United Nations

Artificial Intelligence: 'What we do with it is now up to all of us' - UN Chief | United Nations

Introducing the Preliminary Report of the Independent International Scientific Panel on Artificial Intelligence, Secretary-General António Guterres (1 Jul) said, “the science is here. We can no longer say we did not know. What we do with it is now up to all of us.” Guterres told reporters in New York that the report “is honest about the extraordinary promise of this technology.” Used well, he said, “AI could be the most powerful engine for development, speeding the world's progress on everything from health and hunger to learning and climate,” but warned that the Panel was “just as clear eyed about the harm artificial intelligence can cause.” The Secretary-General said, “the more AI advances without shared rules, the less say governments and people will have in the outcome.” He stressed that “this is a preliminary report” and the Panel “will keep working as the technology evolves.” Joining via video teleconference, Panel Co-Chair Maria Ressa said, “what you are receiving is the floor of our concern, not the ceiling. You know when you get 40 scientists from 40 different contexts and you have to make them agree, you don't drift toward the most alarming claim. You build toward the centre.” Ressa said, “the most contested findings demand the most evidence and the most negotiation. So, everything in this report cleared that bar. It is the minimum we all agree on. And that is alarming enough.” She said, “without facts, you can't have truth. Without truth, you can't have trust. Without these three, we don't have a shared reality. We have no democracy. That's the trend we're seeing, right? That is not speculation. It is in the report with the evidence attached. But we're not telling you the future is written because it isn't.” The other Co-Chair, Yoshua Bengio said, “the decisions we're making today about I will have lasting consequences for individuals, businesses, institutions, democracy, and maybe more than that, our goal here is to help understand AI’s, opportunities, risks, and impacts.” Bengio said, “there's still no known technical guarantees that AI would follow our instructions, norms, or laws. And instead, as she mentioned, growing evidence to the contrary, including AI acting deceptively.” On the other hand, he continued, “it wasn't also the case that I was unlocking useful applications across domains like health, agriculture, science, while its adoption and its benefits, are accelerating unevenly across countries and sectors.” As AI capabilities continue to grow, Bengio said, “the gap between the haves and the have nots is likely to grow,” and added that “to realise the full benefits of AI while minimising its risk, we need to tackle all of these issues and it has to be at the international level because AI, and in particular it's misuses like cyber-attacks, are not limited by borders.” Responding to a journalist’s question, he said, “there are ways to develop AI that will be heavily focussed on, honest answers and factuality. It may not be the most profitable way to develop AI, but it is possible to make those choices collectively. We can choose in which direction we can take the technology. That technology is not some God given thing. It is developed by humans under societal rules and incentives.” The first global dialogue on AI governance will be held in Geneva on 6–7 July.