The biggest warnings about AI from the people who design it

As the New York Times reports, Geoffrey Hinton recently quit his job as a researcher at Google, where he’d been working for more than a decade, so he could talk freely about the dangers of AI without having to consider how it might affect his former employer.

Hinton and two of his students—one of whom became the chief scientist at OpenAI—built a neural network that could analyze thousands of photos and teach itself to identify common objects within them. That startup got bought by Google for $44 million, according to the Times, and helped pave the way for the creation of chatbots like ChatGPT and Google’s Bard.

Hinton’s immediate concern is that the internet will be flooded with fake photos, videos, and text, so the average person won’t “be able to know what is true anymore.” He also said that AI technologies will put jobs at risk and that future versions will learn unexpected behaviors that could threaten humanity.

Musk has said that AI comes with both great promise—and great danger. Like some other technology leaders, he supports government regulation of AI before it becomes too late.

“AI is more dangerous than, say, mismanaged aircraft design or production maintenance or bad car production, in the sense that it is, it has the potential—however small one may regard that probability, but it is non-trivial—it has the potential of civilization destruction,” Musk said in an interview with Tucker Carlson last month.

Meanwhile, the billionaire has announced he’s creating an AI offering to rival Microsoft and Google’s, called TruthGPT.

The CEO of OpenAI has issued a number of warnings about ChatGPT. “We’ve got to be careful here,” Sam Altman said in an interview with ABC News. “I think people should be happy that we are a little bit scared of this.”

Like others, Altman expressed concerns about generative AI being used as a tool for large-scale disinformation, given that it’s getting better at writing code; and about bad actors, because not everyone building it will include safety filters to mitigate the risks.

Regulators and society need to be involved so their feedback on the technology can help avoid negative consequences for humanity, Altman said. Overall, he argued, the benefits of AI will outweigh its risks.

Sundar Pichai, CEO of Google, has said that AI will “scale” the spread of disinformation, creating bigger problems with fake news and images. “[On] a societal scale, you know, it can cause a lot of harm,” he explained in an interview with 60 Minutes.

But Pichai also pointed out that compared to other technologies, he’s seen more people worried about it earlier in its life cycle, which leads to more serious conversations. “So I feel optimistic,” he said.

This group’s online letter, which includes signatures from Apple co-founder Steve Wozniak and Elon Musk, warns that AI tools present “profound risks to society and humanity.” The nonprofit hosting it, the Future of Life Institute, has also published letters on nuclear and autonomous weapons.

The letter calls for a pause in development of AI systems more powerful than OpenAI’s GPT-4. In that time, AI labs and independent experts could jointly develop and implement shared safety protocols that are independently audited.