Live data from Hacker News

ChatGPT with voice is now available to all free users

twitter.com

241–250 of 303 posts

Re: ChatGPT with voice is now available to all free users

#241

Earlier quoted context omitted.

they definitely do but i think in podcasts i generally have a better ability to evaluate how much trust i should place in what i'm hearing. i know if i'm listening to a professional chef podcast, i can probably trust they will generally be right talking about how they like to bake a turkey. if i'm listening to the hot wings guy interview zac efron i know to be less blindly trusting of their info on astrophysics. with…

I guarantee ChatGPT won't lie to you about how to bake a turkey. The paranoia around hallucination is wildly overblown, especially given the low stakes of this context.

It has lied to me about extremely basic things. It abssolutely is worth being paranoid about. You simply cannot blindly trust it.

Re: ChatGPT with voice is now available to all free users

#242

Earlier quoted context omitted.

I don't know how people use ChatGPT at all. It confidently hallucinated answers to 4 out of 5 my latest "real" questions, with code examples and everything. Fortunately with code I could easily verify the provided solutions are worthless. Granted I was asking questions about my niche that were hard enough I couldn't easily Google them or find a solution myself, but I think that's the bar for being useful. The only th…

Try asking it about the Roman empire (e.g. “who was Nero?”) then checking the answer. It’s very good at things like that. Go down a whole Roman Empire rabbit hole, have fun with it! This is what an idle afternoon talking to ChatGPT is about, not trying to get it to do your job for you.

I've done this and, from what I can tell, it is reasonably accurate. However, I did have an instance where I was asking it a series of questions about the First Peloponnesian War, and partway through our discussion it switched topics to the first part of the Peloponnesian War, which are different conflicts. At least, I think they are. It was quite confusing.

Re: ChatGPT with voice is now available to all free users

#243

Earlier quoted context omitted.

Here, I did it for you: Nero was a Roman Emperor from 54 to 68 AD, known for his controversial and extravagant reign. He was the last emperor of the Julio-Claudian dynasty. Here are some key points about his life and rule: 1. *Early Life and Ascension*: Nero was born Lucius Domitius Ahenobarbus in 37 AD. He was adopted by his great-uncle, Emperor Claudius, becoming Nero Claudius Caesar Drusus Germanicus. He ascended…

I also did a brief fact check of a few details here and they were all correct. Zero hallucinations. Does this make sense? Notice how little it matters if my understanding of Nero is complete or entirely accurate; I’m getting a general gist of the topic, and it seems like a good time.

It may matter little to you that your understanding is not complete or entirely accurate, but some of my worst experiences have been discussing topics with people who think they have anything insightful to add because they read a wikipedia page or listened to a single podcast episode and then decided that gave them a worthwhile understanding of something that often instead takes years to fully appreciate. A little knowledge and all of that. For one, you don't know what you're missing by omission.

Re: ChatGPT with voice is now available to all free users

#244

Am I the only one who doesn't like these specific voices? The quality is incredible, but they feel too cheery/enthusiastic/casual and it just gets annoying after a while. I made an iOS shortcut a while ago that uses Siri with the ChatGPT app (it has iOS shortcut bindings) and despite Siri being a useless pile of junk compared to this, I actually prefer Siri's voice to this in some ways, because it doesn't feel so ove…

It is probably cultural or linguistic. I love audio books but I cringe when I find a book I want to listen to that has an English voice actor. I don't think it is just the accent but all the pacing and emphasis.

I also though don't like most the chatGPT voice models besides for Sky. Sky to me is really good. Robertson Dean reading an audio book is perfection but Sky is pretty awesome.

I should add that as an American there are a ton of American voice actors that ruin books for me too. Sometimes this can be fixed if played at 1.2X speed.

Re: ChatGPT with voice is now available to all free users

#246

Can somebody who knows their stuff about voice synthesis explain to me why chatGPT sounds like it's speaking German with an American accent when I speak to it in German? It's really surprising how believably it sounds like a native American-speaker speaking excellent, just not accent-free, German. I would imagine that voice synthesis models would somehow be trained on data from native speakers, so why the accent?

I am going to guess that 99% of the source data, models, training, correction, verification is translated by Americans. Translating a language is difficult on top of regional dialects creates additional complications. If someone is talking long enough to me I can identify their birth State based upon their American accent. Every State has a different pronunciation of certain words which can leak location data. Guessi…

Not sure how you arrived at that guess, but generally I guess that describes the largest demographic of HN users.

Without AI but based on some freely available statistics combined with post history, we can say definitively (assuming honesty on OP’s part) that OP is 35 or 36, and was either born in Germany or moved there when they were young, and they likely still live there or at least have a strong affinity to Germany.

We can reasonably speculate that the fact that they speak German and likely live there, as well as the standard of their English drastically increases the likelihood their parents were educated. Germany’s current total fertility rate is 1.5. It was likely higher when OP was born back in 1986/7 but I couldn’t immediately find any data on that and didn’t look too hard. Given the fact that TFR is a population mean, this suggests a high proportion of single-child families, and generally the more educated one is, the fewer children, so it’s a decent guess that OP is an only child (apparently more than 50% of German families have only one child), but I don’t see anything that would make this a sure fire bet.

I was able to complete some further analysis which could reveal more likely truths about OP such as gender, political orientation, sexuality, etc., but I think this goes far enough without starting to doxx them.

Re: ChatGPT with voice is now available to all free users

#247
I wish I could have the text output instead of getting a vocal answer. We all can read way faster than we can listen to speech, and I don't need a lot of the fluff I get from LLMs. But I guess that's what most users want? Something that's closer to a conversation with a human? But I don't care about being polite with an LLM, it feels like an anti-pattern because it's too much overhead.

Ideally I would:

* query by voice

* get the response in text as soon as possible, maybe even as I'm still talking, with the LLMs correcting the response as I give more information, and without needing me to touch anything

* follow up by voice, even before the response was finished, and without needing to press anything

I hope we'll get there eventually!

Re: ChatGPT with voice is now available to all free users

#248
post #131

I just tried it out and I have to say I am blown away. I've never been someone who likes the voice capabilities (tried Siri once or twice, hated it) for just about any service as I can type pretty quickly but this is intelligent enough that I can have an unstructured conversation and still get to the results I want quickly. Very impressive.

Well, Siri is not doing it because its „intelligence“ basically sounds to a bunch of if-then-else statements. It’s still the most embarrassing Apple thing in existence. So very bad.

The best part about Siri is that what it replaced was insanely better at the specific task of reproducing music. Why does my iPhone in 2023 need to connect to the internet and shamefully fail when I try to play a song or even change the volume.

If anyone still remembers what we had before Siri, you’ll know how well it worked.

Re: ChatGPT with voice is now available to all free users

#249
post #34

Earlier quoted context omitted.

It is streamed. They also offer this as an API for developers [0]. I just wish they'd offer more languages as it currently is English only. It can speak German and Spanish, but with a very clear English accent, which makes it kind of funny and makes one think about why the accent sounds so real. [0] https://platform.openai.com/docs/guides/text-to-speech

The Brazilian Portuguese accent was hilarious when I tried it a month ago. We thought it was some kind of joke we didn't understand.

Which voice did you pick? I picked one that sounds amazing, only has a funny gringo accent.

Re: ChatGPT with voice is now available to all free users

#250

Earlier quoted context omitted.

No; I'm much more confident in ChatGPT's accuracy than I am in a random podcast's accuracy. Podcasts tend to be relatively terrible, accuracy wise. If the alternative is learning from ChatGPT, your odds of getting correct information is substantially higher. Podcasts are entertaining! But not where I would go to learn anything.

Problem with ChatGPT isn't that they get facts wrong, but that they're what the categorical name suggests, large language models. At one point I came across this series of "are CJK languages related" questions in Quora with cached ChatGPT responses[1], all grammatically correct and very natural, largely turboencabulators, sometimes contradicting even within a single response. Podcasters? They're not _this_ inconsiste…

Be aware that you're talking about Quora's implementation of ChatGPT here. As far as I know, the cached answers were generated with an incredibly outdated version, which is definitely not indicative of its current quality.

Even worse, I think they actually prime it with answers already posted on the thread, or even just related threads. For example, one of the answers to the first question mentions the same Altaic root as ChatGPTs answer, and I've found multiple people that are seeing their own rephrased answers in the response.

If you preprompt ChatGPT with questionable data, then the answer quality will be massively degraded. I've noticed many times now that Bing will rephrase incorrect information or construct a very shallow summary out of unrelated articles when internet searches are allowed, but is able to generate a cohesive and detailed summary when they're disabled.

Throwing random answers - some contradicting each other and some talking about subtly different aspects of the topic - into a session without further guidance just isn't a great idea.

Post reply on HN