OpenAI CEO Sam Altman believes artificial intelligence has a tremendous impact on society, but he also worries about how bad actors will use the technology.
In an ABC News interview this week, he warned “there are going to be other people who don’t apply some of the security restrictions that we do.”
OpenAI released its AI chatbot ChatGPT to the public at the end of November, and this week unveiled a more likely successor called GPT-4.
Other companies are racing to offer tools like ChatGPT, giving OpenAI plenty of competition to worry about, despite the benefit of Microsoft as a big investor.
“It’s competitive,” OpenAI co-founder and chief scientist Ilya Sutskever told The Verge in an interview published this week. “GPT-4 is not easy to develop … there are many companies that want to do the same thing, so from the competitive side, you can see this as a maturation of the field.”
Sutskever explained OpenAI’s decision (with safety as another reason) to reveal little about the internal workings of GPT-4, leading many to question whether the name “OpenAI” still exists. But his comments are also an acknowledgment of competition that is killing on OpenAI’s heels.
Some of these competitors may be less worried than OpenAI about creating the same fences as ChatGPT or GPT-4, Altman suggested.
“The one thing I’m worried about is … we’re not going to be the only ones to come up with this technology,” he said. “There are other people who don’t implement some of the security restrictions that we do. The public, I think, has a limited time to determine how to respond, how to manage, and how to deal with it.
OpenAI this week shared a “system card” document that explains how testers deliberately tried to get GPT-4 to provide dangerous information, such as how to make dangerous chemicals using basic materials and kitchen supplies, and how the company fixed the problem. before the product launch.
Lest anyone doubt the nefarious intentions of bad actors looking for AI, phone scammers are now using voice-cloning AI tools to sound like relatives of people in need of financial help—and extort money from victims.
“I worry that the model could be used for large-scale disinformation,” Altman said. “I’ve gotten better at writing computer code now, [they] can be used for offensive cyber attacks.”
Considering he heads a company that sells AI tools, Altman is well aware of the dangers posed by artificial intelligence. This may have something to do with the history of OpenAI.
OpenAI was founded in 2015 as a non-profit organization focused on safe and transparent AI development. model name suggests).
Tesla and Twitter CEO Elon Musk, who is also the founder of OpenAI—and who has contributed heavily—has criticized this change, last month’s notes: “OpenAI was created as an open source (that’s why I named it “Open” AI), a non-profit company to be Google’s counterbalance, but has now become a closed source, maximum profit company effectively controlled by Microsoft.”
In early December, Musk calls ChatGPT “great” and warned, “We are not far from a very dangerous AI.”
But Altman has warned the public just as much, if not more, even as he moves forward with his OpenAI work. Last month, he worried about “how people in the future will see us” in a series of tweets.
“We also need enough time for our institutions to understand what to do,” he wrote. “Regulations will be critical and it will take time to understand … having time to understand what’s going on, how people want to use these tools, and how society can evolve is critical.”