AI about to get ‘uncomfortable’, says ChatGPT's Sam Altman
OpenAI CEO Sam Altman said AI products will have to allow a lot of individual customisation that will make a lot of people uncomfortable
OpenAI chief executive Sam Altman has said that the ChatGPT maker's next model will be able to do lot more than what the existing ones can, adding that artificial intelligence is evolving much more rapidly than previous technologies.
In an interview to Axios at Davos, Altman, who was ousted from OpenAI last November before being quickly reinstated, said the evolution and proliferation of the company's technology will require uncomfortable decisions.
According to Altman, AI products will have to allow a lot of individual customisation that will make a lot of people uncomfortable. He reasoned that AI will provide different answers for varied users on the basis of their values preferences and possibly the country they live in.
“If the country said, you know, all gay people should be killed on sight, then no ...that is well out of bounds. But there are probably other things that I don't personally agree with, but a different culture might. ... We have to be somewhat uncomfortable as a tool builder with some of the uses of our tools,” Altman said in this interview on the sidelines of the World Economic Forum.
According to Altman, soon the users might be able to ask AI about the most important emails they received in the day, and let AI summarise for them.
Altman also stated that the AI advances will help in accelerating the rate of scientific discovery. However, the OpenAI CEO does not expect it to happen this year itself. But according to him, it will be a big deal whenever it happens.
Altman, who said his top priority is to launch the new AI model, likely to be the GPT-5, said he is nervous about AI's impact on elections around the world this year. Around 50 countries including India will elect their next government this year.
OpenAI said it will stop people from using its technology to create chatbots that not only impersonate real candidates or governments but also misrepresent how voting works or discourage people from voting.
When asked about OpenAI partnering with the US Defence Department, Altman said the decisions is due to the desire of supporting Washington DC but he conceded that there will be a lot of things to start slowly on.