Artificial intelligence is shaping our life. Many leading forces in this key sector are part of American corporations; in Europe, we are sliding into dependency. Wider questions are raised with the rise of artificial intelligence. Increasingly, people worry about the impact that this technology will have on our societies and economies. While it is true that there may be some reasons for concern, I believe that a rational and systematic analytical approach will serve us all best.
1. What do Machines do well?
It is the purely data-driven tasks where machines perform very well, for example:
- Standardised tasks, executed in mostly stable environments: Take for example the analysis of a pre-formatted complaint letter or the execution of a financial trade or booking that obeys a rule in accounting or trading since a set number of years. This is already being done by robots. Banks have myriads of robots in their company; not the ones you see moving but the computers that execute specialized tasks they have been taught to execute.
- Processes where a lot of high quality data has been collected which creates patterns: This is true for example for diagnostics where decades of research underlie the results of studies, and diagnostic standards are changed only every decade or so. It is true, too, for self-driving cars with sensors monitoring what is happening on the street. From such data, patterns can be modelled with machine learning algorithms. That means that once you know enough, you can simulate it. Neural networks working alike to structures in our brain are only one way to do that. You can build artificial neurons and get them to connect to other artificial neurons. This learning process builds “highways” that will come together as a network. This may not be entirely easy to understand, nor is the process fully transparent – but it works.
- Unrelated data from unstable environments, providing new insights that have been hidden from human beings: either because there is too much data to analyse efficiently/ in a justifiable time, or because there is too little understanding of the behaviour of the data generated. This could apply, for example, to diagnostic tools that detect genetic anomalies related to life habits in large populations, or to changes in stock market prices related to widely scattered events around the world. Such an analysis can be done with different kinds of machine learning algorithms, combined with statistics or even text data.
In any of these cases, our reality is modelled in a particular way. But machine learning is progressing rapidly. As more data becomes available and as it becomes cheaper to run algorithms (many sophisticated machine learning applications are open source, meaning available for everyone to use and modify, without charging for licenses), these models will come to resemble reality ever more closely. Some experts see this development critically; others praise this technological advance as the best way to lead humankind towards a new and bright future. Which side is right? Impossible to say today.
-
Where are humans better?
As humans live, they create reality through communication, perception, action and reaction. Everything that we excel at and will excel at in the future is related to the reality we live in. So far, computers have no personality, no “spiritual” beliefs, and neither cultural identity nor concerns. They also have no ability to resolve diplomatic conflicts, and they have no nationality.
So far, artificial intelligence is the result of collecting myriads of data and training systems to function faster or more accurately with the help of computing power. In either case, artificial intelligence is limited to hard facts. In contrast, as humans we can derive insights from our experience about things that we know very little about. This makes us more competent for complexity.
Here are some areas where I think human abilities will continue to be invaluable. Therefore, these are also abilities that humans should master to stay ahead of machines:
- Empathy and interpretation. Today, computers struggle with irony, jokes and context they don’t know. In contrast, humans learn about a person’s motivation and characteristics by reading his face and looking at his gestures. The other person will define the interaction with you in a similar way. As a human, you also value other people because they are part of a certain organisation, have certain experiences you share, or have gained a specific level of trust. Just take the example of bank advisors or doctors, teachers or family members.
- Physical interaction – and independence. Anything will somehow be automated someday, no doubt. But every computer has a creator, and therefore, I say, a hidden agenda. In contrast, humans are present in flesh and blood – and their presence means that they can take independent actions. As a human, you own the power of your choices as a consumer. You also have a say in our democratic system. USE IT! Imagine what the touch of a hand feels like – one diplomat shaking the hand of another. Don’t give up the influence you have. If you accept to be substituted, technology and the programmer behind it will decide for you.
- Adaptation to change. All the systems we know have been trained on the basis of large amounts of data to excel in one task and one task only. In contrast, humans are much better than machines at changing quickly and radically. We can also combine tasks quickly and integrate them in complex environments that are changing as we speak.
So what does this add up to? Computers can augment your basic education, but they cannot replace it. Gather experience, learn about politics, biology, math, physics: the combination of scientific possibilities lead to great leaps in achievements. All these aspects relate to IT, and the more you know about it, the better we can influence it.
-
Which Role for Europe?
Technology is mostly driven by the United States. It will not be easy to build a counterweight on this side of the Atlantic. We urgently need to massively invest into our digital infrastructure; the EU should push for this. European policy makers should also support greater and faster exploitation of technology to serve the goal of a better life for humans.
We should concentrate on making further progress in areas where we already have strengths. Check the 2017 World Economic Forum report for the advantages that Europe has in terms of competitiveness:
- Education, especially in healthcare and MINT sciences
- European values that emphasise tolerance, diversity and respect
- The potential for alliances within Europe to allow for bigger steps for example in digitalisation
- Free movement of people including a lot of highly skilled talent
- The brand name and unique selling points of different European countries which stand for value and fairness as well as sustainability
That there is a talent war going on is no secret. It is still humans who build machines and technology, and not machines which build machines. Programmes like Erasmus have shaped whole generations. Start-ups including the entrepreneurship movements in big European cities will keep attracting people.
Beyond the next decades, our ability to significantly influence the future depends on education, innovation (speed and scale!) and the freedom of movement within the EU. Tolerance and diversity are the cornerstones of this paradigm.
Anna Penninger works as Senior Consultant for IBM in Switzerland. She is a member of United Europe and prepared this contribution for the Young Professionals Seminar in Amsterdam in June 2017.