Speaker 0:  openai wants to replace the screen with a device that watches, listens, remembers, and talks like another person in the room. the breakthrough is natural conversation, but the battle will be whether consumers trust an always aware ai companion in their homes. so openai is reportedly developing a portable screenless home device containing cameras, environmental sensors, and movable mechanical opponents. instead of waiting for isolated commands like conversational smart speaker, it would use gpt live to carry on continuous conversations and delegate difficult work to more powerful cloud models. what do you think of this, john?

Speaker 1:  this sounds like the amazon alexa, all over again, but a little bit smarter this time. i can't say i'm surprised. it does make sense. again, the always on is the thing that really concerns me. now what would be cool is if you could run a local llm and have this listening all the time and learning and, you know, all that sort of stuff, taking in all this information so that it doesn't get out to the interweb. but, i think this is a good idea because having a smarter alexa, is helpful. like, i was using alexa for a while when it first came out, but there was a lot of limitations. but now that i'm a little bit more aware of data and, and data usage and and and all that sort of stuff and privacy, i'm less apt to get, a speaker like this. but this is still great. like, this makes sense to me. what about you? what do you think?

Speaker 0:  well, i think we are leaps and bounds away from alexa and siri. i think the world is not prepared for what is happening. i have several agents. i don't know. right now, i'm gonna say six on the go that i can talk to, and get them to actively do stuff. i usually type to them, but there's a lot of times where just talking to them is easier. and the more you look around, products like whisperflow and a few others, i've even got a, open source one installed on one of my machines. i hit a special key, the thing, and i dictate to it, and it does what i want it to. these things are like what we wanted the rabbit r one to be because the back ends are now connected to what they call large action models where you're connected to your cowork claude or your opencloud agent or your hermes or your, chatttp work, engine. these things can or open whatever and actively do stuff. so now you hook up a speaker to them. this thing knows who you are, knows your business, knows where all your files are, knows what's going on, and can actively do stuff on your behalf rather than, hey. play a song. you know, it can actively do work, develop, do, create apps, email people, do work, create documents, proposals, whatever you want. so the world changes. and i think what we got before from a voice interface was interesting. it just never, really became what people wanted.

Speaker 1:  no. and, again, i i like this. i like can, that, yeah, it can act on your behalf. it can access different things. but, again, worried about hacking. but, hey. if i can you know what? i thought we could easily create a device and build it ourselves and just link it to a local llm if this track dbt speaker doesn't allow that. i don't think it'd be too difficult. right? you just get a raspberry pi. you get a couple sensors. it wouldn't look too pretty, but, it would be local and be more safe. right?

Speaker 0:  yeah. absolutely. you could do all of that. i just don't know how much consumers are really concerned about privacy on most of this stuff.

Speaker 1:  yeah. compared to you and i don't, i agree. they're not too concerned. but what's interesting is that there's no screen on this. right? but we've talked about this. we talked about a device. we've talked about devices eliminating screens, but, openai is believing that ai should disappear into the background kind of thing. so instead of interacting with it through a screen, you just talk naturally, which i still i still don't interact with any of my llms using voice. i know a lot of people do, but i don't. i just type to it. i don't know why i don't. i've never i've never been one to dictate. maybe because i have adhd, and if i were to actually dictate, yeah, it wouldn't be the same coming out as if i was typing, if that makes sense.

Speaker 0:  it does. but the more i use them, the more i'm like you where i'm a little dyslexic and i need the time for me to sit down and type it, stops my flying brain from sending 10,000 things at a time and makes it focus on this one thing i have to communicate. because otherwise, when you're talking, it's like, okay. i got this to do, this to do, this to do, and they all come rushing out. right? but, you eventually get to the point where it's faster. it's just faster to say, okay. do this. you know, commit that to github, then do the update, then phase one, then send the email and get back to me when you're done. okay. and it goes and does it. you know? you don't anyway, back to this. our revealed, reported target is for 2026 with sales expected in 2027. openai previously stated in a separate trademark case that its jony ive device would not ship before february 2027, but no official launch date. so we're not a 100% sure if this is a johnny eyes device and whether this will be slowed down by the apple openai lawsuit, that we talked about last week. the reported cost is 2 to $300 us, and it requires a chatgdp subscription, optional premium tier for cloud service fees not disclosed. yeah. so it's interesting.

Speaker 1:  yeah. the let's just talk about because we kinda skipped over the camera and sensors because for those of you that are wondering, well, what can it possibly do? right? so this ambient ai device, it could continuously understand which room you're in, who's present, what objects are nearby, the lighting conditions, time of date, ongoing conversations, user routines, household patterns. there's so much. like, it could anticipate needs. like, oh, i'm noticing you're heading into work. traffic is unusually heavy today, which kinda freaks you out, but whatever. that's where we're going. or you usually take your medication around now. have you taken it? this would be great for older people, for sure. better than the alexa. but, i think this is, this is interesting. the flip side of this, which i don't know if we're covering today, is the, so this device that they're building and the fact that apple is suing, openai too. so, there's a bunch of things going on.

Speaker 0:  so, yeah, we can talk about whether we wanna talk about the trade secret lawsuit because we're not a 100% about which projects, are affected by this lawsuit. openai has, reportedly abandoned the io branding. in core follow-up filings. openai has committed not to ship any consumer devices before at least april 2027. i guess that might have to do with some of these, trade agreements. when people like johnny ives leaves, they have to have so many years before they can, do what have you. legal analysts that follow the trade secret, law note that even if openai ultimately wins the case discovery process alone creates a significant burden. so we're gonna see a bunch of stuff coming out about openai, which is probably been why they're delaying the ipo in you know, we talk about all sorts of stuff, but discovery means they get to look at all these emails and all this information and all this crap gets out. and if they don't have all their ducks in order, it might have a real impact on their ipo. so i think they'd like to see that behind them before that happens. right. it's not their first piece of hardware because they do have, this codecs micro for developers. i don't know if you've seen it before. it looks like this. i'm sharing it on the screen. yeah.

Speaker 1:  that's device. yeah. the codecs. yeah. for yeah.

Speaker 0:  so i'm trying to get it to show you. there you go. it shows it for a second then back to this guy. but it's mostly aimed at developers. yeah. and that was launched already. you can buy it off, work louder. it's interesting. so we'll have to see if we can get to a point where consumers are ready for this because the google home speaker, the alexa, and siri all did not

Speaker 1:  do well. alexa home speaker did well.

Speaker 0:  well, i don't think it well, i you say do well. i don't think it's a profitable device for them.

Speaker 1:  i think so.

Speaker 0:  i'm gonna be wrong. i'll look it up later. and, yeah. i guess my point is i don't think the intelligent speaker market up to this point as a whole has been successful, because it doesn't have enough functionality on the back end. we got people who wanna play music. we got people who wanna do some amazon stuff. i use siri to set alarms, but i don't do that. when i'm talking to apple people, they use siri a little bit, but not a lot, because, you know, it doesn't work consistently enough, and there are a bunch of other things. like, these are the reasons why they really haven't taken off in the consumer market. they have, i think, in the geek market where i have a google home and an alexa somewhere. i use the google home just for playing music, and i have the microphone off on it. so the point being is they weren't able to do much before. you could tie in a few things. you could get it to play the radio, tell the news, maybe play a podcast, any, like, audio type stuff. great. now we're gonna you're gonna have functionality. the question is, is that enough? and will consumers have enough functionality versus, say, geeks who have all these things installed with agents and things that can get things handled? are we moving to every person's gonna have multiple agents whether they know how to use them or not, and they're just gonna work for them. and now we're gonna be able to talk to them through devices like this or whatever they have handy. yeah. so,

Speaker 1:  so you were right. amazon historically sells echo speakers at or near cost. and so internal documents had revealed that the division lost over $25,000,000,000 between 2017 and 2021. so

Speaker 0:  okay. yeah. i, thought i read that somewhere or not even out those numbers, but it was and i couldn't confirm it. thank you. well, i guess but i mean

Speaker 1:  if you're not speaking to account the value of the data, right, which is why

Speaker 0:  the value of the data, being able to purchase stuff, like, if you can have hook it up your amazon account, account and automatically order stuff, we have that, you know, reorder this, reorder that, add this to my thing. yeah. the value inside that ecosystem might pay off, but the smart speaker as a whole has been a failure as a product. the market hasn't taken off. they haven't been profitable. again, because i believe the back end hasn't been able to do enough stuff. so now that we're moving to that, like, let's go. everybody wanted to rabbit our run for the large action model when it came out, like, it doesn't do any of the stuff and everyone's disappointed. but now we're getting to the point where the back ends are catching up to that promise. so yeah. you know, at this point, i don't have a computer that doesn't have an active agent on it that i can talk to to get stuff done, whether type two or whatever, because it's just so much more productive for me. me, and i'm not sure everyone else is like that. so is everyone else willing to do this and pay 2 to 300 plus have a subscription to chatttp? at what level, in this economy is the average consumer willing to take this and use it? so we'll have to see. i'm skeptical that the average consumer will do this. we'll see what about workplaces because then you have the data issue because it could help productivity. but i think more of the other product, we were just talking about the codecs micro would be better for that. so we'll see. i just think that i'm curious to see where this is going. it might just be the first proof of concept to a portable thing you carry with you all the time that knows your context, like, the pendant stuff that, you know, comes out and will record everything and then put everything in context out of the real time or later. and maybe match in an ecosystem with a visual glasses and what have you at some point. maybe. but i'm still i don't think this is gonna be successful because i don't think there's the mass market is ready for it. yep. so let us know what you think. openai is not simply building a smarter speaker. it's betting the cat conversation, vision, and contextual memory can become the next computing interface. would you put a camera equipped chatgtp companion in your kitchen if it genuinely made life easier, or is it continuous awareness a privacy line you will not cross? and then it's supposed to have a battery, and you can take it with you. i'm not quite sure how that works, taking it from room to room. because if it just sat in one place, then you'd how many do you need if you wanna continuously use it throughout the day? so let us know what you think. would you jump at this, or is this not something you would be looking for?