Musings: Convene before AI runs amok
A billboard for AI at the PT Expo on Sept. 22, in Beijing, China. Credit: Getty Images/Kevin Frayer
News about the dangers of artificial intelligence as an existential threat should not be dismissed by leaders. AI companies and scientists should organize a meeting like the Asilomar Conference on Recombinant DNA, which molecular biologists convened in 1975. Talks centered on biohazards and regulation of biotechnology, aiming to create voluntary safety rules for recombinant DNA research and genetic engineering. The Asilomar worries were mostly potential risks.
The conference produced safety principles, many still in place today. Thus, it behooves AI companies and scientists to model their regulations and safety guidelines on the principles established at Asilomar.
The existential threat of AI is real, and just because we can do something does not mean we should. Establishing safety guidelines for AI, especially as companies race toward self-improving superintelligence, would be a step in the right direction. Failure to do so may put us on the path to destruction.
— Michael Hadjiargyrou, Centerport
WE ENCOURAGE YOU TO JOIN OUR DAILY CONVERSATION. Just go to newsday.com/submitaletter and follow the prompts. Or email your opinion to letters@newsday.com. Submissions should be no more than 200 words. Please provide your full name, hometown, phone number and any relevant expertise or affiliation. Include the headline and date of the article you are responding to. Letters become the property of Newsday and are edited for all media. Due to volume, readers are limited to one letter in print every 45 days. Published letters reflect the ratio received on each topic.
