Following on from a previous post about my engineering thesis, this post is about the human side of that journey. Over 10 years ago there were technical challenges of computing power and data storage that no longer really exist today. Challenges on the softer side of innovation unfortunately haven’t disappeared so quickly, and still remain just as relevant.
Challenging the Status Quo
As a final year student, I walked into the first meeting with my supervisor armed with a rough direction for a solution and a hunch about why that approach should work. My supervisor was the head of the engineering school and I was told during that meeting that my approach had been tried many times before. And in no uncertain terms – it didn’t work.
That conversation was like a red rag to a bull. I don’t think my results would have been as good as they were without that initial adversity. You need some people like that in your team, those who get spurred on by confrontational challenge and have a healthy disregard for established authority. You don’t want an entire team of those people – that’s a recipe for disaster. You definitely want at least one though, and more importantly you want to know when to listen to them.
Cross Functional Thinking
This was originally a medical problem about diagnosing irregular heartbeats. The engineers reduced it to a mathematical puzzle, looking for a single equation that reliably agreed with the accepted medical answer. From what I learned in psychology, your brain recognises patterns by paying attention to certain parts and disregarding others.
Imagine you’re looking at a picture of the Mona Lisa, and you’re asked whether it’s a forgery. I’m no art expert, so unless there was something obviously wrong I’d probably tell you that it could be real. An art expert, on the other hand, would know exactly what to look for. They know what to pay attention to. It may be detailed brush strokes, but it may also be its overall tone. One requires close inspection, the other requires viewing the painting from a distance.
A computer program that roughly approximates some very basic attentional processes of the human brain is actually quite effective at diagnosing irregular heartbeats. After only a few examples, a program can learn to ‘pay attention’ to important differences and ignore similarities.
Incorporating ‘attention’ into the solution to this medical problem didn’t require a large cross functional team. It only takes one person to introduce something seemingly unrelated, yet actually very useful. The key skill here is to be able to speak in a language that others understand for as long as it takes to win them over. Although they mean practically the same thing, engineers understand the words ‘transform coefficients’, and psychologists understand ‘receptive fields’. Know your audience and adapt accordingly.
Don’t Expect Results to Speak for Themselves
I have a relative who is a cardiologist. He had a handheld device that could do everything my thesis did and more. Fast forward 10 years and now a smartphone could probably do all that and call the patient to let them know the results. So he was understandably puzzled why I would want to concern myself with a problem that he thought was already ‘solved’.
As an engineering student, I was led to believe that it was about striving for better accuracy or faster speed. Which is about as far from the truth as it could possibly be. Even perfect, instant results weren’t going to win over the cardiologists.
The answer to this, which admittedly wasn’t obvious to me at the time, was that it’s not actually about irregular heartbeats at all. Progress on this problem has the potential for applications far beyond medicine. Getting computers to reliably recognise patterns in the world around them is something that has the potential to allow huge advances in a range of technologies. Driverless cars, as just one example.
Part of the responsibility that comes with innovation is to imagine solutions in countless different applications and contexts. The crazier the better. That’s the true value of innovation – it not only solves the problem in front of you, but becomes a solution to the problem ahead of you. The problem you didn’t even know you had yet.