Artificial intelligence can cause significant damage to the world: says the dad of ChatGPT
robort - 2023-06-03 08:40:11
In US Senate testimony, OpenAI CEO Sam Altman called for action by lawmakers to regulate the AI industry and the companies that crowd it.
“ We think regulatory intervention by governments will be critical to mitigating the risks of increasingly powerful models,” Altman said ( via Arstechnica ). “ For example, the US government might consider a combination of licensing and testing requirements for the development and release of AI models that exceed a certain threshold of capability .”
Altman said he was "quite concerned" that the 2024 US election would be affected by AI-generated content. “ As we face elections next year and these patterns are improving, I think this is a significant area of concern… I think regulation on this would be quite wise,” Altman said.
Christina Montgomery, IBM's chief privacy and trust officer, also espoused the OpenAI CEO line. “ IBM urges Congress to take a thorough regulatory approach to AI. That means establishing rules to govern the implementation of AI in specific use cases, not regulating the technology itself .”
Montgomery believes that Congress must clearly define the risks of AI and impose " different rules for different risks ", with the strictest ones " applied to use cases with the greatest risks to people and society ".
The position of OpenAI and IBM has sparked a certain surprise among those present, so much so that Senator Dick Durbin said that it is extraordinary to see large companies present in the Senate demanding a commitment to regulate them. That they know they have something on their hands that could get out of control?
Altman has suggested that Congress form a new agency that licenses AI technology “ above a certain scale of capability” and that it “may revoke that license to ensure it meets safety standards.”
Also, before releasing an AI system to the public, there should be independent audits by “ experts who can say whether or not the model is compliant ”.
The possibility that the AI industry could " cause significant harm to the world," according to Altman, exists. “ I think if this technology goes wrong, it can go completely wrong, and we want to be explicit about that, ” Altman said." We want to work with the government to prevent this from happening ."
The licenses invoked by the OpenAI CEO should be sought for those models that " can persuade, manipulate, influence a person's behavior, a person's beliefs" or "help create new biological agents ".
Finally, Altman commented negatively on the words of those who have requested a pause in the development of AI ( such as Musk and Wozniak ).
According to him, in fact, it is better to publish incremental improvements: " I think it would not be good to build a super powerful artificial intelligence system in secret and then launch it on the world all at once ".
“A big part of our strategy, in this time when systems are relatively weak and deeply flawed, is to find ways to get people to experience them…and figure out what we need to do to make them safer and better,” he said. added Altman.
Senator Richard Blumenthal also believes there is no point in trying to stop the development of artificial intelligence. “ The world will not wait, the rest of the global scientific community will not stop. We have opponents who are moving forward and putting their heads in the sand is not the answer,” the politician said.
The impact of AI on jobs was also discussed during the hearing. IBM CEO Arvind Krishna recently unveiled plans to suspend hiring for approximately 7,800 jobs potentially replaceable by artificial intelligence systems.
In this regard, Montgomery said that " new jobs will be created, many more jobs will be transformed and some jobs will be relocated ". Altman, for his part, believes that " GPT-4 will fully automate some jobs and create new ones that we believe will be much better ".