A major international report on the safety of artificial intelligence says experts cannot agree on the risks posed by the technology and it is unclear whether AI will help or harm us.
The report, chaired by Canada's Joshua Bengio, concludes that “the future trajectory of general AI is highly uncertain.”
The report states that “a wide range of trajectories, including both very positive and very negative outcomes, are possible even in the near future.”
The report was commissioned at last year's UK-hosted AI Safety Summit, the first global conference on artificial intelligence.
Britain asked Bengio, known as the “godfather” of AI and scientific director of the Quebec AI Institute Mira, to chair the report. It was released ahead of another global summit on AI to be held next week in Seoul, South Korea.
“We acknowledge that advanced AI is developing very rapidly, and that there is considerable uncertainty about how these advanced AI systems will impact the way we live and work in the future. We know that,” Bengio said in the report.
The UK government said in a press release on Friday that the report is the “first ever independent international scientific report” on the safety of AI and will play a “key role” in informing discussions in South Korea next week. We will fulfill our mission.”
A group of 75 experts contributed to the report, including a panel appointed from 30 countries, the European Union and the United Nations. The report released Friday is preliminary, with a final version expected by the end of the year.
We focus on general-purpose AI systems such as OpenAI's ChatGPT that can generate text, images, and videos based on prompts.
The report said experts “continue to disagree on several questions, large and small, surrounding the capabilities, risks, and risk mitigation of general-purpose AI.”
One area being discussed is the potential for “large-scale labor market impacts, the risk of AI-based hacking or biological attacks, and the risks of society losing control over general-purpose AI.”
The report outlines a number of risks, including the harm that AI can cause through fake content, disinformation, fraud and cyber-attacks. It also warns of the potential for risk bias in AI, particularly in “high-stakes sectors such as healthcare, recruitment, and financial lending.”
One possible scenario is that humans will no longer be able to control artificial intelligence, and therefore no longer be able to control the technology, even if it may cause harm.
The report says there is consensus that current general-purpose technology poses no such risks, but some experts argue that “acting, planning and pursuing goals” We believe that continued efforts to develop autonomous AI that can do this could lead to such an outcome.
“Experts disagree on how plausible an out-of-control scenario is, when it would occur, and how difficult it would be to mitigate it,” the report said.
This report by The Canadian Press was first published May 18, 2024.
