Human Element Communications

View Original

There’s something different about ChatGPT with voice

I had a fat chat with ChatGPT last week, a few days after its birthday. Nothing serious -- just some back and forth about which film is Leonardo DiCaprio's best (we tend to disagree on this because I struggled with 'The Revenant').

The nature of the conversation was different, though. Instead of spitting out reams of text at lightning-fast speed, the bot now speaks - with intonation and pacing that is surprisingly human-like. (When you enable the voice function, you can pick from five different voices, each created by one of the voice artists OpenAI worked with to develop this feature).

I know it's a cliche. But I immediately thought about how prescient the film 'Her' was as I tried to convince ChatGPT about the quality of DiCaprio's performance in 'The Beach.'

There is something different about having a verbal conversation with an LLM. A psychological threshold is crossed when moving from text-based chat to verbal interactions. Unlike Siri, which is a virtual assistant programmed to perform tasks based on commands, you can have a relatively natural discussion with ChatGPT.

This makes me wonder about where we are heading with all this. Voice-enabled LLMs can be used for teaching, language translation, call center operations, and much more. I can also see their potential for becoming dependable companions for people who are lonely and need a chat.

But I wonder how these advances are changing our relationship with technology and each other. And what this means. (I'm still working this one out).

How about you? Are you using ChatGPT with voice? What has been your experience so far? Any aha moments or existential musings you'd like to share? Let me know in the comments.


If you liked this post, you might also like:

Brendon Bosworth is a communications specialist and science communication trainer with an ever-growing interest in AI.. He is the principal consultant at Human Element Communications.