Expert: Implementing an AI development strategy is a must

France, Germany and the United Kingdom are the examples of European countries, in which programs supporting the development of artificial intelligence are being implemented. This primarily includes financial support for research on AI conducted in a given country and stimulating cooperation between science and business in this field.

This topic appears also in Poland. In May, during the Impact`18 Congress, Deputy Prime Minister Jarosław Gowin announced that work on the national AI development strategy would start in the autumn. He said that its assumptions would be prepared "by a group of technological leaders, with the participation of scientists".

Prof. Dariusz Jemielniak, Chair of Management in Networked and Digital Societies at the Kozminski University, emphasises that the development of AI in Poland would require a "clear strategic decision", which would include elements such as strong funding for research in this area, development of study programs and prevention of brain drain.

"AI is still a relatively low capital-intensive technology, you can conduct groundbreaking research in a country like ours" the researcher adds.

"If we fail to implement the strategy or at least smaller development programs, we will have to face the risk of an missed opportunity for a civilization leap" - he warns. "We will still develop, also in the area of AI, but it will be a walk uphill. As a country, we need such rare opportunities, in which our strengths help us compete with big players" - adds the researcher.

Considering how fast artificial intelligence is developing today, "we must say clearly that the national AI development strategy is needed to give the matter a proper priority" - says Prof. Jemielniak. "A coherent, consistent and well scheduled development plan for these disciplines is needed - I dare say it`s much more important than electric cars or a central airport".

The researcher adds that planning activities related to the development of AI could also improve the situation of research on artificial intelligence conducted at Polish universities. "For example, research on AI or technology studies conducted in my department only moderately fit into in the present National Science Centre funding systems, we have to compete in sociology, management or communication science competitions. Dedicated competitions would greatly facilitate progress. Similarly in the case of training programs: we need Big Data, VR, AR, computational social science, machine learning specialists" - explains Prof. Jemielniak.

Who should participate in creating the strategy? "It should definitely involve business practitioners and scientists - importantly, not only those specialising in technical sciences, but also social sciences and humanities. Funds and programs should follow that allow for cross-cutting activities beyond the disciplinary divisions" - says Prof. Jemielniak.

He explains that the current state of development of artificial intelligence in Poland is not a result of a long-term development plan in this field. "Artificial intelligence is really a set of about eight types of technology" - the researcher reminds. He adds that today`s state of research related to AI consists of research conducted by "several dozen people in each discipline" as well as the spontaneous development of this technology that can be observed in the private sector.

"We have unquestionable successes, but we lack a uniform and coherent bundle of competences in the area and sensible support from the state" - the researcher says. He adds that the excessive bureaucratisation of grants settlement is a huge obstacle. "Even the so-called fast-track research funding from the National Centre for Research and Development is very bureaucratic, it requires a large amount of time and effort to adapt to project logic - which for many businesses is less profitable than acting alone, and this is certainly a worrying symptom" - he adds.

PAP - Science in Poland, Katarzyna Florencka

kflo/ zan/ kap/

tr. RL