Speaker 0:  alright. speaking of hardware and ai, midjourney's hardware, so they wanna build this thing called the orb. and my question is, are the only profitable ai companies, hardware companies, as we just talked about, ai is losing money hand over openai. everyone thinks it's the shining star, which is great, but, you know, we're it's a huge, pile of flaming money. midjourney's hardware, the orb, has been announced. and, you know, the question goes, will it follow the humane ai pin or, the rabbit r one aimed at revolutionizing how we react, interact with technology but ended up failing due to several critical fall flaws? are we gonna see the same thing with this, or is it gonna take on something to the next level?

Speaker 1:  yeah. so the thing is, like, it's it's crazy that, so there was a canadian company called grok with a q that was getting into ai chips, and then they shifted to heart to software, actually. so, these guys are shifting to hardware. it's noteworthy. it does make sense because then they can build, an asic, like an application specific integrated circuit. but, again, it's, are they like, chip design is not easy. maybe with ai, it might be easier, but it's i like, nvidia does essentially have a hold the, on the gpu market. i know there's other companies, that are creating their own. i know, microsoft's working on one, apple, intel, amd, google, but, no one can touch nvidia. so the question is, is do they have the money to develop and fail to get to a hardware that can be used what nvidia produces.

Speaker 0:  well, yeah. no one can touch nvidia today, but, you know, we could've said that about until ten years ago. yeah. you know, and we're like, yeah. okay. and we can see that that is you know, they're collapsing, and we'll see if they can recover. but, you know, they are definitely waning in popularity. and this kind of hardware is a little bit different. content creation. i'm it sounds more like, our startup lumex who came out, and they had the world's not flat. so they've got a three d, camera modeling system that helps you take pictures and model stuff in three d in real time. yeah. so their hardware ambitions could represent a bold, exciting chapter for them because it's unprecedented unprecedented capabilities. imagine ai powered device that not only generate images, but also creates an entire three d environment in real time, providing artists and designers with new ways to visualize and execute their ideas.

Speaker 1:  yes. but, again, chip design is not easy. architecture doesn't matter what you're using it for, but trying to get the architecture right of the chip and everything so that you can like, look at even google, the biggest, like, who owns android. right? look how many problems they have with their first tensor chip that they designed in house. so it's the question is, do they have the staying power to go through multiple iterations to get to something they, that is usable?

Speaker 0:  yeah. i don't know. and i didn't read far enough to know who their product manager is here, but there's a deep dive on tronto starts dot com, because a lot of these other articles that have the great information are, behind paywalls. so yep. i break down these, compare a bunch of articles. so we've got the information. we talk about openai versus midjourney. the thing is, i think midjourney might be using off the shelf hardware from other people and using their own software, where openai is trying to create new hardware for faster ai generative processing, where midjourney is trying to produce something with cameras and chips so we can do three d rendering in a device, faster in real time and capture three d environments, which is different. so i think, you know, you're taking potential off the shelf components and putting them into a product. so you need a great project man product manager integrate them into something sweet, versus what nvidia and openai are trying to create new hardware to accelerate the process behind all this stuff. so they could just plug in somebody else's chips to do the ai work.

Speaker 1:  definitely. yeah. we'll have to see.

Speaker 0:  but i question, you know, rabbit r one, although it's a, flop, it generated a lot of excitement and made a lot of money, selling those. so as long as they didn't take a loss on building and shipping those, they should be doing alright as a company and giving them a chance to come up with a second product. i remember talking about it saying, you know, i didn't really care about the hardware device. i wanted the back end software that does the interaction. yeah. and, you know, that's i'm still looking for that.

Speaker 1:  although, they, i think someone had evaluated it and said it pretty much could run on any android device.

Speaker 0:  yeah. it makes sense because you know, it just sounds like it's just a raspberry pi with a camera not a lot of cpu power. so i cracked the question. we'll keep an eye on midjourney's hardware. yeah. and that's the thing

Speaker 1:  with the rabbit r one. just the thing is it was heavily, heavily underpowered as far as an android device goes. so they might have made money off it based on that.

Speaker 0:  and speaking of we'll keep an eye on midjourney and what they're doing here. it's interesting to see, the text to graphic companies, one, are in a lot of lawsuits, but, two, there's a lot of them out there using different models. they're finding it harder and harder to differentiate. yeah. we'll have to see if this is gonna be a game changer for them. yeah.