[kvasserman]

I think of it this way. LLMs suppose to be good at generating text/writing, right? Well, they are not very good at it. They generate plausible content that superficially makes sense. Most people can easily tell AI generated slop from human writing. I suspect that mimicking human voice is multiple levels more difficult for LLMs than mimicking human content. The level of nuance that humans produce in their speech is probably staggering. So I maybe completely wrong, but I see no evidence so far to support the idea that either LLM's writing or speaking is going to get much better any time soon.

[boa00]

Not sure I agree here

Text is just human thoughts in their most simple form. Writing is about expressing ideas, and there is almost an infinite number of ways to express them. Extremely difficult task, and LLMs only "imitate" it to the best of their training

This is not at all true for voice. There are an infinite number of possible voices, but a finite number of tones and phonemes you can use to express the text.

It's a much easier technical problem; it's just that it's much harder to gather proper data (you cannot just scrape Reddit and hope for the best, as LLMs do). And voice gets like 1/100th of LLMs' funding

[kvasserman]

Ironically, one of the thing that makes written word by AI recognizable as AI is that it's too perfect. Too polished. Now think about speech patterns, they are way more than voice frequencies, tones and phonemes. One can say the same phrase gazillion different ways, with different pauses, cadence, inflections, intonations and even pitch. Humans speak "imperfectly." It's very contextual too: in many situations, we voice the same words very differently. Again, it's possible that I don't know what I'm talking about, but every example of machine talking that I've heard, I felt it was too mechanical, precisely because it was lacking the nuance of how real humans speak.

[ben_w]

Perhaps, but for what it's worth, when I first heard OpenAI's TTS demo, I assumed they were faking it and a human was speaking because it had "um"s and "err"s.

Right now, the main thing making these things recognisable is there's so few voices. The voices themselves are basically celebrities, albeit in the same way as some annoying D-list celebrity who somehow managed to get a bajillion contracts for advertising cheap tat.

Given that LLM slop is currently rapidly degrading the trustworthiness of search results (even moreso than SEO already had), it's probably for the best if the major AI providers don't release a bunch more voices.