OpenAI vs DeepSeek vs Big Tech vs... Will There Be ONE Winner?!

Transcript is auto-generated and may contain errors.

Hey folks, I'm just headed to CrossFit here. I think one of my mics is busted. Um, one of the videos I went to go upload this weekend. The audio is just like not um but it wasn't dead silent. It was almost like this just like low lowrade buzzing or distortion kind of thing. So, trying out the other mic this morning. I hope it doesn't die on me. Um, we're gonna go to the comments though for today's topic and it's a shorter drive to CrossFit. So, kind of reserve this one for this drive. But the can't remember the commenter's name. I apologize. uh but it's about AI and if we or if I believe that we're going to reach a point where we have basically one of these larger companies that is so far dominating uh the AI space that there's not really competition for others. I think that was kind of the framing like is there sort of like do I believe there's one company that comes out on top.

Um so friendly reminder if you want questions answered about software engineering career development um I'm not an AI expert by any means I do not want to come across that way uh but if you have questions about it related to software engineering and stuff then happy to try and answer leave it in the comments otherwise you can send a message to dev leader on any social media platform that's my main YouTube channel and it's Nick Cosantino on LinkedIn if you write into me I can keep you anonymous. If you leave a comment, it's there for everyone else to see. So, just keep that in mind. Okay. Um, I think, you know, super high level, my thought on this is that I do not see one company that just comes out on top and then, you know, uh, makes it such that there's no one else that can compete.

Like, there's sort of one victor here. Um, I don't see that being the case, especially because right now, number one, I don't see that being the case currently. We have many competitors in the space and I think that's a good thing. And the other thing is that even if we look at other technologies, I realize the questions about you know AI models and things like that but if we look at other technologies even if we have a company that is sort of leading I I don't know of a particular industry where there is only one okay so like I work at Microsoft let's talk about the operating system right there is more than one operating system it's not just Windows If we look at uh video cards, right? Um for a long time, it was not necess unless you're, you know, you want to just be a fanboy or fan girl, it's not just like Nvidia was the only uh company, right?

Um that was really leading here. If we look at processors, there's been points in time where um Intel was stomping AMD, AMD was stomping Intel. There's always been um back and forth in different technology spaces and um I personally think that we're going to see this kind of thing uh with models. I think that we'll see um this kind of thing sort of express itself in different ways. Um and I don't know enough about I don't know the characteristics of models or even different technologies in the AI space to elaborate on this further but uh we may find that some companies have a sort of a niche for certain aspects of these things. So, just as an example, I anecdotally, and not even my own anecdote, just from hearing from others, I'll hear people say, "Oh, like you're using this model for uh trying to have an agent run." Like, no, not a good idea.

You should go use this other model. It's just better at doing it. And it in my head I'm like I haven't really had the need to flip-flop between between models on different things because I don't know like anytime I've tried I find if I just like prompt in different ways I can usually get you know some GPT model to do mostly what I need. Uh but you know people are already calling out for agent stuff. I've seen people saying like don't use a GPT model use claude. Um, I've, you know, anecdotally hear this for other types of things. Creative writing, use this model versus that one. So, I'm I'm wondering if we will start to see more of that. And then drawing that direct comparison to like, oh, there's just one company that's totally, you know, stomped the rest out. that kind of conversation goes away because it's it's harder to to even make that comparison, right?

If it was like, hey, we're only talking about general purpose models. Maybe one company always does come out on top, right? Or maybe you run into this situation, I'm just making this up, right? where you have a model and to access it's more premium. So, there's a higher price point, but it's going to push you that extra few percent higher than the other ones. And then maybe um there are, you know, similar models that aren't quite as good, like just like a few% off, but they're way cheaper. So just as an example for the as a commodity you have like a cheaper model to access becomes more popular as a result. But if you really need like that super cutting edge I wasn't even in sport mode. Nice. Um if we really need that extra cutting edge model that we're willing to pay that premium. Right.

I'm just making this up, right, as examples. So, I think that there's so many ways that this could go that I just don't see a space where we like only have one company coming out on top. And I don't mean to say that like it's impossible. I don't have a crystal ball, right? This is just my my speculation about how I imagine this stuff unfolding. Um, something else I wanted to like bring into the mix. I was thinking about this last night cuz when I was going to bed, I was like, "Hey, like what video am I what topic am I going to pick? This is one that I was thinking about." But I saw a headline at some point yesterday. And I'm uh I'm guilty of like being just a headline reader. Uh just so you know. So in this case, I only read the headline and I didn't read further into it.

um or like the headline and like the you know the 140 character uh abstract of the of the article. Uh, and I hadn't thought about the combination of these two things before, but I think this was about um China and I can't remember the name of their um their quantum computer something Wukong um like Wukong Legacy or I can't remember. Um but I they have a a quantum computer, just one, right? Um, you can tell I don't know exactly what I'm talking about, but they were leveraging their quantum computer for training AI models and they were sharing like this is like kind of like a world first. They were talking about some of the error rate. Um but it kind of dawned upon me and again I don't know enough about this to be able to speak to the details but um my understanding and

I have such a limited understanding of quantum computers but my understanding is that if it's going to be allowing us to do like quantum computers in general are going to allow us to do things like and I have to come back to this point in just a moment but I'm assuming things like uh a lot of or you know some type of math massively in parallel then I wonder if that could be like the way forward for for training. Um now before I go any further on that because the extent of my knowledge drops off immediately um quite obviously as I mentioned quantum computers I don't know anything about them but something I've always found interesting and if you're someone who's watching this and you know about quantum computers or at least enough to exp or elaborate further on the usage.

I always find it fascinating that it's like when I hear about quantum computers, it's like we're going to be able to use a quantum computer to do something like brute force uh like encryption bypass that would have taken like millions of years or you know hundreds of thousands of years and it can do it in like 13 seconds and like that's pretty incredible. That's like a fascinating feat, but it seems like quantum computers are very good at very specific things. I still have not yet seen exactly what that specific thing is. I have not seen a comparison to classic computing to understand what that thing is that it does so much better. And then anytime I go to watch a video or read something on this, it's like, "Oh, well, they're using cubits." And I'm like, "No I understand because this is the only thing anyone talks about." I'm like, "I understand there's cubits involved.

What I don't understand is give me the classic comparison about what a computer is doing that a quantum computer does, you know, like orders of magnitude or exponentially better because I still don't know." Um, and when you say something like for what it's able to accomplish with encryption, it's almost or decryption, depending on how you want to say it, it's almost too much of a jump because I'm like, what is the operation? It's not just does encryption. It's it must be doing something. What's the operation that we can compare to classically? Because right now I really struggle to understand that. That's why I'm saying my assumption right now is that it's some type of mathematical operation. And the reason why that we would be able to do something like break encryption at like what what seems like it should have been, you know, many many years of uh processing time is just because it's able to do it like in in parallel is the the closest thing that I could try to draw a comparison.

Anyway, the point being that I wonder if what we're going to start to see more and more of because we have all the AI talk. We have, you know, all these companies buying up data centers, filling them with GPUs, right? Like people like companies buying power plants and stuff like it's getting pretty pretty wild. But I start to wonder does the does the quantum computing race start to tie into this as well where one of the reasons like people can nail um and I don't even know if it's the right use case right I'm kind of drawing are making a huge jump here but I'm curious like the for folks when I say folks like companies are getting quantum computers pushed further and further. Are they able to start leveraging that for training models? And that becomes like the ridiculous competitive advantage. Um, I just don't know what types of operations those are able to do uh that much more effectively compared to a classic computer.

I wanted to share that in this video because I was hoping that someone who does understand this stuff further could just like drop in the comments and be like, "Listen up, idiot. Here's how it works." And then I can understand and then I can follow up with uh a little bit more research. But thought it was kind of interesting because I had never even I don't know like put those two ideas together. I never put uh quantum computing and artificial intelligence in the same bucket at all. But I wonder if that becomes a sort of an interesting path forward. But yeah, I think you know main point here is that I don't really see it being one company that just you know air quotes wins. Um, I think that over time we may find ourselves in situations where there is one company that seems to have a clear advantage maybe even across the board for different uh, you know, characteristics of models and that kind of thing.

Uh, and then over time that shifts. I think it can go back and forth, but I just don't imagine a a world where there's like just one company that comes out on top. And I'm curious like do we see that with other technologies again like as a Microsoft employee uh and I've I've used Windows my entire life. Like I I actually don't like using uh the Mac OS. I find I don't know why but I find it like just not intuitive for me. Uh I don't like using Linux. Um but like those came before I was working at Microsoft. So, it's not like some weird like loyalty to an employer kind of thing, but like we don't have a like a statistically you could say there's a clear winner in different segments, but I don't know like that's I think I can't remember the stats like Windows isn't even the most popular popular operating system in the world.

It's Android, right? Um, if you're talking about like I don't know like personal compute then perhaps I think Windows is still probably winning. Um, but like you know there's a lot of Mac users. Um, point is like you know there's still even if someone wanted to say well Microsoft and Windows have the biggest market share there. It's not like there's no other people in the space. There's a really cool graphic, I don't know who published it, but it's like a it shows operating system popularity over time, and it's like a a fiveminute like animated chart. Um, it's super cool, but you got to watch it in like uh like double speed or something cuz it's it's over many many years. U and it's just super cool to see that like you have these operating systems, you're like, "What the heck even is that?" And then you end up seeing at some point like the rise of Windows where like it's just it's everything.

Um and then you see the mobile OS's come in. But it it's super cool. Um but my point is I just don't I don't really see places where we have no other competitor in technology. And I don't imagine we're going to have that um as like an end state with AI. like that's the you know one company reaches the end and then then that's it. I think it's a super interesting question though and uh who knows like maybe you know I'm trying to draw comparisons to other technologies and stuff. I realize I said OS's a few times but I'm I'm driving and my brain's not able to kind of uh branch out far enough to other things. I'm driving a car, right? It could be there's not just like one car manufacturer. Um even like with Tesla and electric vehicles, there's other vehic electric vehicle manufacturers.

Um but I wonder too like maybe AI is different. I I don't see why it would be, but maybe it's different enough of a technology where we don't see that where there's competitors in the space and there is this opportunity for one to like come out on top just seems kind of unrealistic to me. And I think for what it's worth, like in my opinion, it's it's a good thing for us as consumers that we don't have one person coming out on top or one company coming out on top. Competition is a very valuable thing for consumers. It gives us options, makes people compete with each other to try and win us over as consumers. So, I think we as consumers, we do want competition. But I think that's my thought on that. I think it's a cool question. Thanks for submitting that one. And again, just a friendly reminder, if you want questions answered, just leave them below in the comments.

I think that's the first one that I did kind of like purely on AI, and you can tell that like I'm not an expert in that. I'm not an expert in quantum computers either, apparently. So, um, but yeah, thanks a lot. Otherwise, I'll see you next time. Take care.