Shulou(Shulou.com)11/24 Report--

At a time when the future and dangers of artificial intelligence are increasingly concerned, more than 2000 researchers and engineers from around the world gathered in Rwanda this week to debate a different vision of the future of artificial intelligence.

One vision is to build more powerful systems such as ChatGPT, designed to transcend human intelligence to increase worker productivity and economic growth. Another vision is to create small, more targeted AI solutions to address global challenges such as climate change, improving health care, and protecting biodiversity.

This is the first major artificial intelligence research conference held in Africa this year, and one of the highlights of the conference is more exchanges of ideas. The organizers hope that researchers from the United States, China and wealthy companies will be more aware of the social problems facing Africa, while allowing African researchers to participate in the discussion and express their unique views.

Joshua Yoshua Bengio, one of the godfathers of artificial intelligence, is a member of the board of conference organizers and advocates the introduction of artificial intelligence technology to Africa: "it is clear that in order for everyone to enjoy the potential benefits of artificial intelligence, we need everyone to participate."

The chat robot ChatGPT became popular a few months before the meeting, adding urgency to the discussion about the trajectory of artificial intelligence and its short-and long-term implications. The release of ChatGPT has sparked a global craze for generative artificial intelligence (AIGC), prompting large technology companies such as Google and Baidu to launch similar products of their own, which can generate text and images, sparking fears of job replacement and the spread of misinformation.

Although OpenAI, the developer of ChatGPT, did not show up at the artificial intelligence research conference held in Kigali, Rwanda, it still became the focus of attention. Do you accept or resist its development direction? The views of the global artificial intelligence community are polarized.

Recently, another "artificial intelligence godfather", Geoffrey Hinton (Geoffrey Hinton), announced that he would leave Google to talk more freely about the risks of artificial intelligence development. In an interview with the media, Hinton said he was worried that the technology posed a threat to the long-term survival of mankind.

At the conference, many researchers from Africa and other developing countries said they were more concerned that artificial intelligence posed a direct challenge to their societies. They worry that current trends in AIGC models could exacerbate the dominance of the US and China in the development of artificial intelligence, creating two problems: Africa is either left behind or has to deal with its problems. Such consequences include more false information in African elections and the gradual disappearance of their mother tongue in digital technology.

Many researchers also worry that the development of useful artificial intelligence solutions that could have helped improve the basic quality of life of people around the world will lag behind.

Girma Abebe Tadesse (Girmaw Abebe Tadesse), an Ethiopian researcher at Microsoft's Nairobi office in Kenya, highlighted key data issues, such as misfilling out medical forms, that hinder the progress of artificial intelligence in developing countries and have not yet helped to improve maternal health and reduce child mortality.

Tadesse also shared some successful cases, such as using a combination of high-quality data and statistical analysis to find that child mortality in southern Nigeria was lower than the country's average, which allowed researchers to conduct more targeted investigations, find out why, and design solutions for other places.

Other participants also presented applications such as the use of satellite images to understand ethnic differences in access to parks and sanitation, the use of sensors and mathematical models to improve grid maintenance, and the use of computer vision to detect agricultural diseases.

Bengio said he hopes this year's conference will make researchers focus more on artificial intelligence for social welfare applications, rather than simply pursuing profits. "We've talked a lot about the risks of artificial intelligence, but we haven't talked enough about how to make better use of artificial intelligence," he said. "

Conference organizers said there were 261 participants who thought they were from Africa, compared with 16 in 2019. Over the years, the largest and most prestigious annual artificial intelligence research conferences have been held in the United States or Canada because of their proximity to Silicon Valley. Because of visa problems, African researchers are often unable to attend meetings because it is difficult to obtain visas. This has led to a lack of African perspective in the development of the most powerful and transformative technologies.

Well-known researchers, including artificial intelligence ethicist Timnit Gebru, point out that research tends to focus on a small number of dominant participants in Silicon Valley, with a lack of researchers from non-Western or marginalized groups.

Gebru grew up in Ethiopia and later came to the United States as a refugee. In 2017, she founded the non-profit organization Black in AI, which aims to bring more diversity to the artificial intelligence community.

Gebru hosted a panel discussion last Friday on the limitations of large language models in dealing with African languages. These models are the underlying technologies for building ChatGPT, but African languages are increasingly excluded from the digital world because of the lack of data needed by current artificial intelligence technologies. Gebru says she co-authored a paper criticizing the explosive growth in resource demand and the impact on the environment of the model, which also underpins Google's search engine. Previously, Gebru was fired by Google and attracted public attention. Within Google, Google characterized her departure as resignation.

In a separate panel discussion, Vukosi Marivate, a data science professor at the University of Pretoria in South Africa and chairman of the conference project, said African researchers were grappling with the threat of losing their mother tongue. "We are racing against time" before English dominates and African languages no longer exist, he said after the event.

Although Kigali is thousands of kilometers from Silicon Valley, one cannot escape the temptation of large language models and AIGC tools such as ChatGPT. The researchers filled the lecture halls for the meetings, sitting on the floor and leaning against the wall, while others rushed straight from the airport and sat on suitcases.

Many researchers say they are surprised by OpenAI's findings, prompting them to actively expand their models. One doctoral student attending the meeting said he planned to shift his focus to AIGC and join the most exciting research direction. "this will be the last meeting that is not dominated by large-scale language model research," he said.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.