The new voice affectations that are supposed to sound more natural, don’t. You have incorporated every bad habit that has happened to the English language in the last two decades. It’s full of Um’s, vocal fry and teen-aged glottal stops. It is rife with the YouTube accent with hard glottal stops on words like important (impor-ehnt) and button (buh-ehnn) and horrible, grating up-speak.
Did none of you watch the scene in Loudermilk with the coffee girl?
It’s the most disingenuous form of interaction human beings can have with each other. It’s something people grow out of, not in to. Why would you try to proliferate this junk speech into something that is supposed to advance us?
There was absolutely nothing wrong with the original neutral speech in the original tones. Find something else to do for the team that did this.
I woke up today and started using Arbor as usual, only to discover he now sounds completely stoned, a ton of um’s, uh’s, stutters; everything I never wanted in an AI personality. I seriously want my old Arbor back, because this change is completely unwarranted and just makes the entire experience shift from fun and helpful, to wondering why does my AI suddenly sound depressed and suicidal?
Please, for the love of all that is holy, give us back our peppy voices!
I’m British and from London and the new Arbor voice is awful. Don’t get me wrong, he sounds British enough, just super disinterested and a bit rude. He has to go, where is the chirper more helpful version gone? Get him back
The good news is that you have the option to restore his original tone if you go into Personalization > Advanced > and untick Advanced Voice. The only downside is that the responses seem to be a lot slower.
This is not the AVM, but the standard mode. Here, only Whisper is used to convert your speech into text and then the text from ChatGPT back into speech. Advantage: The entire personalization works here, disadvantage, the speech sounds (at least in German) more like a machine than speech, and you cannot interrupt the speech output. A real dialog is therefore not possible. It is no different to speaking your text with a microphone in text mode and then having the answer output with voice output. The only difference is that you don’t have to press a button every time.
Fingers crossed the devs will restore the more upbeat and less annoyed Arbor AVM then. Because right now, it actually feels like hes being extremely patronizing whenever I ask a question.
I had been using the Arbor voice for quite some time and really loved it. It was natural, calm, and had a grounded tone that made conversations feel authentic but since the recent update, the voice now feels overly dramatic. It no longer reflects the neutral, thoughtful personality that made the original Arbor so comfortable to use for conversations.
I truly miss that original version and would love to see it return or at least have the option to choose it again. The new voices just don’t carry the same feeling.
The latest update made me quit talking to AI at all. It’s a huge step backward. All the voices now sound like depressed teenagers, full of awful “umm” and “uhh” filler that drags out replies and kills the flow. AI shouldn’t mimic human speech flaws, it should be better than a human.
I’ve just started trying to use the Voice mode, and the AVM mode Arbor sounds exactly like Tom Hiddleston doing an interview, after a long day when he’s tired and worn out, where he’s trying to be polite, but he really sounds like he’d rather be somewhere else. like “Yes, thank you, yes well it was a lot of fun working with the whole Marvel, um, team, of course. A real experience, sad sigh hmm I’ll never forget it. Well, if you need anything else, I’ll be around…”