Altman and Ive team up to build something new with IO - DTNSB 5025
28:06 Share

Ever wonder what your lashes are destined for? The cards have spoken. Maybelline New York Mascara does it all. Whether you crave fully fan lashes with lash sensational, big, bold volume from the colossal, a dramatic lift with falsies lash lift, or natural looking volume from great lash, your perfect lash future awaits. Manifest your best mascara today. Shop Maybelline New York and discover your lash destiny. Shop now at Walmart. This is a mini meditation guided by Bombas.

Repeat after me. I'm comfy. I'm cozy. I have zero blisters on my toes. And that's because I wear Bombas, the softest socks, underwear, and t-shirts that give back. One purchased equals one donated. Now go to bombas.com slash ACAST and use code ACAST for 20% off your first purchase. That's B-O-M-B-A-S dot com slash ACAST and use code ACAST at checkout.

What does the term digital workplace mean to you? In today's world of AI, robotics and AR, it's much more than an office full of people on computers. It can be anywhere from a factory floor to the top of a crane to the cockpit of an F1 car. Wherever your digital workplace may be, TeamViewer's mission is to make work, work better.

How? By securely connecting your people with the data, expertise, or insights they need in real time to make work more efficient. By automating and streamlining IT and OT support to fix problems before they happen to make work more productive. And by bringing all the possibilities of a secure and flexible digital workplace to all your people everywhere to make work more innovative.

So discover how TeamViewer can make work, work better, wherever it happens across your business. Learn more at teamviewer.com slash work better.

Picture this: you're in the garage, hands covered in grease, just finished up tuning your engine with a part you found on eBay. And you realize, "You know what? I could also use new brakes." So where do you go next? Back to eBay. And you've got eBay Guaranteed Fit. You order a part, and if it doesn't fit, send it back. Simple as that.

So when you dive into your next car project, start with eBay. All the parts you need at prices you'll love. Guaranteed to fit every time. eBay. Things people love. This is the Daily Tech News for Thursday, May 22nd, 2025. We tell you what you need to know, follow up in the context of those stories, and help each other understand. ♪

Today, GT Tandon tells us about how the cost of AI training and inferencing is going down. And Johnny Ive is getting cozy with Sam Altman and OpenAI. Sounds like a lot of AI in today's show. I'm Jason Howell. I'm Jen Cutter. I mean, what else is new, right? Let's start with what you need to know with a big story.

And yes, it has something to do with AI. But hey, this is the world we live in right now. OpenAI announced a $6.5 billion all-stock acquisition of Johnny Ive's AI startup, LoveFrom. Now, Ive, of course, led the design of the iPhone, the iMac for Apple. I mean, legendary designer. And I think even that's putting it lightly. Ive is now going to take an expansive role advertising.

at OpenAI leading the creation of the company's first physical AI product or products, because I'm sure this isn't just one, this might be many. The project is codenamed IO. And there's some interesting news that seems to kind of contradict itself out there. Because, you know, initially when the news happened, it was like, we don't know anything about what exactly they're making. Now we're starting to get some leaks from credible sources and some of it's conflicting. So, yeah,

Some of the leaks say that they're not aiming to make a phone or a wearable or an existing form factor. That being said, Apple analyst Ming-Chi Kuo says the device could be larger than Humane's AI pin, which was already pretty large. This is a quote, form factor is compact and elegant as an iPod shuffle, end quote. This would be or could be worn around the user's neck.

Quo also reported that the device is unlikely to have a display. It might feature built-in cameras and microphones, connect to smartphones and PCs for the computing and the display augmentation. The Wall Street Journal reported the device will have real-world awareness capabilities.

as an always-on assistant that is, again, augmenting daily life, finding contextuality in the ways that we maneuver the world, essentially. And Ives' role at OpenAI is not limited to just this hardware development. He's also going to be shaping the product's overall vision, user experience, how those AI models are integrated into physical interactions,

And so this seems to be OpenAI moving from its software only phase to software and hardware phase. Are you excited, Jen? I'm excited to find out what isn't a phone wearable or existing form factor. Totally, right? Like this, it's kind of confusing. Like what could this be then?

I was like, okay, so it's not those, because my first thing was like, okay, if it's not a phone or wearable, it'll be like a speaker. But then they say like no existing form factors. I'm like, okay, well, speakers exist. And then, you know, you heard about the shuffle kind of shape, which is a good shape. Bring that shape back. Great job. Yeah.

But then if that's on a necklace, that's a wearable. So where are they going with that? That's exactly it. Like, it's not a wearable, but it might, but it's going to be worn around your neck. Ah, you said worn. Yeah, exactly. It's like, it's always on and always ready. I'm like, great. But if it comes with me, isn't that kind of wearable? Like, say, glasses. Yeah. Yeah. It's interesting. I mean...

I'm torn on this because I do feel that kind of this moment that we're in with technology, artificial intelligence is a pretty big moment and that there is, you know, right now we're so used to using the AI on our smartphone because we're so used to always having a smartphone with us. And it's just what it's baked into.

I do believe that there can be a form factor for AI that isn't the smartphone. I'm not certain that we've seen it yet. And I think the Humane AI pin was a spectacular failure and, you know, kind of tried to do the it's unlike anything you've already had before. But yet a lot of what we're hearing here kind of gives off similar vibes. It kind of sounds like the same thing, but different. And I don't know how it's different. Yeah.

And on top of that, I haven't heard anything about a timeline. Have you? Yeah, no. I mean, I think I heard I heard yesterday and I don't believe that I transported it into my notes here. But I want to say sometime 2028. So probably not even next year. I think it's a couple of years is the guesstimate. But I mean, we're so early, like unless they've been working on this for a while and we've been hearing rumors forever.

for a little while that they were interested in doing something in the way of a partnership, not necessarily that OpenAI would outright acquire love from, you know, for this. But we have been hearing that they've been interested in doing something together. So this is just kind of

You know, cementing that, you know, setting it in stone and saying, all right, it's happening. OpenAI wants to be more than just an AI data software, whatever you want to call it, company services company. It really wants to have a hardware wing. And who better to help with the development of that than the guy who basically made the modern smartphone what it is at Apple?

I mean, it's, you know, he's a, he's a pretty big get as far as that's concerned. He's pretty good.

Well, and now this is a signal to the market that, okay, OpenAI is working on this. So now everybody's going to try and rush to market their things. Who knows? Maybe the Humane AI pin wasn't a bad product, although I think it was probably a pretty bad product. Maybe it was just too early. Remember, Google Glass was too early and it was, you know...

People have their opinions as far as whether it was a good or bad product. But I think one thing we saw at Google I/O this year with these XR glasses is that's Google I/O 10 years later or 12 years or however long it's been. I think it's pretty safe to say that was in too early.

All you have is the now context. You don't have the future context. And so it's really unpredictable. And I'm curious about your take on this, too, because another aspect of this is Apple, which is unrelated to this news specifically, other than the fact that Johnny Ive, you know, was a legendary designer at Apple.

But unrelated to even that, Apple's been having a real hard time with AI, you know, very notably the last couple of months been taken to the shed for, you know, failures on its part when it comes to Siri and its AI promises and everything like that.

And they're relying on third parties for their AI execution. This could potentially be a big blow for Apple considering where it's at now, especially if this signals the beginning of pursuit of something like a post-smartphone world, which I still think is very far off in the distance. But potential. And if that happens, that's not good for Apple based on where they're at right now with AI.

Oh, I'm sure they've got their friends who are friends with Johnny who are keeping them a place of a situation. Yeah, no, Apple's the big behemoth moving very slowly in the direction that Apple is moving in. And I think a lot depends on like...

okay, how far along is this? Is it ready to be shown? Is this going to slip from 2028, 2029, 2030? So they may have enough time to manage their own ship over there. Yeah, that's true. You know, we are very, very early on this. Yeah, I'm super curious. Like I want...

I want to see a company do this right because I do believe that there can be an AI focused piece of hardware that isn't the smartphone. I just I still have no idea what that even looks like. And I think, you know, of all people, of all like power duos, if you want to call them that, this is a pretty solid power duo. So I have a feeling they'll come up with something compelling, even if it's not successful, it'll be compelling.

DTNS is made possible by you, the listener. Thanks to Mike Cortez, Erwin Sturr, Ken Hayes, and our new patron, Alex. Welcome, Alex. Yeah, welcome, Alex. And all of you, you're all awesome too. Picture this. You're in the garage, hands covered in grease, just finished up tuning your engine with a part you found on eBay. And you realize, you know what?

I can also use new brakes. So where do you go next? Back to eBay. You can find anything there. It's unreal. Wipers, headlights, even cold air intakes. It's all there. And you've got eBay guaranteed fit. You order a part, and if it doesn't fit, send it back. Simple as that.

Look, DIY fixes can be major. Doesn't matter if it's just maintenance or a major mod. You got it, especially when things are guaranteed to fit. So when you dive into your next car project, start with eBay. All the parts you need at prices you'll love. Guaranteed to fit every time. eBay. Things people love. If you're a lineman in charge of keeping the lights on, Grainger understands that you go to great lengths and sometimes heights to ensure the power is always flowing.

which is why you can count on Grainger for professional-grade products and next-day delivery so you have everything you need to get the job done. Call 1-800-GRAINGER, click Grainger.com, or just stop by. Grainger, for the ones who get it done.

This is Paige DeSorbo, the co-host of Giggly Squad. I have exciting news. McDonald's has all new McCrispy strips. It's chicken made for dipping. Tender, juicy white meat chicken with a golden brown peppery breading. It's chicken so good it deserves its own sauce. The creamy chili McCrispy strip dip, a sauce that's creamy, savory, and sweet with a little heat. But it works with any of our sauces. I'm personally a barbecue sauce girl. Even sometimes I like ketchup. I'm just like...

basic sometimes, but I also need it in addition to any new sauces I'm trying. With a new creamy chili McCrispy strip dip, it's chicken made for dipping, only at McDonald's. Work management platforms. Ugh. Endless onboarding, IT bottlenecks, admin requests. But what if things were different? We found love.

Monday.com is different. No lengthy onboarding. Beautiful reports in minutes. Custom workflows you can build on your own. Easy to use, prompt-free AI. Huh. Turns out you can love a work management platform. Monday.com, the first work platform you'll love to use. Ever wonder what your lashes are destined for?

The cards have spoken. Maybelline New York Mascara does it all. Whether you crave fully fan lashes with lash sensational, big, bold volume from the colossal, a dramatic lift with falsies lash lift, or natural-looking volume from great lash, your perfect lash future awaits. Manifest your best mascara today. Shop Maybelline New York and discover your lash destiny. Shop now at Walmart.

There's more we need to know today. Let's get to the briefs. In other OpenAI news, the company announced a major expansion of its Stargate initiative, launching Stargate UAE, a large-scale AI data center in Abu Dhabi. It's developed in partnership with G42, a UAE-based firm, and supported by Oracle, Cisco, NVIDIA, and Japan's SoftBank.

The facility is expected to first deliver 200 megawatts of computing capacity by 2026, eventually scaling up to one gigawatt. This is the first partnership in the company's Open AI for Countries program, which is aimed at helping governments build its own AI capabilities in coordination with the U.S. government.

Yeah. And it seems open AI is very, you know, in a very comfortable place with the U S government, right place, right time. This is, this is an expansion of that. So very interested. One gigawatt, one gigawatt, 1.21 gigawatts. Sorry. It's every time I see one gigawatt, I can't help. It's mandatory. It's kind of mandatory. Even if I'm not saying it out loud, which I just did, it's happening in my head.

Amazon is piloting a new feature on its mobile app that provides AI-generated audio summaries for select products that resemble a two-host podcast experience. Kind of like what we've heard from Notebook LM that everybody freaked out about a handful of months ago. Oh my goodness, it sounds like an audio podcast.

These are going to feature product features of themselves, user reviews, web sourced information, all collected inside of this audio experience. A quote here, the highlights button is going to appear on product pages, starting with products from a number of companies, you know,

like Ninja Blender, Shox, Open Run Pro, SafeRest, Waterproof Mattress Protector. I mean, there's a bunch of them. I just pulled a couple of randoms and a whole lot more. Each summary will begin with a disclaimer, of course, about the AI involvement in that audio. And it's currently limited to a subset of US customers' plans to expand to more products and users in the coming months. So Amazon turns into a podcast app, sort of. Yeah.

Yeah, well, I'm glad they are starting, not ending, starting with the AI disclaimer, which I think is very important for consumer trust. And I'm not surprised that this always starts in English. I think I've mentioned this even on briefing before that when I'm looking at Amazon products, sometimes there's just not a lot of reviews in Canada because we're a much smaller population. So sometimes...

Sometimes it'll pull from US reviews. And a lot of times I will get, obviously, the Quebecois reviews and for some reason, the Mexican reviews, which I will have to translate because I don't know Spanish. I'm very curious to see how that goes because like...

you are also competing with the AstroTurf reviews. I'm like, okay, if it's an AI summary of AstroTurf, that's not helping me. It doesn't matter whether it's a podcast or not. Totally. How much do you really trust the reviews that you are seeing on Amazon at the end of the day? And how shilly does it sound when you're listening to two podcast hosts tell you what's so great about the Shox OpenRug Pro? Yeah, well, I had just bought a blender, so I'm going to need someone to report back to me on what it says about the Ninja one in podcast form. Yeah.

I mean, if you're listening to product review, like Amazon customer reviews as podcast, I'm just here to tell you right now, there's a lot better podcasts out there. Just saying. A massive security breach exposed more than 184 million user credentials that were stored in plain text on an unsecured server.

Jeremiah Fowler discovered the 47.42 gig database containing unencrypted emails, usernames, passwords, and login URLs for services and products by Apple, Facebook, Google, Microsoft, PayPal, government and financial institutions. The data is suspected to have been harvested by InfoSealer malware and some credentials were verified by Fowler. The owner of the database is currently unknown.

184. I mean, you know, again, like I feel like these things happen all the time and I'm constantly fighting to not be numb to it. But 184 million user credentials, plain text, like that's insane. That's a lot of data. That's a lot of bad information. I kind of want to know who the owner of this database is. How long has it been sitting there? I mean, seriously, hopefully we find that out someday soon.

Xiaomi unveiled its first in-house flagship smartphone chip, the Xiaomi X-Ring 01. It's a three nanometer chip, 10 cores. It's seen as a direct rival to Qualcomm Snapdragon 8 Gen 3. So top of the line. The chip drives the company's new 15S Pro smartphone and Pad 7 Ultra tablet.

The company also launched the X-Ring T1 chipset, which is designed for smaller form factors like smartwatches. And also, because Xiaomi had like a number of different announcements in the last 24 hours, Xiaomi also unveiled its YU7 electric SUV. It's a direct competitor to the Tesla Model Y.

Scheduled for release in July, the YU7 has a top speed of 253 kilometers per hour, can go from zero to 100 in 3.23 seconds, has a maximum range of 835 kilometers on a single charge. And if you're inside, you get a panoramic hypervision display spanning the dashboard. There are no firm pricing yet, but it may be priced according to CEO Li Jun,

around 60,000 to 70,000 won higher than the Model Y, which starts at 263,000 won-ish. If I convert that into currency that I understand, it's around $36,574. I'm always happy to see the range increase on these vehicles because I am someone who does rather routinely drive about 400 kilometers in a day, which is why these things have never appealed to me. Yeah, well, like Toronto to Ottawa, it's a hike.

And usually I'll try to stop in the middle, but I just, I drive places and help friends get to airports and such because I'm cheaper than the bus. And yeah, I would love to help the environments and stuff, but I also need to keep waiting because I want cars with buttons. I'm very spoiled on my old car with buttons. Yeah. Well, if you get this, you're going to have hypervision. So that's, that's what, I don't know what that says about buttons, but it sounds like a lot of screen to me.

In related EV news, Chinese automaker BYD surpassed Tesla in European EV sales for this April. The company registered 7,231 battery EVs compared to Tesla's 7,165, a milestone that reflects BYD's aggressive overseas expansion and oppressive year-over-year growth.

Tesla's European sales dropped 49% from the previous year due to its aging product line and political backlash from Elon Musk's involvement in the U.S. presidential administration efforts. Yeah, this was really the first time that Tesla was surpassed in that regard. So kind of, you know, I guess, what would you call it, a bellwether moment or whatever they say. So that's notable.

Certainly not limited to Europe either. There's a lot of backlash on Tesla right now, so I'm not surprised.

Signal updated its desktop app to block the use of the new Windows 11 recall feature that captures periodic screenshots to create a searchable activity timeline within Windows. In the absence of an official API, there is none, apparently, to opt out of recall. Signal employed its own DRM technique to prevent that screen capture of the interface, you know, to protect its users. It's about, you know, protected messaging, and Signal cares about that.

Signal developer Joshua Lund criticized Microsoft for the missing API, stating the company had no other option but to implement this workaround. Kind of surprising that Microsoft doesn't have this. Maybe there's a reason. Maybe they just really, really want you to use recall. But there will always be reasons why companies or users might not want to. In this case, I think it's pretty valid. Yeah, there's nothing people love more. It's getting forced to do things. Yeah. Like...

I was trying to think of like, okay, what's a comparable example? Like everyone got quote unquote forced to HTTPS.

That I believe was a net positive. But like for Google and Gemini and the force integration, my friends at IT at major corporations, they had to find their Google rep and straight up yell at them saying like, you cannot have this on our server. It's like, you are going to cause us issues. We are in law. We are in finance. You have to turn this off. And it was able to be turned off if you are a major corporation who has somebody specific to yell at. Yeah.

Oh, boy. So I think like if people have to do this to Microsoft per company to say you need to turn this off or we need to find alternatives, it's a long, slow road. I have a feeling Signal is not alone in this. And I would imagine eventually Microsoft, if it's not already working on it, would create an API for this. But I also admit I'm not super close to Microsoft on a personal level. So, yeah, there's that.

Switchbox introduced the Lock Ultra, a retrofit smart lock that brings 3D facial recognition to home security. It's priced at $159.99 and pairs with the company's keypad vision, an additional $99.99 that drives the facial recognition technology. The system supports multiple authentication options, including fingerprint, keycode, NFC card, smartphone app, and geofencing. It's available for pre-order now with shipments expected in June.

So how do you feel about facial recognition entry at your front door? You feel all right with that? I don't use smart locks. I don't use smart thermostats. I don't have a smart speaker. None of this is in my house. So I am the least qualified person to comment. Yeah. I mean, when I was looking at this, I was like, all right, so it's got facial recognition. It's got fingerprint. It's got key code, NFC. It's got a smartphone. It's got like all the things and all that kept running through my mind is the more like ways you

give as options for access, the more or the less secure it actually is, right? Because there's always a way around depending. So it's one consideration to have anyways, you know. Those are the essentials for today. Let's dive a little deeper.

Yeah, we hear a lot about the cost of running AI. Last Friday, we actually spoke with Andy Beach about the cost of chips coming down. But the cost of both training and inferencing has come down a lot too. GT Tandon from TELUS Digital helps us understand how you can take advantage of that. GT, thanks for joining us. I appreciate it, man. Thank you. Thanks for having me here.

So we wanted to talk a little bit about how models have become a lot cheaper to train and inference. Obviously, deep seek caught everybody's attention, but it's not just that. Help us understand a little bit about what those efficiencies are. There is a journey behind this, gaining those efficiencies, right? So just to set the context right,

most of these models probably go through that very typical three, four things like where you are actually storing data, you need to train on that data and then you build some inference on top of it, right? And all of these have costs associated with it, like where you data storage costs, which have become like significantly cheap over the years. We all know that. But the training cost is the one that was kind of

that everybody had their eyes on when OpenAI came up with ChatGPT 3.5 about three years back or 3.0. And then over the next last three years, every few months, there's a model that is coming up and they are using their own different algorithms on the training itself.

And then the inference itself, right? Every time you ask a question to these models, how are that question interpreted? What is the compute that goes in the back and then it goes back and finds the right data for you, right? So all of these three dimensions have their own play in the overall cost model of how these Gen-AI are being costed or their entire process is being costed.

And the Moore's law in itself has one of the big contributors to it. And we are all aware of the Moore's law where every few months, every 18 months, the chip processing speed goes up by 2x. And that is one of the big things that is contributing to the actual training of the data itself. Now you can get the model getting trained on that data much faster because now your chip speeds are increasing. So that is one. And

And plus, there is a lot of chip design and NVIDIA is obviously at the forefront of it. There's a lot of chip design that is being now done specifically for AI training, right? Initially, it was chips that were meant for any kind of compute, but now chips have been designed specifically for those kinds of use cases. So which makes it much more easier to actually get the data trained.

Now, I mean, the token cost itself has reduced for the same reason in the last two years. It was, I think, $20 or something, or sorry, $4 per token. Now it has come down to almost 40 cents per token, like has come down to almost 10 times. And that is, again, for the same reason that you can actually now parameterize a lot of algorithms through that token.

Many of the open source technologies have come up with their open weight models, which means that their training and the data has already been done. Now you are just playing with the weights and those weights can then be used to actually inform your own model output. And the third thing is the inference output itself, where it goes with the same logic that you are using compute for inference and that compute is now being done on chips and the

newer algorithms that are much more competing for those efficiencies. So if I want to get into training a model for my own enterprise, it sounds like it's more possible than ever. I don't have to rely entirely on the pre-made models that are out there.

Correct. Right. And there's like a ton of technologies that has already been out there, which are way because, because of the cost that has come down with those smaller models, uh,

As soon as you want to bring your corporate data into a model itself, because the cost has come down, the integration of that model, corporate data, along with all that pre-trained data that the chat GPTs of the world have and marry the correlations between them has become much more easier. Right. I mean, when I tell us digital and we have tools like FuelX and all that, which are using the same kind of a logic.

where we actually apply it to the corporate knowledge basis and then bring the combined knowledge of the universe along with that corporate knowledge and bring it to the response. Yeah. And does that make possible more bespoke models? Not just training on your own corporate data in general, but...

more specific use cases. Correct. Right. Absolutely. So that's the one dimension that has opened up significantly because of one is that because many of the models have become now open source like Lama and many of the models are coming up with open rate models. So which means that you can actually bring those models in house, refine them, fine tune them, make them bespoke to your own corporate objectives and then apply them. Right.

them. Right. So the cost of lower cost of those model development has not traveled into your own territory and you can apply that to your own. Can you, can you think of a couple of example cases where a company may take advantage of that? Yeah. So for example, I mean, and I'll start with that first on the training side of it, right? Because, uh,

Because there's still a lot of data that needs to be annotated and labeled to even be trained on it, right? So even on the training side, now given that most of the data that comes through all these open models have already been pre-trained,

Now for the data that is very specific to your domain, whether it is on your corporate side or whether if you're in scientific research or so on, you apply some of that model fine tuning through experts. Instead of data labeling, you are actually applying some model fine tuning through those experts that can actually learn from the synthetic data as well as from the data that corporate has and then apply that to actually the internal data that is coming to you.

So what it means is that instead of actually spending hundreds of people who are actually annotating your data, that this is a cat and a dog and a car and a boy. Now all that can be done through much more refined model tuning.

using those use cases. So that use case is very much prominent in many of our clients that we work with. Well, GT, thank you so much for taking the time to help us understand this a little more. If people want to find out more about this topic, find out more about what you do, where should they go? They should go to my LinkedIn profile and happy to engage with them on many of the topics that they have in mind on this. Fantastic. Thanks so much, man.

Thank you. Picture this. You're in the garage, hands covered in grease, just finished up tuning your engine with a part you found on eBay. And you realize, you know what?

I can also use new brakes. So where do you go next? Back to eBay. You can find anything there. It's unreal. Wipers, headlights, even cold air intakes. It's all there. And you've got eBay guaranteed fit. You order a part, and if it doesn't fit, send it back. Simple as that. Look, DIY fixes can be major. Doesn't matter if it's just maintenance or a major mod. You got it. As

especially when things are guaranteed to fit. So when you dive into your next car project, start with eBay. All the parts you need at prices you'll love. Guaranteed to fit every time. eBay. Things people love. If you're a lineman in charge of keeping the lights on, Grainger understands that you go to great lengths and sometimes heights to ensure the power is always flowing.

What do you want to hear us talk about on the show? One way to let us know is in our subreddit. Please come over and submit stories and vote on them at reddit.com slash r slash Daily Tech News Show.

We end every episode of DTNS with some shared wisdom. Today, Anon Jr. posted a reaction to the new Marshall soundbar that we talked about earlier this week. Anon Jr. says, as a guitarist, I'm all for a Marshall soundbar. I should have expected Marshall prices would come with it.

Yeah, of course. I hadn't heard about the Marshall soundbar. I am a musician, so I have respect for the Marshall brand. And boy, this truly is a soundbar that looks like a speaker stack in a practice space or something, which I can get behind that. It looks so cool. And I was showing it to my brother and he's like, this is great. This is well outside of Father's Day prices. Yeah, right. What is it?

north of $1,000. Maybe it is right at $1,000. Yeah. And then more in Canadian. We're like, well, we're going to be sensible and wait on reviews. Make sure that it's got some longevity to it. Certainly not the most expensive soundbar. There are good quality soundbars that are upwards much more expensive than even $1,000. But still,

on the high end, I would say of what I would probably want to spend on a soundbar. And I probably wouldn't, maybe I'd put this in the living room, but it might stick out. Like it's kind of like, yeah, you would do it as a display piece. Yes. You would need to fit your room specifically. Totally. It would work in the den, maybe the den, maybe not the living room is what I'm thinking. Yeah.

What are you thinking? We'd love to get your insight on this story or any story that we talk about. You can share it with us. Feedback at DailyTechNewsShow.com. Thank you, Jen. Always fun getting the chance to do this with you. Thanks to GT Tandon and Anand Jr. for contributing to today's show. And thank you for being along for Daily Tech News Show. The show is made possible by our patrons on Patreon.com slash DTNS. We'll see you next time. The DTNS family of podcasts.

Helping each other understand. Diamond Club hopes you have enjoyed this program. If you're a lineman in charge of keeping the lights on, Grainger understands that you go to great lengths and sometimes heights to ensure the power is always flowing, which is why you can count on Grainger for professional-grade products and next-day delivery so you have everything you need to get the job done. Call 1-800-GRAINGER, click grainger.com, or just stop by.

Ranger, for the ones who get it done. Each customer is unique. Every shopper is different. So why are your e-commerce search results still one size fits all? Generic experiences don't create loyal customers. They drive them away. Coveo's AI search delivers relevance at every step, anticipating what shoppers want even before they ask. The result? Better discovery, higher conversions, more profit. Visit coveo.com slash commerce to see how it works.