AI2050 Charitable Foundation

History

2022: Creation of a charitable fund with a total capital of $125 million

On February 16, 2022, it was reported that the former executive director Google Eric Schmidt (Eric Schmidt) announced the creation of a charitable fund with a total capital of 125 million, dollars which will assist research in the field. artificial intelligence We are primarily talking about research aimed at solving the cardinal problems manifested in the use of artificial intelligence technologies, including bias (- AI bias phenomenon approx.), TAdviser The possibility of harm and abuse. The list also includes geopolitical conflicts and the scientific limitations of the technology itself.

Former Google CEO Eric Schmidt is creating a fund to solve "key" AI problems
Photo: Reuters/Financial Times

The fund, called AI2050, plans to pay the entire amount over the next five years. In addition to Schmidt himself, the fund is also led by James Maniika, head of Google's technology and society division.

As Schmidt himself told the Financial Times, the list of cardiac AI problems was formulated on the model of Hilbert's Problems, a list of 23 cardinal mathematics problems presented by David Hilbert (Hilbert) in 1900 at the II International Congress of Mathematicians in Paris.

"A lot of people are expressing concerns (about AI), but few are really working to solve (cardinal problems). If we can find the next generation of researchers capable of perfectly timely discoveries in these areas, this will be a great result, "Schmidt said.

Business, governments, members of the scientific community and civic activists widely discuss how AI will affect the future state and development of society and how to neutralize the most "toxic" consequences of its widespread introduction. Among them are the malicious creation of "deep-fakes" and the manipulation of public opinion through would-be social networks.

The goal of the AI2050 fund is to contribute to the creation by 2050 of such AI technologies, the impact of which on society will become universally favorable. The Fund will finance projects that will allow the use of AI to objectively assess and eliminate socioeconomic inequality. In addition, funding has been promised to projects developing so-called "liquid" neural networks that can change their parameters in real time, significantly improving their ability to analyze changing data.

Developers of artificial intelligence technologies, as noted by the Financial Times, are often criticized for not taking into account ethical components when creating AI and making mistakes in their programming and training, which lead to numerous manifestations of bias and bias in machine decision-making.

Schmidt points out that his goal is to help avoid mistakes made in developing "today's technologies."

"I don't think we understand the impact of social media on society - both positive and negative. And AI has the potential to have an even greater positive and negative impact, simply because of its ability to understand and manipulate human behavior and beliefs, "said Schmidt.

His fund will not fund developers hired by large corporations, although according to Schmidt himself, AI research, which began in academic environments, is easily and often commercialized.

This is also due to the fact that even at the largest universities at the Stanford or Oxford level, it is very difficult to get funding for AI research. But corporations willingly invest in them, but they are interested in narrow-practical goals and solutions, and not global issues.

This is not Eric Schmidt's first investment in this area, for example, a year ago he invested about $150 million in AI-based biological research. The emphasis on interaction with independent researchers allows us to get away from the development of purely applied developments of commercial companies and concentrate on promising technologies, often without an obvious scenario of rapid commercialization, "said Dmitry Gvozdev, general director of Information Technologies of the Future. This - on the one hand - is an investment in the development of the direction as a whole, on the other - an opportunity to gain control over potentially promising technology. And when Erich Schmidt warns in his articles and books about the threats and opportunities that lie in the penetration of AI into various branches of human life, he knows what he is talking about - many of these threats and opportunities are formed with his participation.

The first grants from the AI2050 Foundation have already received six researchers. Google is not involved in financing these grants.^[1]

Notes