The UN wants to put artificial intelligence at the service of humanity

Artificial intelligence (AI) is not a field that can self-regulate, the host of the UN’s specialized agency for technologies pleaded before the start of a summit in Geneva bringing together experts, diplomats and human robots.

• Read more: Harvard is using an artificial intelligence that will almost replace the teacher

• Read more: If AI destroys humanity, how will it happen?

As research into AI, especially generative, continues to expand, the United Nations is calling on the international community to develop rules and safeguards to ensure that these technologies benefit humanity rather than endanger them.

American Doreen Bogdan-Martin, secretary-general of the International Telecommunication Union, told reporters at the end of the summit that “I hope we will come up with a clear plan to move forward.”

The “World Summit on AI for Social Benefit”, organized by the ITU on Thursday and Friday, is expected to bring together around 3,000 participants, including industry leaders, academics and policymakers.

Ms Bogdan-Martin called for finding a “balance” to “restrict innovation without limiting it”. But “doing nothing is not an option. Humanity depends on it.”

Summit attendees include Amazon CTO Werner Vogels, Google DeepMind COO Lila Ibrahim and legendary Real Madrid and Spain soccer goalkeeper Iker Casillas.

From fighting fires to providing humanitarian aid to supporting medical care, health and sustainable agriculture, they will be joined by a range of specialized robots, including nine humanoid robots.

The United Nations hopes to use AI to advance its goals on issues such as health, climate and poverty. According to its Secretary-General Antonio Guterres, “this summit will help ensure that artificial intelligence will benefit humanity”.

“Moment in History”

At great technical complexity, AI systems are too impressive to worry about. Although they could save lives by enabling a quantum leap in medical diagnosis, they are exploited by authoritarian regimes to engage in mass surveillance of citizens.

The spread of false images on social networks, more real than life, has warned of the dangers of opinion manipulation and dangers to democracy.

There are no international rules on AI, but UNESCO prepared a recommendation on ethics in 2021. And at the end of March, hundreds of academics, employers and figures said they needed six months to develop more powerful AI systems. “Major Threats to Humanity”.

According to AI expert Gary Markus, “this powerful new technology is spreading so widely and forcing itself into our lives, when we’re not really ready”.

“We are at a critical moment in history where we can either succeed and put in place the global governance we need, or fail and find ourselves in a situation where a few institutions control the destiny of a very large number of people.” He said during a presentation at the ITU Summit.

For its part, the EU hopes to finalize the world’s first regulation aimed at driving innovation in AI by the end of the year. But more than 150 major companies have warned the plan risks damaging EU competitiveness, particularly against the US.

Gabriela Ramos, who oversaw the development and adoption within UNESCO of the first global tool to promote the ethics of AI, said regulation was necessary rather than attracting investors.

On the other hand, he does not think it is necessary to create an international organization to oversee AI, but calls on countries to set up institutions such as drug regulatory authorities, which will be responsible “at least before allowing major developments”. in the market”.

These technologies, he told reporters, must be scrutinized to ensure they comply with “human rights” and not just commercial and geopolitical considerations.