Sam Altman says that it is “the greatest technology that humanity has developed so far.”
OpenAI CEO Sam Altman admitted he was “a bit scared” about creating ChatGPT and what people might use it for, but stressed that it could be the biggest technology ever developed.
“We have to be careful here,” Altman said Thursday in an interview with ABC News, adding that he is “particularly concerned that these models could be used for the large scale misinformation“, since now that “these systems are getting better at writing computer code, they could be used for offensive cyberattacks.”
The programmer says that the potential of ChatGPT is huge and very exciting, and he highlighted the need for society to be as involved as possible, considering that feedback will help deter attacks. possible negative consequences that technology could have on humanity.
“I think people should be happy that we’re a little scared about this,” he said, warning that “there will be people who don’t put some of the safety limits” that has been forced on it. “I think society has a limited amount of time to figure out how to react to that, how to regulate it, how to manage it,” he added.
Altman, 37, also mentioned that despite fears that artificial intelligence (AI) could replace many jobs — which he says is possible — it is a tool that largely functions “under the control of humans“, so it could also lead to “much better ones”.
“The reason for developing AI is in terms of impacting our lives and improving our lives, and on the upside, this will be the greatest technology that humanity has ever developed until now,” he said.
This week the company presented the fourth version of the popular artificial intelligence system, which, according to the developers, is more powerful. The creators claim that the new tool named GPT-4, in addition to text, accepts images and also outputs text, allowing you to chat with the system by inputting complex files, such as entire books to summarize them, in order to do so. faster, more precise and humane.
The businessman acknowledged that this version uses the deductive reasoning instead of memorization, a process that can lead to strange responses.
“What I try to warn people about the most is what we call the ‘hallucination problem,'” he said, since “the model will confidently state things as if they were completely made-up realities.”
“The correct way to think about the models we create is a reasoning engine, not a database of facts,” he said. While the technology could act as a database of facts, she explained that’s not what’s “special.” “What we want them to do [los sistemas] is something closer to the ability to reasonnot to memorize”.
In relation to education, Altman assures that “it will have to change” as it has happened on other occasions. “When we got the calculator, the way we teach math and how we test students changed completely,” he recalled.
“We can all have an amazing educator in our pocket that is personalized for us, that helps us learn,” the OpenAI executive suggested, adding that he and his team want users to think of ChatGPT as a tool.