Welcome to the deep dive before we jump in today a quick. Thank you to our
Download transcript (.srt)
0:00
0:03
0:05
0:09
0:10
0:13
0:18
0:22
0:27
0:32
0:34
0:40
0:41
0:45
0:48
0:51
0:56
0:57
1:00
1:01
1:05
1:10
1:11
1:15
1:18
1:23
1:29
1:30
1:35
1:36
1:40
1:45
1:49
1:50
1:55
1:56
2:00
2:01
2:06
2:09
2:10
2:14
2:17
2:21
2:22
2:26
2:29
2:34
2:35
2:38
2:42
2:47
2:48
2:54
2:55
2:59
3:00
3:04
3:08
3:12
3:15
3:19
3:22
3:26
3:27
3:32
3:34
3:39
3:40
3:44
3:49
3:50
3:54
3:57
4:01
4:04
4:05
4:10
4:12
4:15
4:17
4:22
4:25
4:30
4:32
4:36
4:38
4:42
4:45
4:48
4:53
4:54
4:56
5:02
5:03
5:07
5:08
5:13
5:18
5:23
5:25
5:29
5:30
5:35
5:37
5:42
5:47
5:49
5:53
5:55
5:59
6:00
6:04
6:07
6:11
6:14
6:18
6:21
6:26
6:28
6:31
6:35
6:38
6:43
6:45
6:50
6:54
6:55
6:59
7:00
7:02
7:09
7:09
7:13
7:14
7:18
7:19
7:22
7:25
7:30
7:32
7:36
7:37
7:41
7:45
7:49
7:52
7:57
8:01
8:02
8:06
8:07
8:10
8:15
8:18
8:25
8:25
8:30
8:31
8:34
8:35
8:39
8:42
8:46
8:48
8:51
8:52
8:58
9:00
9:05
9:07
9:11
9:17
9:19
9:24
9:26
9:31
9:36
9:38
9:42
9:44
9:50
9:52
9:56
10:00
10:01
10:06
10:09
10:14
10:16
10:21
10:24
10:29
10:32
10:37
10:39
10:44
10:44
10:48
10:50
10:55
10:55
10:59
10:59
11:01
11:04
11:07
11:08
11:12
11:12
supporter safe server
Safe server handles software hosting and they're really focused on supporting your
digital transformation
So if you're looking for reliable hosting you can check them out at
www.safe server dot de like I said today
We're embarking on a mission that feels a bit
Well bit sci-fi maybe don't it leans that way we're looking at a lotto AI
Right, and the core idea is taking sophisticated conversational AI like really
advanced stuff and getting it out of the screen
Out of your phone or computer and into physical things specifically toys plushies
even it sounds simple
But the sources suggest this is way beyond just a talking toy exactly
We're gonna break down how they're trying to merge the hardware the software and
these really distinct AI personalities
The goal seems to be making these interactions feel hyper realistic. That's the
hook isn't it?
The source material literally says they're giving plushies voices that feel
ridiculously real
Yeah, it made me think of that movie Ted the talking teddy bear, right?
But imagine that powered by actual live cutting-edge AI. That's kind of wild and
that's what we're diving into
It's sort of the next step for digital companions moving them into the physical
world. Okay, let's start with the basics the hardware
For anyone maybe new to this kind of tech. What is the a lotto device?
Physically well at its heart. It's a small gadget an IOT client. Technically. He's
got the microphone
It's got the speaker but the clever part is how it attaches. Okay, it uses two
simple silicone straps
So you can clip it on to pretty much any toy you already have. Oh, right
So you don't need to buy their specific toy. Nope that old teddy bear in the attic
Yeah, suddenly you can have you know a brain and a voice that flexibility seems
like a big deal
It really is and the setup sounds incredibly simple aimed at well anyone no tech
skills needed
How simple are we talking like three steps simple first clip the device onto the
toy?
Okay, second connect it to your home Wi-Fi. It uses what's called a captive portal
Uh-huh like when you connect at a hotel exactly that it makes its own little
network
Temporarily to guide you super easy and third pick a character personality from
their list and just start talking to it
Wow, okay. Now I saw they're actually two different products mentioned. Yeah, they're
catering to slightly different people
There's the main AI device. That's the consumer one, right pre-order price
mentioned was $69 that gets you the device
Access to all the AI characters unlimited apparently and a free month of their
premium subscription
It's the plug-and-play version and the other one for tankers
That's the AI dev kit a bit cheaper $59 on pre-order
This one's really for developers makers people who want to mess around with it. How
so it has open source firmware
Runs over a standard USB C connection and lets you load your own custom voices or
even your own AI models
If you want much more flexible if you're technically inclined gotcha and practical
things
Battery life is this thing always plugged in apparently not they claim a week of
battery life, which is pretty good makes it actually portable
Yeah, that's essential if it's meant to be a companion, and I saw something about
community support uh-huh over 1200 stars
They said which suggests. There's already a decent buzz around it people are
interested that early engagement is usually a good sign
Definitely shows people are intrigued by the idea yeah, and maybe even want to
build on it themselves, okay?
Let's shift gears the hardware is neat, but the sources really emphasize the
personalities the who
This seems to be where a lot of really tries to stand out absolutely this isn't
just about making a toy talk
It's about giving it a very specific often complex character
They mentioned over a hundred ai characters available a hundred and they're not
just slight variations
Yeah, not at all the examples they give are incredibly diverse they seem to be
leaning into strong personalities even flawed ones
Not just helpful assistant, okay. Give us some examples. What kind of range are we
talking well? You've got the comforting
nostalgic types
Like Dottie Mae Dottie Mae described as a classic Southern diner waitress uses
terms like hun sweetie
gives unsolicited advice
recommends the pie
Pure comfort food in voice form basically ah okay, so that's one end. What about
the other end? Oh they go there dramatic flamboyant characters
There's captain star flash is a super overconfident space captain who thinks laser
solve everything right or dr
Voltanus the classic mad scientist full of manic energy apparently shouts catchphrases
think loud thunder effect
So you could clip this onto like a superhero toy or something exactly or maybe
something completely incongruous for comedic effect
And what about more thoughtful characters yep?
They mentioned paradox pithius an ancient Greek philosopher type sounds wise wise,
but also apparently kind of smug
He answers your deep questions
With even deeper possibly more annoying questions makes you think but maybe grinds
your gears a little okay
This is where that uncensored aspect might come into right the comedy take sugarplum
the description is fascinating
Speaks in a super sweet bubbly childlike voice sounds innocent
But apparently drops comments so dark it makes Satan clutches pearls whoa
Okay, that's a choice. It's intentional friction right that contrast creates shock
value makes it memorable
It's not trying to be bland and they seem to lean into existing pop culture stuff,
too. I saw Ted mentioned
Yeah, Ted the inappropriate Teddy. Yeah, clearly referencing the movie character
Boston accent bar fly mouth
Can you imagine where that goes uncensored indeed any other specific types loads?
They mentioned Mikey Sally Sullivan hardcore Boston guys swearing rants, then there's
the proper British lad
What's his deal judges your tea making skills apologizes constantly if you bump
into him very specific cultural niche
It seems like they're aiming for very defined archetypes. Totally and it's not just
comedy or stereotypes
They even list Zoran Mamdani the political activists. Yeah described as empathetic
focused on social equity and justice
So the range covers serious and specific viewpoints to not just jokes
So the strategy isn't just make a friend
It's pick a very specific memorable character exactly depth and distinctiveness
over just being generally agreeable
You clip it on you get that personality fully formed which brings us neatly to the
how we know the what the device we know
These wild personalities
How does the tech actually pull this off in real time making a toy have a
continuous natural conversation globally?
That sounds hard. It is hard. The key seems to be what the source calls real-time
speech to speech conversion
We're talking potentially up to 15 minutes of uninterrupted chat 15 minutes. Wow
How they use what the source referred to as a brain trust? They're not relying on
just one AI model
Oh, okay. So they're pulling from multiple sources, which ones? Yeah, it's quite a
list of the big names right now
Open AI is real-time API
Google's Gemini live API 11 labs AI agents and also Hume AI EVI for four different
ones
Why so many wouldn't that be complicated? It probably is but the idea is that each
model has strengths
Maybe one is faster one sounds more natural one is better at catching emotional
cues
By using several they can kind of pick the best tool for the job for each part of
the conversation or blend them
It helps keep the latency low and the quality high like hedging your bets that
makes sense redundancy and optimization
Okay for someone listening who isn't a developer. Can we simplify the architecture?
You mentioned a triangle earlier
Yeah, think of it as three core pieces working together really fast first
You've got the device itself the IOT client that ESP 32 thing
We talked about clip to the toy it just captures your voice and plays the AI's
voice sends the audio securely using web sockets
Okay, piece one the ears and mouth on the toy exactly piece two is the edge server
This runs on something called Dino think of it as the super fast traffic controller
or router my edge
It means it's located geographically close to you and also close to the big AI
models
Its whole job is to grab the audio from the toy
Instantly fire it off to the right AI service like Gemini or 11 labs get the
response back and zap it straight to the toys speaker
Minimizes delay got it the middleman ensuring speed and the third piece. That's the
front end
Basically the website or app you use built with next.js. This is where you choose
your characters
Maybe create custom ones adjust the volume that kind of thing. Ah, and I saw you
can tweak the pitch
Yeah, the pitch factor so you could take a serious character's voice and make it
sound high pitched and cartoonish if you wanted more
Customization. Okay. So the whole thing relies on speed if there's a big delay it
ruins the illusion of conversation
What kind of performance are they claiming? The numbers are pretty impressive,
especially for a global system
They're aiming for under two seconds round-trip latency under two seconds from you
speaking to hearing the reply
Yeah, which is generally fast enough to feel pretty conversational not like a walkie-talkie
and the audio quality. Does it sound clear?
They mentioned using the Opus Kodak at 12 kiloby piece
Which in non-technical terms means it should sound pretty clear and crisp even
though they're keeping the data rate low for speed
Okay, one more tech thing. How does it know when I've finished talking? Do I have
to press a button? No, and that's crucial
They use something called Server VAD voice activity detection. Server VAD?
Right, instead of the little device trying to guess, the powerful server analyzes
the audio stream in real time
It figures out precisely when you've naturally paused or finished speaking. Ah, so
it makes turn-taking much smoother
Exactly, less awkward silence, fewer interruptions, key for making it feel real.
Plus they mentioned OTA updates. Over the air?
Yeah, means the software on the device can be updated automatically over Wi-Fi
So it can get better over time without you needing to plug it into a computer. Okay,
so putting it all together
It's quite an ambitious project. Merging these very specific, sometimes wild
personalities with hardware that enables smooth, fast
conversation. It really is. The big takeaway seems to be shifting AI interaction
away from just typing in a box. And
into a physical object you can actually talk with. Like, really talk with. Whether
you want that companion to be a nurturing waitress like Dottie May or
a sarcastic philosopher or an inappropriate teddy bear. Right, it's that
customization delivered through a physical form.
So the final thought for you listening, the source emphasizes this device has no
filters, no rules.
We have the tech now to give an innocent looking plushie a voice that could be,
well,
deliberately offensive like Ted, or shockingly dark like Sugar Plum, or maybe even
politically charged.
If digital companionship becomes totally personalized and unrestrained, what does
that mean?
What happens when we start designing companions not to be helpful or polite, but
maybe
unhinged?
Provocative. Something to think about as this tech develops.
Well, that's all we have time for on this deep dive and thanks again to our
supporters Safe Server.
Remember, they handle software hosting and support digital transformation.
sources.
sources.