© 2024 88.9 KETR
Public Radio for Northeast Texas
Play Live Radio
Next Up:
0:00 0:00
Available On Air Stations

OpenAI releases latest ChatGPT — it can talk, laugh and even sing like a human


OpenAI released the latest version of ChatGPT, which can talk and laugh and even sing. Here's NPR tech correspondent Bobby Allyn. No, really. It's him.

BOBBY ALLYN, BYLINE: We all know AI voice assistants can talk - Alexa, Siri - and ChatGPT has been able to do the same. But OpenAI's latest version adds some human flourishes to conversation that have freaked some people out. In this livestream demo, a guy named Rocky tells ChatGPT he has an upcoming interview with OpenAI.


AI-GENERATED VOICE: OpenAI, huh? Sounds vaguely familiar. Kidding, of course (laughter). That's incredible, Rocky. What kind of interview?

ALLYN: The sarcasm, the laughing, the cadence, the exuberance at the end. Rocky turns the camera on himself and asks, OK, well, how do I look?


AI-GENERATED VOICE: Maybe just run a hand through your hair or lean into the mad genius vibe.

ALLYN: These back and forths lit the internet up. Here's Winsor Sineus reacting on TikTok.


WINSOR SINEUS: The way the AI was talking to this man, rizzing (ph) him up, bro - giggling.

ALLYN: It immediately drew comparisons to the 2013 sci-fi romance film "Her," where Joaquin Phoenix falls in love with a computer operating system voiced by Scarlett Johansson.


SCARLETT JOHANSSON: (As Samantha) You have several thousand emails regarding LA Weekly, but it looks like you haven't worked there in many years.

ALLYN: The AI goes through Phoenix's hard drive. He says, oh, yeah. I saved some emails because they were funny. The AI says...


JOHANSSON: (As Samantha, laughing) Yeah, there are some funny ones. I'd say there are about 86 that we should save. We can delete the rest.

ALLYN: Sure does sound a lot like ChatGPT. OpenAI CEO Sam Altman stoked the speculation himself by writing one word on X the day the new ChatGPT came out - her. In an interview with NPR, OpenAI's chief technology officer Mira Murati said the model is not based on the movie.


MIRA MURATI: It says more about our imagination, our storytelling as society than about the technology itself.

ALLYN: She says the voice similarity and even the flirtiness is based on user input. What a Silicon Valley way describe it. She says when the new ChatGPT voice assistant is released in the coming weeks, it won't have what tech people call latency, meaning it won't have the awkward pauses between speaking.

MURATI: The latency doesn't get in the way of the interaction, and it's quite a joy to interact with this model.

ALLYN: Hopefully more joyful than the ending of the movie. Bobby Allyn, NPR News.

(SOUNDBITE OF MUSIC) Transcript provided by NPR, Copyright NPR.

NPR transcripts are created on a rush deadline by an NPR contractor. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.

Bobby Allyn is a business reporter at NPR based in San Francisco. He covers technology and how Silicon Valley's largest companies are transforming how we live and reshaping society.