The discussion on how universities should look like in the era of artificial intelligence is a red herring. Instead, we should concentrate on how to function as a society in the world where AI-based solutions will play an increasingly important role. Mateusz Gaczyński, Deputy Director of the Department of Innovation and Development, Ministry of Science and Higher Education, in conversation with Maciej Chojnowski
Maciej Chojnowski: During the debate of the Coalition for Polish Innovations and the Foundation for Polish Science on relations between science and business you said that the Constitution for Science meant the completion of a broader legislative process. What does it consist in?
Mateusz Gaczyński*: The Constitution for Science is a unique umbrella for all changes to support development and innovation and to bring together scientific community and business in Poland which we have been implementing since the end of 2015. They consist in two acts on innovation, the act on the Łukasiewicz Research Network, and the Constitution for Science. The system is supplemented with the act prepared by the Ministry of Development regarding IP Box, i.e. tax reliefs for new technologies or results of research and development studies.
I dare say we have a very well developed system to support both sides that want to collaborate on joint projects and to invest in research and development. There is nothing to be ashamed of, even if we compare with the countries that are higher in innovation rankings than Poland.
In the “Artificial development policy in Poland” there are two important subjects connected with education: Virtual Research Institute and doctoral school. Looking from the perspective of the Ministry of Science and Higher Education, what is key to support artificial intelligence development in Poland?
The doctoral school or the Virtual Research Institute are very important because they will provide staff and study results, which can later be applied in economy. However, if we talk about development of the economy based on data, our advantage is that every year Polish universities award diplomas to about 100 thousand graduates specializing in the fields where machine learning and AI elements may be used almost instantly. An engineer means someone whose professional DNA will always contain certain knowledge on digital technologies.
At this point, we come to another matter. There is the issue of the extent Polish entrepreneurs see opportunities offered by artificial intelligence. According to the last year’s Digital Poland Foundation study, very few Polish companies see the potential of applying digital technologies that would be connected with something more than just having a website or e-mail.
It is true that Polish entrepreneurs are rather reluctant to implementing new solutions. But even if business and science decide to collaborate on some projects, other problems appear. Universities complain about specialists being poached by businesses, while entrepreneurs are discontent with graduates lacking a holistic approach and being unable to work in teams. The academic community retorts that they would be better prepared if companies shared their data or organized traineeships. Do you think the Ministry could help with that situation?
It is true that graduates who want to work for commercial entities are not prepared. But the situation is the same in Poland, the United States, France, Germany, and Great Britain. OECD studies show that entrepreneurs from all countries perceive higher education deficiencies as a problem. No matter what the education model is like and how well it correlates with business, companies will always be dissatisfied with the “product” issued by universities. The objectives of the two systems are somewhat different.
But you can try to encourage future employees to learn fast. Then, both the entrepreneurs and the teachers would be contented. However, this process must start before the university studies. A person who knows how to learn fast can be almost fully retrained by a company in the space of six month or one year.
Education profiling to satisfy the needs of a company is also possible, provided that it doesn’t interfere with the curriculum. That is already happening. There are studies whose curricula are developed in collaboration with entrepreneurs: in machine, automotive, electronic and pharmaceutical industry. In Kraków there are postgraduate studies for pharmacy graduates where you can learn industrial pharmacy.
We are wrong to believe that we can create something that would possess all traits of intelligence and that, at the same time, would exclusively serve our purpose
Recently, Polpharma, the University of Warsaw and the Ministry have held trilateral talks on the idea to create an education path at the University of Warsaw that would take into account the skills sought after by Polpharma. Both parties still need to agree on what can be done. The University has to develop a framework that would substantiate the award of the master’s degree to graduates. That is the proof of education quality, especially in the context of the Polish Graduate Tracking System (ELA) launched by the Ministry several years ago. Today, people, who are about to decide what faculty they want to chose, can refer to the system to see, in black and white, which option is the best.
Do companies and universities collaborate spontaneously on the idea to develop a course of study? Does this form of collaboration result from the specific nature of the field that is sought after? Or is it rather supported institutionally, e.g. by the Ministry of Science?
The Act 2.0 lays down clear rules stating that there is a need for interaction between the universities and the social and economic environment. As far as “professional” faculties are concerned, we have extended the period of traineeships from three to six months. The industry believes it is a bare minimum. The formal framework for cooperation is there.
A lot depends also on how open university authorities and companies are. Companies signal their needs to universities if they see an opportunity to develop their business. Bright future lies ahead for energy and biotechnological sectors, for example.
Also universities adopt a more open approach because they know that if they don’t create an attractive offer for people, they will lose their renown. Let us not forget that all this is happening in the times when the number of students is shrinking. The fight for good people, including good students, is getting more and more visible.
Obviously, some universities consider themselves excellent and believe they do not need to collaborate with the industry. And they may be right, in the short run. The big question is if that approach will stay viable in 10-15 years. Personally, I do not think so, as the competition for good people is global.
In his book called “Robot-Proof” Joseph Aoun, president of Northeastern University in Boston, writes about „robot-proof” education. He believes that in the era of AI it is very important to change the way you think about a university. Universities should open to people willing to acquire new competences throughout their professional career to adapt to the labor market. Can that vision come true in Poland?
Today, we cannot tell how education should look like when AI solutions will have a much stronger influence on our social or economic reality. Therefore, the ideas you mentioned should be treated rather as an example of how universities try to justify their raison d’ être in the reality that can prove dramatically different from what we are prepared for.
In his book “Life 3.0” Max Tegmark claims that, when confronted with artificial intelligence capabilities, people will be able to find a job mainly in the areas requiring social interaction and empathy. Something that you are not taught at a university. It’s a cultural process resulting more from family and social issues rather than from education. But we don’t know what human relations will be like when we will have to deal with more advanced artificial intelligence.
The problem is that we are trying to write a scenario about our future but we are still clinging to our past. We are unable to go beyond what has been hammered to our heads in the education process: this is how society looked like and this is how work evolution looked like. But we don’t know how all that will work when something faster, more efficient and cheaper will appear in our lives. Will we be needed for anything at all?
In my opinion, the discussion on how universities should look like is a red herring. Instead, we should concentrate on the whole model of our functioning as a society or mankind in the world where AI-based solutions will play an increasingly important role. Artificial intelligence meeting the criteria referred to by Tegmark will be a “being” far more efficient than us, at least from the evolutionary point of view. The question is what relations we will have with it.
According to the EU idea of artificial intelligence we should build solutions that are focused on a human being and that are reliable and ethical. The creators of the vision of responsible artificial intelligence say that it must be founded, among other things, on education, but add that the education should change. The belief that engineers deal only with technical matters and humanists are devoted exclusively to humanities must be revised. The ethical and humanist component is important not only in the case of engineers but also managers. And it’s not about 30 periods of ethics, but about treating this challenge seriously, as people designing and implementing such solutions usually don’t have to face such problems. Does this idea make sense?
In my opinion that transition has already happened. iPhone has been successful not because it can do more than any other smartphone, but because it was designed in collaboration with people who knew what product would be enjoyable for the user. The same stays true for creative industries. In the sphere of design we observe close collaboration of engineers and people dealing with aesthetics connected with anatomy, in a broad sense.
It all plays out according to more natural rules. Let me refer again to the idea of learning throughout all your life. Its key aspect is that people should be open to what is happening around them. There is a reason why it is said that innovation comes out best in melting pots where different ideas, visions of reality, and cultural elements clash. And this is why California is an innovation incubator.
Creation of ethical solutions also requires taking different points of view into account.
Yes. But how to guarantee such an approach in the case of solutions which, in theory, should think on their own? Although people have the Ten Commandments, they don’t always keep them.
If we don’t “load” those rules to the AI systems, what conclusions will AI thinking on its own have? It may consider us as a redundant protein burden and decide that it will do everything better without our help. This is the basic law of evolution: a being that is less adapted is replaced by another being that is better adapted.
Saying that we will design artificial intelligence with a great amount of ethic rules sounds like wishful thinking. It sure is perfectly natural, but in my view we are wrong to believe that we can create something that would possess all traits of intelligence and that, at the same time, would exclusively serve our purpose. We want to create a genie that would grant us an unlimited number of wishes but that would never harm us. I don’t think it is possible.
What you are talking about is the subject of a recently published book by Stuart Russell. In his opinion, the AI paradigm ought to be changed in such a way so as to put the human in the artificial intelligence decision center. Should we support that approach and adapt the education program?
We come the question of who develops artificial intelligence. At present, the biggest outlays are invested by the military. By definition, armed forces preparing for a conflict do not put the well-being of their enemies in the center of their attention. Military experts design and develop solutions for their maximum efficiency.
What is more, although we can expect that in the culture of western democracy, in the culture of Christianity, development of artificial intelligence will comply with the paradigm where the human is put in the center of the world, the approach of other cultures investing a lot in development of artificial intelligence may be somewhat different. One of the AI experts once said that he wasn’t afraid of artificial intelligence but of people who would use the solutions based on AI.
Studies show that people are afraid of contacts with humanoid robots. Maybe we should teach the next generation not to fear them
Ethics is a convenient concept but I think it is also very difficult to apply. But aside from military issues, stock exchange algorithm parameters are set to generate maximum profit! They don’t care if a given operation results in closing a factory or in increasing or decreasing the value of currencies causing problems in one country or another. Of course, the algorithms operate in a certain framework, only that the framework is defined by humans. And what will happen if the framework is extended because of the need of maximizing profits for any of the parties to the transaction?
So, although I agree with a deeply humanistic idea that we should develop artificial intelligence with the human as the decision-making center, I have serious doubts about whether it is even possible. Even if we developed AI with the human in the center, what would be their role? Who is to decide what artificial intelligence will do? Someone like Pope Francis or Pol Pot?
Agreeing on the goals and values of 8 billion people sounds like a great challenge indeed. Even Russell admits that.
So, paradoxically, maybe it would be a good thing if we get replaced by artificial intelligence. In his science-fiction book “Diaspora” Greg Egan depicted the society a part of which functions in a digitalized form. Those advanced beings, possessing all aspects we associate with highly developed humanistic culture, travel between galaxies with extreme ease and explore things which would never be possible for us in a protein form.
No process can hibernate us for as long as it will take to travel to a planet located 50 thousand light-years away from the Earth. But let’s imagine we are able to digitalize our consciousness, to save it on some kind of a tangible carrier and to send it along with a robot to which, after we reach the destination, we will be “loaded”. Then, we might explore deep space.
But that wouldn’t be us. Not entirely…
Surely, we wouldn’t have our physical bodies. But our consciousness would be there. Totally independent from our physical form. We could function as a protein form, as a robotized form, or as a combination of both. We could also function as information, which has its pros and cons.
Ray Kurzweil is an enthusiast of that approach to the future. But others emphasize how important the material aspect is: our physicality and psyche are inseparable so any attempt to decouple the mind from the body is doomed to failure.
Assuming that technological progress goes faster than it has until now, it is possible that in several decades humankind will produce protein forms. Kurzweil claims that around 2045 we will be able to decide what form we want to exist in: Do we wish to be an element of information cloud or do we want to be a being loaded in one form or another? We could preserve the unity of body and mind: digitalized psyche that could be loaded to a protein body immune to any disease. Additionally, the body would look exactly as we always wanted it to be.
The problem is that we don’t know all that. So it is difficult for us to tell what changes we need to better function in that reality. In my opinion, the only thing we can do is to make everyone as flexible in the education process as possible.
Studies show that people are afraid of contacts with humanoid robots. Maybe we should teach the next generation not to fear them. Soon, biotechnological progress may lead to a situation where we will not be able to tell who is a 100% human and who has been improved in one way or another.
Are you optimistic or fearful about the future?
Being afraid of the future is not going to change anything. The only thing we can do is to be open, flexible and ready for different scenarios. Fifty years ago nobody even imagined that they would have a pocket-sized device that would provide them with any information they wished in the blink of an eye. And that using it wouldn’t require thorough preparation or exceptional knowledge. Some say that with our smartphones and computers we have already become a bit electronic, as it is hard for us to imagine to live without such devices.
For that reason, calls to reduce our carbon footprint do not make any sense because we are getting increasingly addicted to products which make the carbon footprint bigger. Obviously, it’s very good that we are trying to limit our environmental impact but we mustn’t forget that there exist some limits which have been set by our model of consumption. And, unfortunately, that model is not based on self-limitation.
Some believe that new technologies will help us cope with those problems.
Unless we are capable of producing fully clean energy, nothing will change. And I am pretty sure it is not going to happen any time soon.
*Mateusz Gaczyński is the Deputy Director of the Department of Innovation and Development in the Ministry of Science and Higher Education. He is responsible for the research and innovation policy, including the instruments of the cohesion policy and related EU programs. Before joining the Ministry of Science and Higher Education, he worked in the Ministry of Economy (sustainable development, climate changes, Lisbon Strategy), Permanent Representation of the Republic of Poland to the European Union (cohesion and employment policy), Polish Agency for Enterprise Development (entrepreneurship), and the Ministry of Labor and Social Policy (labor migration). He is a graduate of Jagiellonian University and the National School of Public Administration.
Przeczytaj polską wersję tego tekstu TUTAJ