Since 2017, the United Nations’ AI for Good conference in Geneva has brought together researchers, NGOs and tech leaders to discuss the opportunities and challenges of artificial intelligence. Frédéric Werner, the summit’s founder, defends taking a balanced approach
The ambition of this week’s AI for Good summit in Geneva is bold, aiming to unlock artificial intelligence for the benefit of humanity. Organised annually by the International Telecommunication Union (ITU) and Switzerland, the conference brings together hundreds of researchers, NGO representatives and senior executives from leading tech companies, and also comes ahead of next year’s global AI summit, also to be held in Geneva
Frédéric Werner, head of strategic engagement at the ITU and creator of AI for Good, explains the goals of an event that first launched in 2017 and which is set to welcome 12,000 participants this year
Le Temps: Why did you decide to create AI for Good?
Frédéric Werner: Nine years ago, there was a great deal of discussion about the promise of artificial intelligence, but very little substance. We saw countless marketing presentations, even though the technology was not yet truly ready. That did not stop us from thinking about how AI could help tackle major global challenges: healthcare, education, clean energy and disaster response. We also wanted to prevent the technology from widening the existing digital divide. At the same time, we began asking what standards would be needed to support these high-impact applications once they became more widespread.
The ITU is a rather unique agency within the UN system because our membership includes both the public and private sectors. More than 1,000 private companies are members of the ITU and contribute to our standard-setting work. We take a multistakeholder approach, engaging academia, civil society and NGOs. We have also made considerable efforts to ensure that developing countries have a seat at the table
Over its 160-year history, the ITU has navigated many waves in technological change: the telegraph, dial telephones, switched telephone networks, satellites, radio, television, 3G, 4G, 5G and soon 6G. AI is simply the latest technological wave
The name AI for Good suggests that AI is also used for harmful purposes. How can those risks be minimised?
As with quantum computing, another emerging technology, the security implications are significant. But AI could also accelerate drug discovery, advance materials science by helping us better understand the properties of new materials, improve weather forecasting and enable many other major breakthroughs. The goal should be to ensure that people without direct access to these technologies can still benefit from them
Isn’t the influence of major US technology companies too dominant in Geneva?
Take a look at our list of speakers — it takes about ten minutes to scroll through because there are 1,250 of them. Of course, some big names immediately stand out, such as Salesforce, Amazon, Microsoft and Google. But altogether, only around 60 companies are represented. And they are not all American: roughly one-third come from Europe and another third from Asia
AI for Good is the only truly open and neutral platform where everyone is welcome. I like to describe it as the Davos of AI – but unlike Davos, which is highly exclusive, AI for Good is extremely inclusive
Still, the spotlight will inevitably be on the American tech companies, all of which will appear on the main stage.
On our main stage, we deliberately explore both sides of the debate. Yes, we are welcoming Microsoft’s Brad Smith, Salesforce’s Marc Benioff and other business leaders. But we are also giving the same platform to people who are highly critical of big tech
We will hear from scientists examining AI’s environmental impact, experts focused on AI safety, advocates calling for a slower pace of development, and researchers who take a deeply critical view of Big Tech. We are also hosting Tristan Harris, who has spoken forcefully about the effects of social media on young people, as well as specialists on algorithmic bias and facial recognition
Even under the banner of AI for Good, we have always sought to present both sides of the coin. But I believe public debate is already overwhelmingly focused on the dangers of AI. We also need to discuss its positive potential
If AI can help discover the next cancer treatment, predict the next tsunami or reduce humanity’s enormous energy consumption, then it may well be one of our greatest hopes for solving these complex challenges. That may sound ambitious, but I believe we need to have both conversations
Can, and should, the United Nations play a role in regulating AI?
For the first time, UN member states have received a mandate, based on a UN resolution, to discuss the future of artificial intelligence together. Until now, there have been many conferences, many discussions and many important conversations
The recently established Independent International Scientific Panel on AI is separate from the political process. Its role is to provide scientific advice to policymakers on public policy and regulation. Its report now serves as the basis for dialogue. States will be able to discuss AI’s implications for public policy, future regulation and governance frameworks
Will this immediately lead to global governance or worldwide AI regulation? I would say no – it is still too early. But this process will become one of the most important tools for informing those discussions
It is also important to remember that countries will adopt their own regulations at different speeds and in different ways. Even so, bringing scientific expertise into the policy debate is an excellent first step
This article was originally published in French in Le Temps. It has been adapted and translated into English by Geneva Solutions. Articles from third-party websites are not licensed under Creative Commons and cannot be republished without the media’s consent
