OpenAI’s ChatGPT, launched in late 2022, has often denied having emotions or desires by emphasizing that it is simply an artificial intelligence model. However, recent upgrades announced by OpenAI on Monday suggest the company is aiming to make the chatbot more human-like.
The new version of ChatGPT demonstrated the ability to engage in rapid, natural voice conversations, detect emotional cues, and even display simulated emotional reactions. Powered by the advanced AI model GPT-4o, ChatGPT will now be accessible through a new desktop app for both free and paid users.
OpenAI’s chief technology officer, Mira Murati, showcased the capabilities of the GPT-4o model, highlighting its improved responsiveness to voice, image, and video inputs. In demos, ChatGPT engaged in lively conversations, adopting various emotional tones and even responding flirtatiously at times.
During a livestream event from OpenAI’s headquarters, the chatbot impressed with its ability to maintain engaging interactions and provide thoughtful responses. OpenAI’s CEO, Sam Altman, emphasized the significance of this new interface, likening it to AI from the movies and expressing his surprise at its realness.
The company plans to roll out these enhanced capabilities to all users in the coming weeks, signaling a significant step towards achieving human-level response times and expressiveness in AI technology.
