AI Trends at CES: It's Getting Real, and Awkward

AI Trends at CES: It's Getting Real, and Awkward

Show Video

- We're here at CES. We're gonna do something a little bit different. We're gonna have kind of a seated, more nuanced discussion about all the AI that we're seeing at the show. So I'm Emily. I'm here with my coworker, Brian. - Hi. - We cover different parts of AI, but we'll just kick it right off.

So Brian, what are you seeing at the show? - Well, I've primarily been here covering PCs. So laptops, desktops, components, and AI is in all of it. So we've got AI PCs and AI parts and AI features coming to all of it. - Mm-hm. - How about yourself? - So I mostly, I cover usually all the big names, Google, OpenAI, Microsoft, but a lot of them aren't here. So this isn't as much of a software thing.

So that gets to my other beat, which would be electric cars. In that space, it's all about self-driving and there's a lot going on with just the new administration and Elon Musk and the robotaxi push. So it's been self-driving cars for me.

- Yeah, well, and then outside of the PC stuff I cover, I mean, we're seeing AI in all sorts of wacky configurations. (Emily chuckles) We're seeing dedicated devices. Some of them are useful, some of them are not. We're seeing new services.

Smart pens that take your notes for you. - That's great. - Barbecues that tell you how things are cooking in your barbecue. AI is being applied to so many different things at this show. It really is the star of CES this year. - Yeah.

Well, last year, it was kind of like also the star, but we're kind of figuring out what that term meant. I think this year, we're kind of realizing there are so many different uses of the term, and you had a good theory about like there's the GPUs and the hardware and then the devices. So like talk a little bit about that. - Yeah, so when we talk about AI, you might be thinking things like ChatGPT, which are generative AI software experiences that people have.

But AI also is the underlying technology. So the chip architectures are changing to neural networks. - Yeah. - The way we process data is shifting from the traditional model to a new parallelized neural network type approach, and it's having a major effect on all of computing, whether it's branded as AI or not. So those two things overlap a lot because the technology feeds the experiences, but it really muddies the waters when we're trying to differentiate one thing from another when it's all called AI. - Right. Yeah, and the chip thing is very kind of dense and confusing to some people.

I mean it's a piece of hardware you usually don't see, but it's required to run all this stuff that all these tech companies want to do. So that's why Nvidia gave the keynote. They debuted some new chips, gaming mostly, right? - Yeah, so the Nvidia GPUs are mostly gonna be used for gaming.

You know, that's for heavy-duty visuals, but that same hardware can then be applied to a lot of different things because it's AI based and it has that neural technology built in, it can then be applied to a lot of different things. It gives horsepower for computing for all kinds of applications from new entertainment experiences to heavy-duty data science and everything in between. - Right, right. - So AI encompasses all of that, and that's hard for some people to understand because we're still feeling out what that is, and they're all still figuring out how it's applied to different things.

- That's true. - So it's all changing fast. - Most people here don't totally know what they're talking about. (Brian laughs) Like the companies are, they wanted to start using it. They wanna say AI, but it's a watered down term, and maybe some of the tech's not even there. Like Honda had a huge debut of their EVs at the show and they're saying, "Oh, they're gonna have level three self-driving." So the whole thing with self-driving cars is a scale up to six. - Mm-hm.

- And so like Tesla's at two. Ford, GM a lot of other companies are also at two. And so that's when you can take your hands off the wheel, and the car could drive itself on highways, but you have to look at the road.

So I wrote a piece, I actually found that very boring because you're just sitting there like this for like a whole road trip and you have to stare straight ahead. So level three is really a lot better because you can take your hands off and you can look at your phone and do all this. You can do a second task, but it's not really legal. (laughs) So there's no, I mean, Mercedes is the only car company that has permission to operate any level three system in the US and it's in one state, it's in Nevada. - Oh, really? - So we have, in this case, Honda saying we're gonna have level three self-driving cars.

And they do not have legal clearance for that. There's Waymo, which is also at the show, they allow passenger rides like Ubers that drive themselves. But that's a different industry. It's kind of like more business versus a car you could buy yourself. So it's very much evolving.

- Oh, and that's really one of the things that you see a lot at CES, is that there's this whole process you have to go through before there's a technology that's usable for everyday people. You have to have the underlying tech, the architecture, and the chips, then you have to have developers that have the tools to develop things for it, and then they have to create products and refine those products and get them to a usable point. And that whole process can take a long time.

Even if the revolutionary technology is there, the end use requires a lot of maturation first. - Yeah, I mean it's like the tech is so far ahead of when it comes to market 'cause I'm sure Honda can, in a closed course, operate a level three self-driving system. I'm sure all car companies could, but it's like how do you actually bring that to consumers? And at least with the laptops, it's pretty straightforward, like it has a chip or not. - Well, that's one of the things we're seeing, is that like Microsoft has been pushing Copilot PCs, which have, it's called a neural processor unit inside. And so they've been pushing that for most of the year now, but you have to have that hardware out in the market before people can start developing things for it.

It takes advantage of the hardware, and so it's only this year at this show that we're starting to hear about the features that people wanted when they were announced. - Got it. - So things like- - Like a one-year lag, two-year lag.

- Yeah. So we're finally getting things like language models that can summarize your emails for you or expand on what you've written or generate images for you on your device without having to go out to a server or online. And that's what people immediately thought of when they heard AI PC, and that is what Microsoft wanted to deliver. - Right. - But it just takes time between introducing the hardware and getting an experience that does that reliably for the user. And so we're just in the early stages of that second part where the tools are there, and now the experiences are starting to get out to the market.

- That reminds me of something else I've seen at CES, which is a couple different companies, including Nvidia, and I also saw a separate robotics company making robots or chips and being like, "This is a toolbox, like go forth. Like we're gonna make this robot so you can program it how you want. You're an engineer in this case."

- Yeah. - Yeah, so it's almost like they want people to build AI systems too. And like everyone's just trying to see what this thing can do. - Well, you kind of have two things going on. Two, you know, kind of metaphors to think of here are plumbing. You gotta have pipes before you have water.

But then the other one is that AI is kind of a gold rush. - Totally. - And so companies like Nvidia are making shovels just as fast as they can to make sure that they make money regardless of how this trend pans out in the long term. - That's true, yes. - So- - I mean, some of it, I wonder is like they say it's coming, it's coming, but I know it's very lucrative for them if AI does come and it does change our lives. So... - Yeah.

I mean I genuinely think we're at a point where it's no longer a question of if, it's a question of when. - Yeah. - But it's also a question of how that will be expressed because we have to explore different options to find what works and then develop from there. And that's gonna mean, unfortunately, a lot of overblown hype, a lot of promises that might not get followed through on.

But it also is going to mean some of these big futuristic claims that sound like total marketing fluff are absolutely true and often undersell what's actually gonna happen. - That's interesting. - We're going to get both. - Yeah, yeah.

And some will be in business and some will be at home. Yeah. - Yeah, a lot of what we've been seeing in the last year has been very focused on business because everybody wants to kind of squeeze some of the value and productivity out of it.

But I think really once it proliferates out, it's going to make our technology a lot more personable and let us make things more personal for us. - Right, right. - So it's gonna benefit both sides of the equation. - Right. I'm very mixed on that.

Like so I sat through the Honda press conference, and they had this video of this woman who was talking to her car like it was her friend, and she was saying, "This car is my partner by my side every day, taking me through new experiences." And it's like I know everyone always says they love their car, but I think they're trying to make it so you are actually in love with your car. (Emily laughs) - Yeah. LG is doing a thing where they're calling their AI features affectionate intelligence. - That's my award for worst slogan at the show. - That's just creepy. (laughs) - It's getting very creepy.

It's getting personal. I would say it can be getting a little sexual in the affectionate realm. So it's like they want it to be personal, but I would rather that just help me genuinely like anticipate my needs, but I don't need it to have like a human voice. I don't need to tell it about my day. I just want to know what I'm working on. - Yeah.

I think part of that is an attempt on the company's part to kind of clarify like we've been talking about, try to clarify what the value is for a person and why it's different. - Right. - And you know, if you can be a friend with your gadget, then I guess that's a selling point, even though it's weird. - That's interesting, yeah. Is it a value? They think we don't have friends? - Well, I think it's harder to explain data processing as something that people want in their gadgets. And so it's a matter of putting a face or at least a voice on the technology they're developing. - Right. - And trying to sell the value and the benefits in a way that people understand in an area that isn't clear and isn't easy to understand. - Right. I think it's a miss.

(both laughing) I think it's odd. I think it's very sci-fi inspired. And I know like, and Elon Musk did his robotaxi event. So he had this big event in Hollywood at a studio, and they had like robots dancing and like crazy stuff. And he said on stage like, "We wanna bring sci-fi movies to life." And I do think some of this is trying to get at that, you know, the sci-fi movies where, you know, you're talking to a computer and like with the "Her" movie, you know, you fall in love with it.

- Yeah. - Is that where we want to go? - Well, I think part of the issue is, I mean we talked about that cycle of technology and development and bringing things to market, but there's another side to this where you have to develop the technology, but you kind of also have to let culture change too. And people's understanding of what technology is and how it's used, - Right. - there's a lag there too. - Right. - I was in a QA with Jensen Huang of Nvidia yesterday where he was talking about how his generation was the first one to grow up with computers.

- Yeah. - And that meant in school, you started learning a bit about how to use computers, and then some people went into learning how to program computers, and then computers got applied to making computers better, but then also got applied out to all sorts of other industries. And he said that a similar shift will likely happen with AI because it's gonna change how we interact with computers, but then it changes how do we use it, how do we apply it, and how is it made useful out in the real world. And it's that same kind of shift.

There's just, there's so much unexplored territory still that we're defaulting to what we know and what we know is weird movies and books - Yeah, yeah, yeah. - that aren't always happy and aren't necessarily even accurate to what we're dealing with in technology. - Right. - It's just that's our frame of reference is the "Terminator" when really, it's a computer program that can help us write a better email. - Exactly. - Those are two very different things. - I'm just a little surprised

we're still in the "Terminator" phase. I thought that was like two years ago. (Emily laughs) - Well, I think it just shows that James Cameron makes a great movie - That's true. - and Arnold Schwarzenegger looks like a robot. (laughs) - Right, right. - And that's really an enduring image.

- Yeah. - But I think that's the cultural shift we're talking about though, - Right. - is that people think, "Oh, movies and sci-fi," instead of thinking, you know, email and spreadsheets - Yeah. It's way more fun. - because it's the everyday tools that we're going to be using every day. That's what it's going to be, it's just we don't know how it's gonna look yet. - Right.

I kind of wanna talk about that moment you witnessed with Nvidia CEO. (laughs) - Yeah, so at this Q&A, a bunch of media people in a room, people bringing mics around, there was a sound system set up there at the convention room at the hotel. And I guess Jensen wasn't happy with like the sound levels, what he was hearing his voice too much and not enough of the people out there. And he kind of had a rich, powerful flip out moment- - Which was really rare because he's Mr. Relatable, speaks in layman's terms. - Yeah.

He's a very personable guy and a very smart guy, but he got a little frustrated with the sound crew and kind of chewed them out in front of a room full of media, might have singled out a single sound guy and said, "I'm going to refer to all sound troubles by your name going forward." - Sebastian. - It was a little weird. And I think that's actually one of the things that's kind of scary right now, is that the big tech companies are already big and powerful and scary. - Right.

- Some of them are going to get a lot bigger because of this new technology, - Right. - and that is going to make certain individuals much more powerful. - Oh yeah, he's very powerful. - And so it's upsetting the balance of Silicon Valley and the greater technology world. And some of that's gonna mean it skyrockets new people into positions of power, some of that's gonna be putting powerful people into more powerful positions.

And that's a whole separate world of things to be concerned about. - Yeah. I did see the influence of Elon Musk a lot at the show. Of course, he's not here and Tesla's not here, but over the past year, he's been saying on earnings calls, it's all about self-driving, self-driving, self-driving.

So he wants to, he has shifted the company around AI, like he actually fired the entire team doing EV charging. And if you look on their jobs portal, now all the jobs are in the AI and robotics. Like he's shifting the company around AI and he's betting on it in such a way that even there, the number of vehicles they sold was actually down year over year.

So I think he's not really interested in making cars anymore 'cause he proved he could do it. Now he wants to do self-driving, which is what Steve Jobs said was, you know, famously the hardest AI problem you could ever solve. So he wants to solve it. So now the whole show is all about self-driving.

I mean, there's always a part of the show that's about self-driving, but this year really, it was the focus. And even Honda, like last year, they were talking about EV charging times and lightweight like batteries and like all that good stuff we want for EVs. They didn't talk about any of that this year.

Now they talked about their level three self-driving thing. - Hm. - And you can see like the narrative shifting and changing. And it also depends on the administration because Trump wants to get rid of the $7,500 federal tax credit for EVs, which you know, tens of thousands or hundreds of thousands of people are cashing in on and has fueled the industry. He's gonna get rid of it. Musk wants him to get rid of it because it will hurt his competitors more than him 'cause they're still scaling up their businesses.

So they need the credit. - Yeah. - Tesla advertises the credit on their website prominently as a sales tool, but he thinks, "Eh, it's all right. We'll still come out ahead." So the actual selling of EVs (laughs) is questionable. So we didn't have a lot of car companies here, but we did have self-driving.

And Trump and Musk are also working on relaxing and shifting and changing the self-driving rules to changes where would be allowing companies to put more self-driving cars on the road as part of a test fleet. So if you could only do like 30 and you now have 1,000, so they like scaling up their training of self-driving cars. And the other one is pretty scary.

It's like the self-driving car companies wouldn't have to report all the crashes to the Highway and Safety Administration. - Oh, wow. - So there's a lot of weird stuff going on, but I definitely saw that influence of a very powerful person shape the dynamic at the show. 100%, I was talking about it with other people, and it's now all of a sudden, we figured out EVs, they're cool, but they're still kind of just like cars.

So now we wanna make 'em drive themselves. (chuckles) - Well, I think it kind of speaks to how prominent this technology is going to be that every major player in tech is doubling down and tripling down on their AI development. - Yeah, it's all about AI like no matter what the industry. - So in a similar way to how technology kind of touches every aspect of life, AI touches every aspect of technology right now. And so we're gonna see that both in terms of what's offered to customers and what's used by businesses as well as what's happening in the industry and how that affects the broader world of, you know, politics and money.

AI is affecting all of it because it's such a dramatic shift. - Right. - And I think the fact that each of these major companies are betting so big on it kind of helps you see that this isn't a flash in the pan. - Yeah, I agree. - This is a revolutionary and perhaps generational shift. So it's a safe bet in that it's going to happen and is happening. - Right.

- The real question marks are how does it happen and who benefits? - Right, and how do we get enough chips for it (laughs) 'cause that's the whole thing. - Yeah, but that's a lot of what's happening this year, is that- - That's why that's so much this year about the chips. - Yeah, I mean we've seen major announcements from chip makers like Intel and Nvidia, Qualcomm, AMD.

All of the chip makers made big announcements this year. The NPU hardware that's used in PCs got a major upgrade, - Yep. - plus the software ecosystem that's gonna take advantage of that has more of a presence at this show than anything I've seen at any AI-related event in the last couple of years because the ecosystem is finally starting to take off, and that's gonna mean more practical uses of AI instead of empty promises. - Yep. - And more things that, I mean, readers and viewers are going to be able to try, we're gonna have to explain it, but all of this is really showing that this is moving into the real world beyond just marketing hype and future projection of what could happen. - Right. - We're there now.

It's things that are happening right now. - One last thing I'll say on just the chips and the promises I didn't realize. So the cars, you know, they need chips as well to do all this. And it did kind of come out at the show that all the power going towards the self-driving and on that chip actually could reduce the range of the vehicle, meaning drain the battery. - Interesting. - So it could have the same effect on a PC or a phone. - Yeah, there's- - So there's a lot to be worked out. (laughs) - Well, and one

of the other things is, I mean there's been a lot of concern about like the power usages and ecological impact of these giant server farms that they're using for cloud-based AI. But as we're moving now to device-centered AI, a lot of that becomes mitigated because now you're not gonna be charging and using your laptop significantly more than you already do. It's just allocating that power a little bit differently. And the overall trend in laptop chips every year is more powerful but also more power efficient.

You know, longer battery life is the selling point they push every single year - Right. - because they do get more efficient. And shifting to devices that are getting increasingly efficient is gonna really mitigate some of those concerning problems in a big way. - Yeah, it's incentivized 'cause it's helps reduce cost, and then it helps the consumer. So that's good. - Yeah, well, and for end users, it's super helpful too because if you can run AI on your device, that means that you are a lot more secure with your data.

You're not uploading everything to Google or some other big company - Right, right. - to potentially share or use their training data- - That's sending your text to your friend to like Russia. - Yeah, it also makes it more usable for businesses where you've got proprietary data that you can't share.

Now you can still apply AI tools to that in a secure and private way. I think that's gonna help a lot of people get more comfortable with AI tools in general. - And Apple's pushing it. Yeah. - Yeah, 'cause privacy

is a major concern if you have to share your data with something out there in the cloud. - Yeah, that's a whole Apple intelligence thing, is that they try to do as much on device as they can. They don't really tell you exactly the percentage of what's happening on device, but they do say the servers that goes to are also private, like kind of locked down, don't store your data, all that. - Well, and that's part of this whole category maturing, is they're finding new ways to approach that.

A lot of companies are using a hybrid approach. We've got on-device and in the cloud working hand in hand, and getting a lot more granular and mature about how do they handle data, how do they protect your data, and how do they communicate that clearly to users because it hasn't been very clear so far. And I think the AI companies, a lot of them have watched the growth of like social media companies over the last several years and seen some of the pitfalls there and now know to avoid those heading into this. - That's true, yeah. - And so my hope is that that makes for a more conscientious and responsible industry. - Right. - We'll see how that shakes out 'cause I mean... (laughs) - Yeah, yeah, yeah.

But you're right, all this speaks to your point of like it is actually coming because we're even talking about this stuff. Like this is all implementation questions like data privacy, like where are we getting the chips? So these are the conversations that I think we're seeing this year and what all these companies are trying to figure out. - Well, and you know, one of the conversations that I frequently have is people thinking that AI is in a bubble, and to an extent, it is because what happens with a bubble is you kind of flood the zone with money and talent to develop it when it's new because you don't know where it's gonna go yet. And so you just have to push it as far as you can and then start refining it. - Right. - And some companies will find a way to be both effective and profitable and sustainable. And that becomes, you know, the next Google or the next Facebook that continues on. - Right, right.

- And then a lot of other companies just kind of peter out because they don't crack the code. - Right. - And that's the function of a bubble, that's what Silicon Valley does, is you make all that work to find what does work long term. But that also means that there are winners, there are losers. - Right. - Some come out really big in the end and a lot just disappear. It's volatile, but it also gets some things done.

- Yeah. I guess we'll see. - Yeah. I think that's the biggest thing, is just things are very interesting right now and they're gonna be for a long time. - I think it's a little awkward, - It is. - like with the affectionate intelligence thing. (laughs)

- Yeah. - It's kind of like a teenage years vibe. I feel like I'm just in middle school at CES, like everyone's trying to figure it out. It's a lot of money flying around. I don't know. - Well, and I think part of that is an attempt to make these new things compelling when they're still trying to figure out how to make them useful. Once they're useful, then the use case is really clear, the value is clear, people will invest because it's useful.

- Right. - Until we're there, you've gotta figure out another way to make it engaging and exciting, and sex appeal does that, I guess. - I guess. - Yeah, it's weird. Technology's weird, people, and it's getting weirder. (laughs) - And it's weirder talking about technology in Vegas, which is a very weird place.

- It is. In some ways, Vegas is the perfect place for this show right now because the whole city is kind of in the uncanny valley of overcommercialized and oversexualized and just in your face all the time. - That's true. - I mean that's always been the case with CES.

It's always been a good fit, but more so right now because things are moving in just unexpected directions. - Mm-hm. - All right. Anything else? - Ah, I could probably talk about this for hours.

- I know, me too. I'm trying to be efficient. - But in terms of what we've seen here at CES, I mean, what has you excited or at least interested? - Well, I do think the self-driving stuff, I'm intrigued by because I struggle like to grasp the benefits of it 100% while it's in this immediate state. Like I said, with the boring, you're just staring at the road. Also, I was less alert and I'm just like, how is that gonna shake out? And then I don't know, like do we want that? And then is it gonna just tank the car industry because it's an impossible problem to solve? (laughs) Like you can't, it's gonna take 20 years to solve that problem, like solve. So like what are you gonna do in the meantime? So just there's a lot of questions that came up for me. - I think a lot of our natural intuitive timeframes for how things are going to develop are gonna be proven very wrong - Really? - because with AI, you're not just getting rapid change, you're getting compounding change.

- Right. - I mean, if you look at like generative AI, we've gone from, you know, text where you can get natural-sounding phrases back and forth to now getting natural-looking video based on the same sort of a text prompt in two years. And that same level of development is happening in many different areas that come together. - I guess the way I think about it is like I know you could do it and you can do it in a lab setting or a small group of people like right now. Like .01% of people have a level two self-driving car. I know you can do it.

- Yeah. - I feel like with technology, there's just a huge gap and a huge lag and like the pop general adoption. I mean there's still people who don't know how to use a computer today. - Yeah. - And so I feel like technology has a big problem with true implementation, and that's where I'm like just wondering if it'll ever happen. And in the self-driving car case, cars are very, you know, a lot of people have cars obviously.

So I think about it really broadly, not just like, could it be done in a close course? Could it be done if you had enough money? And also governments getting involved, which theoretically is supposed to be applying to everybody. - Yeah. Well, I think one of the real promises of AI is actually making technology more approachable for those end users, you know. - Yeah, maybe

that's what it needs. - If I can interact with my computer in natural language like you and I are talking right now, that's something you can do, you know, without having to have a specialized knowledge of menus and shortcuts and functions. - Yeah, yeah. - And so a lot of that is opening up and becoming more accessible. But also it's important to remember that AI isn't just a system, it's like several overlapping systems.

And when those start developing, the effect, you get a compounding effect and things coalesce a lot more rapidly now than even they did two, three years ago. So I think things are gonna change. They're gonna start changing faster, which is also gonna take us in weird new directions. But I think we're gonna see some really impressive changes in the next couple years. - Yeah, yeah. Well, we'll see. I mean, last year, AI was I think a true buzzword. This year, we're getting more implementation.

Next year, we'll see. - Yeah, and we're still seeing it in really gimmicky applications. I mean, I got pitched an AI barbecue, an AI-powered kitty litter box.

- Does it like analyze the char on the meat? - It does something like it uses sensors to watch the meat and determine, you know, (Emily laughs) is it good for what you're trying to do? And I don't know, I didn't look too far into that 'cause I've been busy watching other products but- - Yeah, that sounds like an immediate delete email situation. (laughs) - But I think the real takeaway here is just that it's being applied everywhere, both in the serious use cases that are going to stick around and the gimmicky ones that won't. And it's easy to write it off as a gimmick, - Right. - but I don't think it is. I think it's gonna stick around for long term.

We just don't entirely know what form it's going to take - Right. - because it's still developing. - And gimmicky can be innovative. - Yeah. - I like new even

if it's weird. - Yeah. It's what makes CES fun for us anyway. - Yeah, exactly. - Yeah. - All right.

- So that's us at CES. Hello from Vegas. (Emily laughs) (laughs) We'll talk to you more. - Wanna give it like, sound like, "I'm Brian Westover"? - Yes. All right. That's us from CES in Las Vegas. I'm Brian Westover. - I'm Emily Forlini. Thanks for listening.

- See ya.

2025-01-17 17:40

Show Video

Other news

X1 vs XL: AMS vs Toolchanger - What's better? 2025-02-11 16:00
Google, DeepMind on AI future, Deepseek; Trump Plans 25% Metals Tariffs | Bloomberg The Pulse 2025-02-13 19:22
How Infrastructure is Powering the Age of AI 2025-02-09 20:43