Rosalind Evans
A new alert regarding the potential dangers artificial intelligence (AI) might bring to humanity was shared on Tuesday by industry leaders and scientists, including high-ranking officials from Microsoft and Google.
“The threat of extinction caused by AI needs to be globally prioritized, similar to other wide-scale societal risks such as pandemics and nuclear warfare,” the announcement stated.
Among the many influential individuals endorsing this message were Sam Altman, CEO of ChatGPT creator OpenAI, and Geoffrey Hinton, a renowned computer scientist often dubbed the godfather of artificial intelligence. The declaration, posted on the Center for AI Safety’s website, had hundreds of signatures.
Fears of AI systems overpowering human intelligence have escalated with the emergence of advanced AI chatbots like ChatGPT. This development has instigated a global rush to create regulations for this burgeoning technology. The European Union is leading this effort with its proposed AI Act expected to be ratified later this year.
The recent alert was deliberately concise — a single sentence — aiming to unite a wide range of scientists who might disagree on the specific threats or the optimal solutions, according to Dan Hendricks, Executive Director of the Center for AI Safety based in San Francisco, the organization that coordinated this initiative.
“Various top-tier university representatives across different fields are alarmed by this and believe it’s a global priority,” said Hendrycks. “So we aimed to encourage people to openly voice their concerns, as many had only been discussing it privately.”
Earlier this year, over 1,000 technologists and researchers, including Elon Musk, penned a more detailed letter advocating for a six-month hiatus on AI development, citing “significant risks to society and humanity.”
This letter was in response to OpenAI’s introduction of a new AI model, GPT-4. However, leaders at OpenAI, Microsoft, and Google declined to sign and disregarded the appeal for a voluntary industry hiatus.
In stark contrast, the latest declaration received support from Microsoft’s chief technology and science officers, Demis Hassabis, CEO of Google’s AI research lab DeepMind, and two Google executives overseeing AI policy.
Some skeptics argue that extreme existential risk warnings issued by AI creators have added to the overstatement of their products’ abilities and diverted from the push for immediate regulations to address their real-world issues.
Hendrycks suggested that society could effectively handle the “immediate, ongoing harms” caused by AI products that generate new text or images while also beginning to confront the “potential catastrophes looming ahead.”
Drawing a comparison to nuclear scientists in the 1930s cautioning people even before the creation of the nuclear bomb, he said, “Nobody is suggesting that GPT-4 or ChatGPT is currently causing these types of concerns. We’re trying to address these risks before they occur rather than dealing with disasters post-occurrence.”
Experts from nuclear science, pandemic response, and climate change also signed the letter. Among these signatories was author Bill McKibben, who raised awareness about global warming in his 1989 book “The End of Nature” and cautioned about AI and related technologies two decades ago in another work.
“Given our failure to heed early warnings about climate change 35 years ago, it would be wise to consider this issue carefully before it’s too late,” McKibben commented via email.
As AI continues to evolve and shape our world, these concerns remind us of the potential risks that could come with these advancements. While the alarm is being raised on a potential existential threat, the importance of maintaining the balance between innovation and safety regulations cannot be overstated. The focus is not just on preventing AI from causing a catastrophe but also ensuring it is used responsibly and ethically. Indeed, we must heed these warnings, like with climate change, to prevent potentially severe consequences.
