Silicon Valley prophet warns against artificial intelligence: I think we’re all going to die

Technician

From Ariel Zilber

July 13, 2023 | 1:32 pm

A prominent Silicon Valley researcher is sounding the alarm about the rapid advance of artificial intelligence, saying the consequences for humans could be dire.

I think we weren’t ready, I think we don’t know what we’re doing, and I think we’re all going to die, AI theorist Eliezer Yudkowsky, considered particularly extreme by his tech peers, told Bloomberg News.

Yudkowsky echoed concerns voiced by the likes of Elon Musk and other tech figures who have advocated for a six-month hiatus on AI research.

Musk said earlier this year that there is a non-zero chance that AI could become Terminator over humanity.

Where we are right now is exciting, Yudkowski, a researcher at the Machine Intelligence Research Institute, told Bloomberg.

Where we’re going is far more important than that, I’d say.

Yudkowski said he’s concerned that data analysts don’t fully understand the component parts of the latest edition of GPT, the AI-powered language model developed by Silicon Valley firm OpenAI.

It went a little further than I expected, said Yudkowski, who wrote an essay in Time earlier this year advocating air strikes against rogue data centers, about AI research.

I think you get smarter than us.

Yudkowski said AI is acquiring knowledge at a faster rate than humans can comprehend.

The capabilities of how smart these things are have been far ahead of advances in understanding how they work and being able to model their behavior in detail, he said.

We have no idea what’s going on inside GPT-4. If humanity had any common sense, we’d be pouring $100 billion a year into figuring out what’s going on in there.

But the few people who work on it aren’t making progress at a rate where even if we put half the graduate physicists on this, I wouldn’t have much faith that we’d know what’s going on in there in three years. [or] five years, he said.

Concerns about AI systems outsmarting humans and running amok have intensified with the rise of a new generation of highly capable AI chatbots like ChatGPT.

It has sent countries around the world scrambling to come up with regulations for developing technology, with the European Union leading the way with its AI Act due to pass later this year.

Some critics have complained that dire warnings about existential risks voiced by AI makers have helped boost the capabilities of their products and distract from calls for more immediate regulations to curb their real-world problems.

A new Netflix documentary reveals that world governments are in the midst of an arms race to develop AI-controlled weapons.