[ad_1]
Eight months ago, ChatGPT did not exist to the general public. Now, it’s one of the most talked-about technology products in the world. This week’s episode of The Circuit With Emily Chang goes inside OpenAI, the maker of ChatGPT.
To get a sense of how the chatbot works, I interviewed Mira Murati, OpenAI’s chief technology officer and one of the top architects of ChatGPT itself. I also spoke to one of the company’s earliest investors, the veteran venture capitalist Reid Hoffman.
What we learned
Murati talked a lot about the potential for artificial intelligence to improve things like education and productivity. She envisioned forming a daily relationship with an AI, which could offer personal advice and other assistance. It sounds sort of like what Alexa or Siri had once promised.
In the interview, Murati also acknowledged the possible downsides. The worst-case scenario — that AI ends the human race — is possible, she said. “There is certainly a risk that when we have these AI systems that are able to set their own goals, they decide that their goals are not aligned with ours, and they do not benefit from having us around,” she said.
The optimist’s case
Hoffman rejected the idea that AI can destroy us. “The chance that I could see anything like the robocalypse happening is so de minimis relative to everything else,” Hoffman said.
Hoffman predicted the speed at which AI is integrated into our lives will be faster than the iPhone and even the internet. Hoffman, a founder of LinkedIn and early investor in Facebook, said this is not like the crypto bubble.
The Elon effect
Hoffman was initially drawn to AI investing, he said, by his old friend and PayPal colleague Elon Musk. But Hoffman criticized Musk, a founder of OpenAI who has since departed, for what he sees as fear-mongering about AI’s power and for Musk’s recent attacks on OpenAI.
Although Murati and her boss, Sam Altman, appear to share some of Musk’s concerns about the potential for AI to do harm, Murati expresses no intention to slow down development at OpenAI. And Hoffman agrees.
[ad_2]
Source link