The conversation examples are impressive but this made me say holy shit out loud:
"it runs locally on a MacBook Pro M4 Max or better. And we figured out how to train Dessa on only one NVIDIA GB200 NVL72."
If it really can replicate that low latency + that expressiveness on local devices, that's one heck of an accomplishment for "indie" AI.
> LLMs need an insane amount of text. Dessa learned language from hundreds of times fewer words (in fairness, LLMs also have to learn C++ and whatever the hell Angular is).
Now that a larger proportion of text is AI, I'm appreciating human-written text more and more; this got a good chuckle out of me.
There's no way I would be able to detect if this was AI from the examples alone. If I got a phone call from a malicious agent using this tech, I probably wouldn't realize it. My plan is to ask a challenge question about a shared memory with people I personally know, if there's something suspicious about a call. I don't know how else to prevent it. It sucks to be this suspicious, but I can't help it.
There's so many potential scenarios.
One scenario: An AI voice trainer trains on someone's social media video posts and get a pretty perfect copy. Then SIM jacks the target's phone number and calls pretending to need help. AIs could be used to mock ransom people, request bail money, ask for a change on a production server, etc.
People with high public visibility are likely targets, especially if they have money and or power. Impersonating them to subordinates would be an easy means of a social engineering attack.
It sounds like it is trained on a lot of podcasts, judging by a quick few second skims over their demo tracks.
Also, the trend of posting on HN with much more sensationalist title first, getting to the front page, then getting the title adjusted by a moderator should probably be penalized; how is 'most authentic ever' not a marketing term? (not just this post, but a bunch of posts recently got to frontpage by making unreasonable claims)
It also seems like it trained on dialog in plays and novels. Some of these responses seem way too "perfect", like a response you would hear in a dramatic soap opera. Honestly it might be using that too. Just my take from listening to first minute or two of dialog.
These bidi/duplex demos always come off as overly artificial to me. It's cool that the model can speak at the same time as you instead of taking turns, but training it to interrupt as much as possible isn't the way to do it. In the very first example, did anyone here find the laugh after the "winter" answer to be convincing? What was the need for it? Why did the human speaker not get to finish a single sentence or finish a thought before the model took over with a monologue?
OpenAI ruined advanced voice mode for me with that. I’m in the middle of a sentence and it goes “uh huh” and it kills my whole train of thought. It does not make me feel like I’m talking to a real person, it makes me feel like I’m talking to a stupid stupid AI that I hate.
I listened after reading your comment, and in the first half of the example, Dessa sounds drunk to me, or at least tipsy. So weirdly, the laugh made sense to me.
I'm absolutely going to hate talking to these things as customer service agents, with fake vocal fry and odd amounts of intonation and emotion peppered in.
For sure there's gonna be people falling in love with them though! I find all the anthropomorphization creepy though.
I've been hearing people talk negatively about vocal fry for ages now with no idea what it meant. But this comment was finally the one that made me go look it up. And today I learned that I speak almost entirely in vocal fry.
Most recent voice-related tech is using voice as an interface or interaction mode. It’s a means to an and, and the end is the value provided by the reasoning system on the other side of the voice.
This model is clearly not so smart. So the voice is the whole point, there’s not much value beyond it, at least on its own.
What that adds up to is, impersonation. That’s what this is for, impersonating humans.
These full-duplex models are getting very "Her"-ish lately. gpt-live-1 is fantastic. Dessa, though, seems to be aiming at full human parity. There's a huge market for AI girlfriends that they should be looking into in earnest...
human advancement is a function of population, instead of focusing on population decline to solve our resource problems we should focus on efficiency and technology. i am not rooting to revert to the dark ages i really struggle with this ideology of reduction that seems extremely popular. we really take for granted the fragility of human existence we exist in a pico second of reality and could and most likely will be snuffed out eventually, having our own species rooting for this outcome seems suicidal.
> we figured out how to train Dessa on only one NVIDIA GB200 NVL72.
This is the crazy part. It seems pretty smart, and LLMs require 1000x more compute to train. Can this kind of efficiency somehow be achieved with text too?
Not if you want the same model that debugs your C++ to also be able to sort objects by shape and generate a description of a video and pick out a good birthday gift for your cousin
"Oh, uh, yeah, pointers and stacks and stuff, uh – lemme spin up Minimax. Just a sec. Ohhh – ok, see where you're casting that float to a double there, uh, line 72..."
Agreed, it’s incredible and scary how natural and human it sounds. It’s gonna be pretty much impossible for almost anyone to tell if they’re talking to AI
Really impressive demo. I have to say that, but I do worry with this kind of fidelity to a human interaction that people will get more drawn into AI relationships and replace human contact.
I like voice as a mode of interaction with AI, but it’s always been clear with current tech that I’m talking to AI (and yet folks still got drawn into fantasy relationships thanks to sycophantic models etc) - I imagine with voice models like Dessa that could be worse for humans. This kind of fidelity blurs the lines of who or what the human is talking to and lets humans fall into an isolated world where they believe they are “socializing.”
Now, we could argue this is not a negative, maybe it is actually healthy for those who are looking for social connection or lonely, but my gut feel is that is not the case for most. It worries me that this kind of model will actually cause further deterioration of human socialisation and human-to-human contact.
So, I tolerate the "ums" and "ahs" and "uhs" in conversation from humans because they're human. If an AI is going to speak to me (gods forbid), it better be concise and clear and focused, or I'm throwing it out the fucking window.
It is interesting that they've done this reasonably successfully, impressive technology. Novel ideas can have inherent value, absent direct uses they may have, as they can be a stepping stone toward other inventions. But, I'm having a hard time coming up with a use case that isn't predatory or intentionally misleading. Company support agents pretending to be human, for instance. Or, taking the next step on AI girlfriend/boyfriend type interactions. Or, outright scams. Models already do an excellent job with voice cloning, at least for intonation and timbre, adding realistic conversational tics makes it even more likely grandma is going to fall for it when her "nephew" calls needing help.
But, that seems to be extremely common with new tech these days. There's always an angle that is unsavory. Marketing, scams, propaganda, surveillance, killing, that's where the money is, so everything bends toward them.
I also don't get the need for all these imperfections. In some examples it sounds like "she" is half asleep or even tipsy. If I want to talk to an all powered machine that knows my codebase inside out, or that can discuss about technical topics like the universe, science, or a specific technical book, etc... I expect it to be an eloquent speaker, capable of a good cadence, and that's easy to understand. I don't need it to mimic all the imperfections, I want it to be like when I'm talking to a really smart person that keeps me engaged and makes me feel smart.
Especially when you keep in mind that not everyone knows english as their first language. I know he's a bit controversial sometimes but Neil deGrasse Tyson is easy to understand and explained interesting topics in his podcast (I could still understand most of it when I was still getting used to listening to english 15+ years ago).
If conversation is the new interface then we will need UX designers to take care of this.
I came here to say the same thing. I don't want a computer constantly inserting stop words, pauses, laughing unnecessarily and interrupting me. This seems technically impressive but is not something I'd want to interact with.
I know someone who recently came across a scammer where a guy impersonated a sheriff's deputy and tried to get money out of her. She told me the guy sounded like a real sheriff and made her believe the guy until he asked her to pay some money in Bitcoin. I can imagine that this type of scam could go viral soon. Even worse, someone could record our voice and fine-tune a model like this so that it can produce our voice and then talk to our family members to deceive them. It'd be a real trouble.
Totally agree. Those scam calls calling as IRS agent, or outsourced criminal shops that try to steal identity, now this makes it much easier. No more foreign accent. They can hyper optimize to sound like a person from that specific locality.
I actually thought it sounds really good and natural and it freaked me out in a good way.
You can tell that the robots have no idea of what they're talking about, but the voicing, the emotions, etc, really good.. Like talking to a friend on the phone.
Also creepy, because of the potential for abuse of this technology.
But the interesting part to me is that training on voice only is what allows models to learn emotions - voice transmits a lot of emotional information which text doesn't capture. Language has words for a very limited number of emotions - maybe only the peaks - but there's so many more which we can't describe or transmit via text.
You could determine the phonetic tokens which express different kinds of emotions or 'render' text+emotion using a DSL, etc. Exciting stuff.
This feels like a parlor trick where the AI was trained to mimic informal telephone mannerisms. The actual conversations were pretty bad. Lots of style, not much substance basically.
Interesting tech, but if these are the hand-picked examples that best represent it, I suppose we're not quite there yet.
The intelligence sounds lacking. At this point, anyway.
But the idea of learning through voice instead of written text might prove fruitful. After all, text isn't strictly native to humans. It's just a tool we've invented.
Not bad, but I guess I lack the imagination to envision the use cases for such casual conversation. Other than... I suppose a certain demographic would be lining up to use this as the voice for their AI girlfriend.
Also I seem to be missing where/how they are selling this.
Both speakers in the examples sound unnatural. The human doesn't know what to talk about and seems to have little confidence. The robot says things that only mostly make sense and draws on, saying ultimately very little. "How's the winter stuff?" e.g. is not a sentence a person would say. I might say "How are you dealing with the snow?" or "Are you looking forward to winter?," but not "How's the winter stuff?" And there's just so many examples like this. Also, I'm not convinced that this "human-like" conversation with filler words and interruptions and pauses is actually what anyone wants, especially in absence of a face to interpret or a real human being to understand behind it. Even if it worked perfectly, it would feel uncanny and artificial because I know the person isn't real.
They sound just as unnatural als the annoying actors in the 'Gilmore Girls' series my daughter keeps on watching and re-watching, i.e. stupid valley girl or whatever it is called witless drivel. In that sense this model is tailor-made to create direct script-to-screen voice-overs for pabulum like that.
Can you instruction-train a duplex model like this? How would that look? I also wonder, if a model trained only on speech can develop world knowledge like this, could you train an image model only on images of text to essentially act as an LLM?
Call me a skeptic, obviously the voice sounds plausible and very realistic and human-level but where does the intelligence come from if it's only trained on human speech datasets? LLMs are powerful because they can do anything, if you only have a voice corpus I doubt it is as powerful as all written human knowledge. Like when it says "My day was good", what does it actually mean? Or is it just generating plausible slop.
I'm surprised it took this long! Clearly the next step is joint training on text and audio. The speech realism is very good, but the intelligence is pretty clearly 2024-level.
The speech is great, but you can see some cracks even in the first sound bit. near the end he asks the model to speak louder and initially the voice is louder but then goes back to the same level, also losing what she was saying just seconds ago and he has to lead back to the topic.
Sorry I'm lost. The sample on the homepage sounds absolutely awful. Not one bit of the cadence individually and together, the tonal ranges in responses or the overall structure and feel is comfortable to listen to, it's so far off passing as 'real' that I feel like I must be missing something.
>That doesn't sound like the same bar as "most authentic full-duplex speech model ever built"
It's up there. I think most humans would be convinced they're talking to a human. I assumed you've used other speech models. They are okay, but not as natural and imperfect as Dessa is.
If it really can replicate that low latency + that expressiveness on local devices, that's one heck of an accomplishment for "indie" AI.
Now that a larger proportion of text is AI, I'm appreciating human-written text more and more; this got a good chuckle out of me.
I can reliably get my custom models to output a 95% perfect replication of my tone and voice
There's so many potential scenarios. One scenario: An AI voice trainer trains on someone's social media video posts and get a pretty perfect copy. Then SIM jacks the target's phone number and calls pretending to need help. AIs could be used to mock ransom people, request bail money, ask for a change on a production server, etc.
People with high public visibility are likely targets, especially if they have money and or power. Impersonating them to subordinates would be an easy means of a social engineering attack.
Also, the trend of posting on HN with much more sensationalist title first, getting to the front page, then getting the title adjusted by a moderator should probably be penalized; how is 'most authentic ever' not a marketing term? (not just this post, but a bunch of posts recently got to frontpage by making unreasonable claims)
For sure there's gonna be people falling in love with them though! I find all the anthropomorphization creepy though.
This model is clearly not so smart. So the voice is the whole point, there’s not much value beyond it, at least on its own.
What that adds up to is, impersonation. That’s what this is for, impersonating humans.
The conversation with the AI fake wife was a little creepy, albeit honest about what this tech is probably going to be used for.
This is the crazy part. It seems pretty smart, and LLMs require 1000x more compute to train. Can this kind of efficiency somehow be achieved with text too?
It is kinda scary that we're here already, I can see this being pretty problematic for seniors/public prone to getting scammed.
I like voice as a mode of interaction with AI, but it’s always been clear with current tech that I’m talking to AI (and yet folks still got drawn into fantasy relationships thanks to sycophantic models etc) - I imagine with voice models like Dessa that could be worse for humans. This kind of fidelity blurs the lines of who or what the human is talking to and lets humans fall into an isolated world where they believe they are “socializing.”
Now, we could argue this is not a negative, maybe it is actually healthy for those who are looking for social connection or lonely, but my gut feel is that is not the case for most. It worries me that this kind of model will actually cause further deterioration of human socialisation and human-to-human contact.
It is interesting that they've done this reasonably successfully, impressive technology. Novel ideas can have inherent value, absent direct uses they may have, as they can be a stepping stone toward other inventions. But, I'm having a hard time coming up with a use case that isn't predatory or intentionally misleading. Company support agents pretending to be human, for instance. Or, taking the next step on AI girlfriend/boyfriend type interactions. Or, outright scams. Models already do an excellent job with voice cloning, at least for intonation and timbre, adding realistic conversational tics makes it even more likely grandma is going to fall for it when her "nephew" calls needing help.
But, that seems to be extremely common with new tech these days. There's always an angle that is unsavory. Marketing, scams, propaganda, surveillance, killing, that's where the money is, so everything bends toward them.
Especially when you keep in mind that not everyone knows english as their first language. I know he's a bit controversial sometimes but Neil deGrasse Tyson is easy to understand and explained interesting topics in his podcast (I could still understand most of it when I was still getting used to listening to english 15+ years ago).
If conversation is the new interface then we will need UX designers to take care of this.
But the interesting part to me is that training on voice only is what allows models to learn emotions - voice transmits a lot of emotional information which text doesn't capture. Language has words for a very limited number of emotions - maybe only the peaks - but there's so many more which we can't describe or transmit via text.
You could determine the phonetic tokens which express different kinds of emotions or 'render' text+emotion using a DSL, etc. Exciting stuff.
The intelligence sounds lacking. At this point, anyway.
But the idea of learning through voice instead of written text might prove fruitful. After all, text isn't strictly native to humans. It's just a tool we've invented.
Also I seem to be missing where/how they are selling this.
Where is this going?
It's up there. I think most humans would be convinced they're talking to a human. I assumed you've used other speech models. They are okay, but not as natural and imperfect as Dessa is.
There's no male or neutral voice.
All examples are of white male software engineers talking to what I assume will be ultimately marketed as a 'girlfriend' bot (or used for love scams).
The single paragraph hand waving at the importance of 'safe AI empathy' doesn't dissuade me at all.
Sell your idea as what it is, yet another 'perfect' woman shaped object a hypothetical man can own and use how he likes.
Is having a duplex model constantly cutting in with uninteresting quips something people are looking for?