“Can supercomputers and artificial intelligence drive technical, economic and social progress?” — I spoke about challenging opportunities in dealing with smart technologies at the Virtual Conference 2020 this week. Today, I would like to share my thoughts on this topic here on LinkedIn.
How did people imagine the future a quarter of a millennium ago? Certainly not in the way it has now become reality. Back then, James Watt built the first steam engine, horses pulled carriages. A normal, modern-day middle-class car, with combustion engine, radio technology, touch displays and on-board computers: unimaginable.
250 years after the invention of steam engines, we are now experiencing a change that many describe as the beginning of the “second engine age”. The worldwide data stock is exploding, computers are calculating faster than ever. Cray-1 is considered the world’s first supercomputer, and when he started to calculate at the Los Alamos National Laboratory (USA) in 1976, his 500 megaflops of power were the best in the world. The current record holder is called “Fugaku”, is located in Kobe (Japan) and carries out more than a million times (!) as many arithmetic operations per second as Cray-1. One thing is certain: interconnected machines will generate such enormous amounts of data that they can no longer be handled by humans.
What will our future look like in this digitalised world, side by side with machines? I don’t have a crystal ball. And yet I commit myself: the future will not look like we imagine it now.
Two suggestions on how we should think about the future
Describing the dimensions of change from the past helps us to understand the pace and scope of past changes. Yet we can only guess at what is to come, because change is happening faster and faster. Moore’s Law, the rule of doubling computing capacity every ten years, describes exponential growth. And thus, hardly imaginable changes.
We often think too pessimistically about the future.
We are not good at thinking about long periods of time in advance. Therefore, we should not imagine a possible future too concretely, not too much in terms of ideas about things and how they will look in the future.
There is a second problem: we often think too pessimistically about the future. Even in the early days of the first machine age, contemporaries feared mass poverty and impoverishment if the steam engine were to compete with the muscle power of the people. Now that AI is competing with intellectual laborers, unemployment scenarios are coming back into focus. But why?
As bad as we are at anticipating long periods of time, we are just as good at continuously adapting to innovations. In fact, we are much better at it than we think. The steam engine did not leave the masses unemployed but brought humanity the highest level of prosperity in world history. And even now, various studies have concluded that the digital change could result in just as many or even more jobs than before — because although it makes some job profiles redundant, it also creates new ones.
If we really internalise these two insights — not thinking too much in concrete product ideas and not too negatively — we can gain a better view on the future. And above all, a sharper view of how we can create it. In our sense, according to our ideas…
Innovations that no one can imagine today
Will AI and supercomputers advance the world technologically? They certainly will. Just different from what we might imagine. They will bring innovations that no one today can even imagine. We will work and interact with intelligent machines and robots; they will become part of our everyday lives.
AI and robots are no longer great mysteries in our lives, they do not embody the evil in our reality.
But we must abolish the idea that they look like us, maybe even compete with us and want to take over world domination. We do not know what they will look like, because we do not know how people will shape them so that they are most useful to us.
We will not even notice many of them, because they are stuck in objects that we have long been familiar with. It’s already the case today. AI and robots are no longer great mysteries in our lives, they do not embody the evil in our reality. Instead, they facilitate, support and help us, and we often don’t even notice it — in the car, for example, or on the smartphone, when streaming or shopping online.
AI and supercomputers are radically changing the way we live and work — and if we do it right, it will be for the better. This will determine whether they will also help us advance economically and socially. It is not just a matter of boosting economic potentials. They are enormous — McKinsey expects them to exceed those of the steam engine.
Why ethics must play a key part in the development of machines
For them to advance us socially, ethics in particular must play a central role in the development and use of AI and robots. Even if we do not know what future innovations will look like: we can determine the requirements they need to meet. Because we people define what machines should be allowed to do — and what not.
There are already numerous incidents of AI discriminating people. This is not acceptable!
One important example is diversity. There are already numerous incidents of AI discriminating people: Because you are a woman, you don’t get the job. You don’t get a credit because of your skin colour. That is not acceptable! Such mistakes are caused, for example, by too small and narrow-minded data sets when training AI — by too little diverse data sets, but also by product and development teams who are not aware of their possible bias. We must work together on this and invest in (further) education.
For us to design technology for the benefit of all people, the most diverse perspectives must be brought into the development process.
If we do it right, this structural change is not a negative change, but a tremendous opportunity. We can design AI and supercomputers to enrich our everyday lives. Whether they will do so is not up to the technology — it’s up to us. That is why we should not only invest in artificial intelligence, but also in human intelligence. Investment in AI should always be combined with investment in digital skills. Because we need smart, diverse minds to shape our smart future with smart technology.
Further information on the Virtual Conference 2020 can be found on the event’s homepage.