– To work with artificial intelligence you don’t need expensive labs and you don’t usually need supercomputers. Most things can be done with a laptop with a graphics card. If you need more computing power, you can simply buy cloud computing hours – says Dr. Łukasz Borowiecki, co-founder of 10 Senses, a company providing data science and machine learning services, in an interview with Business Insider Polska. In his opinion without clear financial support from the state, we risk losing competitiveness in the field of AI for the benefit of other countries.
Dr. Łukasz Borowiecki, co-founder of 10 Senses: I don’t think any country is really prepared. Few people expected neural networks to be such a powerful tool for solving all kinds of computing problems. So, when scientists around 2012 proved the superiority of this type of algorithm, it turned out that there were very few people working in this field. Everyone was surprised. Of course, Silicon Valley’s biggest tech companies reacted the fastest, buying experts and then a few startups off the trunk. However, at national government level, the response has been slower. So today by reading official government documents, one can discover, for example, that the British are afraid of American domination and the flight of experts abroad, or that the Germans think that they have lost in recent years and are technologically backward in the field of artificial. intelligence. We’re not the only ones wondering if we’re ready.
I think so. Artificial intelligence economists widely agree that technology will impact many sectors of the economy, as has been the case with the internal combustion engine, electricity and the internet. This would mean that the development of artificial intelligence would have a direct impact on the long-term competitiveness of the economy.
PwC predicts that following the implementation of artificial intelligence solutions, UK GDP will be 10.3% higher. in 2030, which would represent an additional £232 billion. What will this implementation consist of?
For example, the British see in artificial intelligence an opportunity to increase the competitiveness of their own industrial sector. Currently, industry’s share of UK GDP is only 10%. and this is very little compared to other European countries. Artificial intelligence has many industrial applications. that’s why The British hope that investments in AI will allow local companies to regain competitiveness against, for example, German companies.
What is the greatest strength of the British AI strategy?
First of all, it should be emphasized that the British strategy is being pursued on solid foundations. Britain has an excellent academic base and there are many world-class teams dealing with artificial intelligence at its universities. Additionally, the UK has a strong startup ecosystem that makes it easy for tech companies to fund rapid growth. Britain’s artificial intelligence strategy builds on these two strengths and focuses on rapidly expanding the number of PhD students and research teams, attracting overseas experts, facilitating access to data and high availability of capital.
To become a major player in AI, a fundamental change is needed – this echoed from time to time during the working meetings of the groups working on the AI report at the Ministry of Digitization. Did Britain also need fundamental changes, or did that just happen naturally?
Rather naturally. Documents related to the British Artificial Intelligence strategy were developed in 2017-2018. However, already in 2015, the Alan Turing Institute was established, which is a national research institution dealing with data science and artificial intelligence. In previous decades, the British have funded various artificial intelligence research teams. Thus, while the strategy currently presented is centered on the rapid development of AI, it is however the next stage of many years of investment in this type of technology.
Where is Britain in work on AI?
They are among the world leaders. For example, DeepMind operates in London. This startup was founded in 2010 to develop artificial intelligence technology. Four years later, Google bought the company for over $500 million. Currently, the founders of DeepMind pride themselves on having the largest team of artificial intelligence experts in the world. According to them, nearly half a thousand employees hold doctorates. Among the successes of DeepMind is the construction of the AlphaGo program, which beat the world champion of the Chinese game of Go for the first time. This was considered a great achievement because the game is much more difficult for the computer than, for example, chess. It should also be noted that DeepMind employs many Polish scientists, so already at this stage the UK AI ecosystem is a magnet for experts from other countries.
When creating the Polish AI strategy, what should we mainly focus on?
My opinion is that the most important task is to increase the number of people with the appropriate qualifications. First, we should rapidly increase the number of people studying data science, artificial intelligence, and related faculties. Moreover, we should ensure the development of research teams whose work would be coordinated around specific problems in the field of the development of artificial intelligence and its implementation. Great Britain can be cited as an example, where it has been decided to allocate 250 million pounds for work related to the development of self-driving car technology itself.
Could some elements of British strategy simply be copied?
In my opinion, the set of institutions within which the UK AI strategy is implemented is very reasonable. On the one hand, we have the Alan Turing Institute, which is a scientific unit that coordinates research in the field of data science at the national level. On the other hand, it is assumed that additional institutions will be created for the purpose of implementing the strategy. The Artificial Intelligence Council is to serve as a discussion forum between business, academia and government. The Office for Artificial Intelligence should take care of the implementation of the strategy. In addition, there are plans to create a Data Ethics Center, which should deal with the development of appropriate standards so that algorithms and data are used in a safe and ethical manner.
This distribution of responsibilities seems reasonable. In particular, the creation of separate institutions in charge of strategy development (IA Council) and implementation (IA Authority) seems to facilitate the coordination of development work.
Looking at the British AI strategy, I understand that without the involvement of the state, especially in finance, it will not be possible to create a thriving AI environment?
Currently, more and more countries are announcing their own artificial intelligence development strategies. Within these strategies, specific sums will be allocated for staff training, research and implementation. In this situation, without clear financial support from the state, we risk losing competitiveness in the field of AI to other countries. It is difficult to expect that the task of technological development will be taken over by private Polish companies, since they do not have the financial capabilities of Western companies.
Here, however, it should be emphasized that investing in artificial intelligence does not require as much investment as in the case of other technologies. To work with artificial intelligence, you don’t need expensive labs, and you don’t usually need supercomputers. Most things can be done with a laptop with a graphics card. If you need more computing power, you can simply purchase cloud hours.
What is the impact of artificial intelligence on the economy?
There are a lot of unknowns at this point. The most common assumption is that artificial intelligence will have a huge impact on the economy. However, it is difficult to predict precisely because we do not know which algorithms and solutions will be available in 5 to 10 years. Historically, many breakthrough technologies have had a delayed effect on the economy. This is due to the need to train human resources and develop infrastructure. For example, electricity generation and transmission technologies were developed in the second half of the 19th century. However, they did not result in increased production efficiency inside factories until after World War I.
An often debated question is the impact of artificial intelligence on the labor market. Over the past few decades, the progressive automation of industrial production has systematically reduced the need for labor. However, they found employment in the service sector, where, in addition, the work was generally less physically demanding. Therefore, the automation of production outside certain industrial regions has not had a negative social impact.
However, the development of artificial intelligence is likely to automate work in the service sector. Algorithms will enable the implementation of many tasks that are performed today by doctors, lawyers and civil servants. And while the vision of a world without work seems very distant, the real problem is the threat of a systematic erosion of the labor market.. I mean limiting the availability of “good jobs”, those that guarantee financial security, access to health care and retirement benefits. This process is already underway. For example, in the United States, median incomes have increased slightly since the 1970s, while the costs of health care and higher education have increased very rapidly. Americans are aware of these negative trends and this has been reflected in their policy choices. Of course, automation is not the only factor increasing inequalities, but we must not lose sight of the impact of technological change on the labor market.
Read more about the UK’s AI strategy in the Digital Poland Foundation report: “Review of the Strategy for the Development of Artificial Intelligence in the World”.