Artificial intelligence cost little – Newspaper Kommersant No. 58 (7503) dated 04/05/2023

The growth of private investment in artificial intelligence (AI) technology in the world has stalled in 2022 after a decade of growth, according to a report from Stanford University. The pause in increasing investment is accompanied by the rapid development of this technology and the regular release of new projects to the market. As a result, experts say, the world community is facing questions about the ethical issues associated with the deployment of AI and the measure of its regulation. Countries will have to decide whether to bring the use of AI to human ethical standards, which will greatly limit the technology and reduce its effectiveness, or to give developers more freedom in the interests of competition and development.

In 2022, private investment in artificial intelligence in the world amounted to $91.9 billion, 27% less than in 2021, when the figure was a record $125.4 billion, according to a report from Stanford University. “Index of artificial intelligence-2023”. At the same time, last year’s figure is 1.5 times higher than the level of 2020 and 18 times higher than in 2013. The decline in 2022 is the first in a decade.

The USA remains the leader of private investments in AI – $47.4 billion. China – $13.4 billion, the EU and the UK – $11.04 billion. in the UK – 99. In 2022, the largest volume of investments was directed to such areas as medicine and healthcare ($6.1 billion), data processing and management and cloud technologies ($5.9 billion), fintech ($5.5 billion).

Global investment in AI start-ups fell by 34% to $45.8 billion in 2022, according to research released earlier by CB Insights.

The fall in investment came after a record high in 2021 and amid a comparable decline in global venture capital funding. According to experts, the reduction in 2022 will not become a trend – now the AI ​​market needs some cooling (see “Kommersant” dated March 20).

The Stanford study notes that despite such financial results, AI in 2022 has moved into the “deployment stage” – the technology is rapidly progressing, it has new opportunities. Both large companies and startups are included in the process – AI technology is no longer controlled by a small group of actors. In 2022 and early 2023, new large-scale AI models were regularly released, capable of performing an ever wider range of tasks, from text analysis to image generation and high-quality speech recognition (ChatGPT, Stable Diffusion, Whisper and DALL-E 2).

At the same time, identified errors in the use of AI raise questions about emerging ethical issues about regulation.

It remains to be decided whether countries will take the path of bringing AI to human ethical standards, which will severely limit the technology and reduce its effectiveness, or, in the interests of competition, its development will not be significantly limited.

Note that at the end of March, the Italian authorities announced a temporary shutdown of user access to the well-known chat bot ChatGPT due to the leakage of personal data. After that, it became known that Germany could also block access to ChatGPT for privacy reasons.

The US and China, the leaders in AI investment noted by Stanford experts, are increasingly competing with each other. The advantages of the Chinese approach (large-scale state support and free circulation of data) were recently proposed by the US Chamber of Commerce experts to counter light regulation and the introduction of common approaches to it together with other countries (see “Kommersant” dated March 10).

Another important topic for the field of AI is the attitude of the population towards this technology. According to an Ipsos survey, Chinese citizens are most positive about AI-based products and services – 78% of those surveyed versus 35% in the United States. Poll VTsIOMpresented at the end of 2022, 32% of Russians do not trust AI technologies (10 percentage points less than in 2021).

Venera Petrova