.

Modern imaging of the human brain (From the collection of Deutches Röntgen-Museum)

Explore more:

– The Darker Side of Invention