OpenAI CEO Sam Altman warns that other AI developers working on tools like ChatGPT won’t set safety limits — and the clock is ticking

In an ABC news interview This week, he warned “there will be others who don’t put in place some of the safety limits that we have in place.”

OpenAI released its AI ChatGPT chatbot to the public at the end of November and this week it reveal a more capable successor called GPT-4.

Other companies are racing to provide tools like ChatGPT, leaving OpenAI to worry about a lot of competition, albeit with the advantage of having Microsoft like a big investor.

OpenAI co-founder and chief scientist Ilya Sutskever told The Verge in an interview: interview published this week. “GPT-4 is not easy to develop… there are a lot of companies that want to do the same, so from a competitive perspective, you can consider this the maturity of the field.”

Sutskever is explaining OpenAI’s decision (with safety reasons being another) to reveal very little about the inner workings of GPT-4, leaving many question whether the name “OpenAI” still makes sense. But his comments are also an acknowledgment of the wide range of opponents that are stalking OpenAI.

Altman suggests that some of those competitors may be far less interested than OpenAI in placing barriers on their ChatGPT or GPT-4 equivalents.

“The thing that worries me is… we won’t be the only ones making this technology,” he said. “There will be others who don’t set some of the safety limits that we put in place. I think society has a limited amount of time figuring out how to react to that, how to adjust to that, how to handle it.”

OpenAI this week shared a document “system card” which outlines how their testers knowingly tried to get GPT-4 to deliver dangerous information, such as how to make a dangerous chemical using basic ingredients and household items kitchen, as well as how the company fixes problems before the product launches.

For fear of anyone suspecting the bad intentions of bad guys looking to AI, phone scammers are now using voice-cloning AI tools to pretend like people’s loved ones are in dire need. financial help—and Successfully extorting money from victims.

“I am particularly worried that these models could be used for large-scale misinformation,” Altman said. “Now they are getting better at writing computer code, [they] can be used for cyberattacks.”

Considering he leads a company that sells AI tools, Altman was particularly revealing about the dangers posed by artificial intelligence. That might have something to do with History of OpenAI.

OpenAI was founded in 2015 as a non-profit organization focused on the transparent and secure development of AI. OpenAI switched to a hybrid “limited profits” model in 2019, with Microsoft becoming a major investor (the amount of profit it can get from the deal is limited, as the model’s name suggests). suggest).

Tesla And Twitter CEO Elon Musk, who is also the co-founder of OpenAI—and who has donated heavily to it—criticized this change, note last month: “OpenAI was created as an open source (which is why I named it “Open” AI), a non-profit company to serve as a counterweight to Googlebut now it has become a closed-source, profit-maximizing company effectively controlled by Microsoft.”

Early December, Musk calls ChatGPT “frighteningly good” and warned, “We’re not far from dangerously powerful AI.”

But Altman has warned the public as much, if not more, even as he is promoting OpenAI’s 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 organizations to figure out what to do,” he wrote. “Regulation is going to be very important and it will take time to learn… having time to understand what is going on, how people want to use these tools, and how society can co-evolve is very important. .”