Since its launch in late 2022, OpenAI’s ChatGPT has generally rejected suggestions that it has emotions or desires by responding that it is just an artificial intelligence model. Updates announced by OpenAI on Monday showed the company apparently trying to make the chatbot act more like a human.
In demos, the new version of ChatGPT was capable of fast, natural voice conversations, picking up emotional cues, and displaying its own simulated emotional reactions.
During a livestream from the company’s San Francisco headquarters on Monday, OpenAI CTO Mira Murati announced that ChatGPT will be powered by a new, more powerful AI model called GPT-4o. The model will be available to free and paid ChatGPT users through a new desktop app, as well as the existing mobile app and web version.
Murati said the GPT-4o model allows ChatGPT to respond more quickly to voice, image and video input than previous OpenAI technology. In demos, she and other OpenAI employees held quick conversations with ChatGPT, which responded using a lively, expressive female voice and nimbly kept up the pace when interrupted.
ChatGPT adopted different emotional tones during the conversation and sometimes responded as if he were experiencing feelings of his own. When an OpenAI employee said that he had been talking about how “useful and amazing” the chatbot is, he flirtatiously responded and gushed, “Oh, stop it, you’re making me blush.”
“This feels so magical, and it’s wonderful,” Murati said, adding, “over the next few weeks we will roll out these capabilities to everyone.”
Elsewhere in the demo, ChatGPT responded to OpenAI researcher Barret Zoph’s greeting by asking “How can I brighten your day today?” When Zoph asked the chatbot to look at a selfie of him and say what emotions he was displaying, ChatGPT responded “I’ll put on my emotional detective hat” and warmly said “You seem to be feeling very happy and cheerful… whatever.” passing by, you seem to be in a very good mood.”
In a blog post On Monday, OpenAI CEO Sam Altman highlighted the importance of the new interface. “It seems like the AI in the movies; and I’m still a little surprised that it’s real,” Altman wrote. “Reaching response times and expressiveness at a human level turns out to be a big change.”
OpenAI via WIRED Staff