Artificial intelligence has popped up everywhere. Employers, police officers, and doctors use it. It even powers military drones and self-driving cars. Companies such as OpenAI and Google scramble to outdo each other with ever-more-powerful AI tools.
Now some experts ask: Is everything moving a bit too fast?
That’s the gist of a petition signed by some of today’s leading tech industry figures. They say it’s time to pause AI development and think through the risks.
The signers include billionaire entrepreneur Elon Musk and Apple co-founder Steve Wozniak. AI experts such as Yoshua Bengio, Gary Marcus, and Stuart Russell also signed.
“Recent months have seen AI labs locked in an out-of-control race to develop and deploy ever more powerful digital minds that no one—not even their creators—can understand,” says the petition. It calls for AI labs to stop training new systems for at least six months. Specifically, it asks them to stop training any system more complex than GPT-4.
In fact, the release of GPT-4 prompted this entire petition. But what is GPT-4? And what makes it such a big deal?
This AI-powered language model is the big brother of the famous ChatGPT. ChatGPT can answer human questions and produce convincing text. GPT-4 goes even further. It can understand pictures. It can even turn a hand-drawn sketch into a working website.
Some signers have concerns about the growing intelligence of AI. They worry it might become too powerful and too hard to control.
But for many signers, that fear still belongs in the realm of science fiction. Instead, they worry about AI’s impact on the job market. Tools like GPT-4 and the image-generating DALL-E threaten to steal jobs from writers, artists, and designers.
Others dread the spread of misinformation. Tools like ChatGPT can write believable text—but they don’t always get their facts straight. Criminals can use AI tools to run new kinds of scams. Already, some scammers create digital “voice deepfakes” to trick people over the phone.
“Current technology already poses enormous risks that we are ill-prepared for,” writes Gary Marcus, one of the signers. “With future technology, things could well get worse.”
Not everyone agrees with this petition. Some approve of the six-month pause, but they find the petition vague. They are also skeptical of Elon Musk’s involvement. Musk owns Tesla, a company that produces AI-powered self-driving cars. In the past, he has pushed back against the sort of accountability he’s now recommending.
Will OpenAI and other pioneers heed this petition? It seems unlikely. But many nations, such as the United Kingdom, have already made moves to regulate AI.
Even if AI development doesn’t stop, we can choose to slow down. We can count the costs for ourselves. The world asks: Will all these AI tools help or hurt?
Maybe that’s for us to decide.
One who is wise is cautious and turns away from evil, but a fool is reckless and careless. — Proverbs 14:16
(Tools like ChatGPT and GPT-4 from OpenAI prompt new concerns about the rise of artificial intelligence. AP/Michael Dwyer)