According to the common understanding (and Wikipedia), the technological singularity refers to a point in time when artificial superintelligence will abruptly trigger runaway technological growth, resulting in unfathomable changes to human civilization.
As a marketer first and futurist second, I have always had a slightly different definition of the singularity. I believe it is the time when companies and governments simply can no longer keep up with technological change.
I think we have reached that point.
Over the long span of human history, technological changes have sped up remarkable. China had an almost 2000 year monopoly on the product of silk before Japanese travelers stole a few worms and abducted a handful of crafty maidens. Its dominance in ceramics lasted a thousand years before the Europeans caught on. The printing press wasn’t replaced for 300 years, and the loom lasted a hundred. The PC endured for 40, and processor speeds double every 18 months. Soon quantum computers stir things up again.
If technologies and the products they bring forth last for centuries and decades, that is still very much in line with human life expectancy and our ability to cope with change. But there are a number of fields where technological change is becoming unmanageable. We may call this an “early singularity”.
Here are some of the most obvious:
Despite emerging regulation in certain countries, governments are still pretty much in the dark about the future of finance and the impact of cryptocurrencies. At a recent government consultative meeting one advice to the Central Bank of China told me “We simply don’t know what to do. By the time we have new rules, everything will have changed again.” Whether they are an asset, a currency, or an equity, banks, and exchanges are precariously behind when it comes to understanding, assessing, or regulating these new technologies.
It’s all the rage, and it’s happening now: the complete transformation of the manufacturing landscape. New machinery, robots, a bit of cloud computing and some machine learning for good measure.
Yet if you are equipping a new manufacturing plant today, it will be almost entirely obsolete by 2020, when 5G kicks in. 5G offers the bandwidth and low latency that makes full and safe automation possible. Only the best and most expensive is easily upgradeable. The speed at which technologies in the IoT and IIoT space are changing is breathtaking.
The regulatory schemes around the world will have to change drastically over the next decade. The FDA has moved to allowing more experimental treatments for critically ill patients, but there is a slew of new treatments, therapies, and especially diagnostic tools coming out that they will never get around to licensing before they will be replaced with something else.
While some companies bet on growing human organs in animals, others bet on stem cells and 3d printing. And then we have artificial organs. Very likely not all of these will win the day.
There are numerous technologies for measuring glucose levels by drawing blood yet also efforts using Apple’s smartwatch. Lots of startups are working on diagnosing heart disease through markers in the blood, and then Google came along and announced it only has to look at your eyes.
It is even worse in the medical praxis. There is so much innovation going on yet nothing has the time to sink it and make it into clinical practice. “We are too busy treating patients and don’t have enough time to adopt new technologies,” a hospital surgeon told me recently.
Robotics and Machine Learning
There are thousands of robotics companies out there, but almost every single robot is obsolete the moment it goes on sale or is announced. Add to it machine learning and the race becomes unwinnable. Hundreds of startups focussing on programming robots have been put out of business by robots that learn by themselves, in swarms, or through machine learning. Cloud-enabled, connected robots will learn so fast from each other that any form of training or programming them will soon be obsolete.
A number of companies are competing at the forefront of artificial intelligence, you know who they are. The problem with AI is that it doesn’t forget. Once you reach a “cross-over point” where the AI is able to learn at superhuman speed, the race is over for the competitors.
Nick Bostrom in his book “Superintelligence” explains that brilliantly. In short, if two AI systems compete for the same domain, whoever is ahead even by days or weeks can effectively put the competitor out of business.
Material science is advancing so fast that by the time an airliner goes into production, every single material on the aircraft could already be replaced with a lighter, more efficient, and smarter material. HCG, a Taiwanese maker of toilets, has recently abandoned production of an ultra-clean, antibacterial, smart toilet because advances in material science have made the use of the recently developed coating obsolete.
Energy and Environment
Finally, energy. Whether it’s solar, wind, or even coal, by the time you got all the permissions, planned and built a new power plant, there are more efficient panels, turbines, filters and furnaces available.
I am sure you can think of examples in other industries. We are still building old-fashioned ships that need large crews, while Rolls Royce is working on a global control center of unmanned ships. Many of the autonomous transport vehicles in development will never be used — drones are proving much more efficient. Even leading innovators like Tesla find themselves in a conundrum. You can update the software remotely, but in order to keep your Tesla really state of the art, you would have to replace parts every 2-3 months.
Is the speed of change hampering investment?
A friend of mine planned to open a fully automated restaurant in Taipei, Taiwan. He found suppliers, robots, the right software, and the right layout, and by the time he was done finding a location and securing the financing, technology had changed so much that all his plans were obsolete and overpriced.
By the time a Taiwanese industrial conglomerate had finished work on their first indoor farming concept, new LED lighting and water filtration system had made their concept farm look entirely out of date.
Another company had decided to develop a hand-held scanner specifically for 3D printing applications when along came QIone, a mobile phone app that does the same thing for free. Then there is the company that made an iPad app to visualize furniture in your apartment, when along came the Hololens and put them out of business.
Never in the history of humankind has technology changed so quickly. Even the highly efficient German Patent Office has recently announced they would need 300 more people to clear up the backlog of over 20’000 applications. Brazil is considering granting “emergency patents” to clear a backlog of over 200’000 applications.
For most people, the rate of change is already far too chaotic. But we are starting to see the first companies abandoning plans for investment and innovation because by the time a project is finished technology may have changed so much that it is not just looking old, but also not financially viable.
How can we cope with the pace of innovation? How should companies manage change? Where do you see “early singularities” in your industry? Let me know your thoughts.