TUE Martijn Heck

Martijn Heck is the scientific director of the Eindhoven Hendrik Casimir Institute.

10 May 2023

Where’s the real-world impact of the grandiose claims being made in high-impact journals, wonders Martijn Heck.

A few months ago, I heard a venture capitalist complaining about the unrealistic value universities put on their high-impact Nature and Science papers. According to him, it takes at least ten times the effort to turn those academic results into a real product. I fully disagree. I’ve almost never seen a Nature or Science paper in my field that can be turned into a real product. It typically takes at least ten times more research before startups or other industry players will even start thinking about turning it into a product. So I’d say your high-impact journal result is typically less than 1 percent of whatever product you claim it will be. At best.

That contrasts sharply with the impact claims being made in such journals. Just about every article touts the likes of a new computing paradigm that will replace digital electronics, a transistor that will make computers run a thousand times faster, a million times faster internet or a sensor that fits in any wearable and monitors you like a Star Trek tricorder.

Look, I don’t mind inflating impact a bit. But it seems to me that the level of exaggeration scales exponentially with the journal impact factor. Whereas your typical IEEE journal stays relatively down to earth, and dutifully explains and quantifies how a CMOS circuit lowers the noise or how a new laser design improves efficiency by a few percent (big deal, by the way!), every article published in Nature and Science claims to change the world.

“We built an optical chip to make the internet a million times faster!” No, you didn’t. You used a bunch of 19” racks full of expensive equipment coupled to a nonlinear waveguide on a fully equipped optical table. Nothing integrated, no chip. “We have a new THz transistor, so computers become thousand times faster!” No, you have an element that you excite with a huge femtosecond laser to trigger a sub-picosecond current. Apart from this being a far cry from a computer, we forget that THz transistors already exist and aren’t the speed bottlenecks in processors anyway. “We have this new nonlinear material, which will revolutionize neuromorphic computing!” Well, you have a new nonlinear material – that’s it.

ASML special

Why do we allow this? If we reason that the title, introduction and conclusion of an article are part of it, then these segments should be held to the same scientific standards as the methods, results and discussion parts. Clearly, that’s not always the case, and often we can argue that these claims aren’t well justified. Should all those papers be retracted, due to unscientific claims? As much as I’d like to give that a try, by writing a comment to the editors, this isn’t my point here.

The issue at hand is that parts of academia are creating their own world, their own ivory tower. I’ve realized that ever since I started writing grant proposals, about a decade ago, I use the term “real-world” a lot. Real-world technology. Real-world usage. Real-world impact. I have to do this because otherwise, proposal reviewers don’t understand that I want to work on an actual technology, which can actually be used and has an actual impact on the people and society around us. I happen to like IEEE journals, so my focus on the real world doesn’t bother me. But that’s not the case for many academics, universities and – most importantly – ranking institutions and funding agencies, which all love the exaggerated claims high up in the ivory tower, far above the real world.

This isn’t a rant on all that’s wrong in academia. This is actually a long, but well-argued introduction (take note, Nature Publishing Group!) to a very short problem statement: the Netherlands is making roadmaps for semiconductors, photonics and optics. Where are the research and education in real-world transistors, real-world lasers and real-world lenses? You know, the stuff that will be part of our gadgets and pervasive in our society in 2030?

There’s nothing wrong with a tall ivory tower. We need that blue-sky thinking above the clouds. But only as long as we have the engineering scientists and engineers to build the elevators. Otherwise, the only real-world impact academia will be making is when they fall out of their ivory window.