America Votes Against Data Centers - DTNS 5388

0:00Huyen Tue Dao: With the Spark Cash Plus card from Capital One, you earn unlimited 2 % cash back on every purchase. And you get big purchasing power, so your business can spend more and earn more. Stephen, Brandon, and Bruno, the business owners of SandCloud, reinvested their 2 % cash back to help build the company's retail presence. Capital One, what's in your wallet? Find out more at CapitalOne.com slash Spark Cash Plus. Terms apply. Hey, I'm Pablo Torre, host of Pablo Torre Finds Out, and I have chased down a lot of strange stories, but nothing quite like what's inside a McDonald's quarter pounder with cheese.

0:35Huyen Tue Dao: 100 % fresh beef, cooked right when you order it, hand-topped onions plus cheese and pickles in every bite, and right now, you can get one with medium fries and a drink for just$8. Only at McDonald's. For a limited time only, prices and participation may vary. Prices may be higher in Hawaii, Alaska, and California. And for delivery, fresh beef not available in Alaska and Hawaii. Offer not valid for delivery. This is the story of the one. As a shipping manager, he's running the dock, stocking supplies, and making sure every order goes out on time. And Grainger helps him stay prepared. From shipping supplies to PPE to hand tools, he counts on Grainger to keep products moving and trucks rolling.

1:20Huyen Tue Dao: So his operation runs smoothly no matter the shift. Call 1-800-GRAINGER, click grainger.com or just stop by Grainger for the ones who get it done. Do you snore? Did somebody hand you a CPAP and did you quietly stop using it? The mask that never quite seals, the hose tugging at you every time you roll over, a machine sitting abandoned in a closet. Excite OSA works a different way. It treats a root cause instead of managing the noise, and you use it while you're awake. 20 minutes a day, while the coffee brews, while you set your fantasy lineup. An FDA-cleared prescription device for primary snoring and mild obstructive sleep apnea.

2:06Huyen Tue Dao: Then you go to bed like a person, nothing strapped on, nothing plugged in. In clinical studies, 90 % of patients saw a reduction in snoring. No sleep study required. Prescribed online, shipped to your door. ExciteOSA at springsleep.com slash sleep20. Code sleep20 for 20 % off. Excite OSA is an FDA-cleared prescription-only device for primary snoring and mild obstructive sleep apnea. Not for everyone. Individual results vary. Risks and benefits at springsleep.com.

2:55Huyen Tue Dao: this is the daily tech news for wednesday november 4th 2026 we tell you what you need to know give you important context and really do help each other understand today data centers were on the ballot in the u.s and they seem to have lost the day have the data centers conceded they have conceited. I'm Tom Barrett. I'm Wint Wett Dow. Let's start with what you need to know with that big story. Yes, the U.S. held elections yesterday. You likely heard about it somewhere, even if you're not in the United States. But the part that's interesting to us on Daily Tech News Show was the 41 ballot measures across 10 different U.S.

3:35Huyen Tue Dao: states that asked the electorate questions about data centers. 28 of those measures proposed restrictions on them. 13 would have facilitated development. Mostly voters voted against the thing that would have been beneficial to the folks who want to build data centers. Let's take a look at a few of them. We're not going to go through all 41. Edgewater, Florida residents approved a measure 77 % to 23 % that put a ban on data centers into the city's charter. So this isn't just a piece of legislation. It's now part of the city charter. Can't put a data center in Edgewater. Folks in Perry, Ohio, by the way, Ohio had a lot of these folks in Perry, Ohio passed a measure that enacts financial and operational restraints on data center builders.

4:22Huyen Tue Dao: If you want to get a tax break from Perry, Ohio, you got to play by these particular rules. Citizens of Trenton, Ohio, voted to just outright ban large scale data centers. They defined them as I think 100 megawatts on up. The cities of Urbana and Hubbard voted by 77 and 76 percent respectively to ban data centers in their cities. And then there was Harwood, North Dakota, which voted against an amendment that would have mandated public hearings before large industrial projects. So that was one of the few that went in the favor of the people who want to build data centers. Of course, many candidates ran on policies related to data centers.

5:03Huyen Tue Dao: I'm going to focus on the governors because the governors are the ones that can actually make legislation happen versus senators who have a bit, you know, they have to work with the rest of the Senate. Both candidates for governor of Texas ran against building more data centers with governor-elect Hinojosa calling for an immediate moratorium until a legislative session can be issued or can be held on the issue. In other words, it's got to build any until we can pass some legislation to protect people from their effects. Texas accounts for nearly 30 % of the U.S. data center capacity set to come online in 2028.

5:37Huyen Tue Dao: So that could impede quite a bit of these. Ohio's governor-elect Amy Acton has called for an immediate conditional moratorium with a new series of requirements to be met before construction. And Georgia's governor-elect Keisha Lance Bottoms has called for a moratorium on new construction sites for data centers as well. Overall, 26 candidates in 21 states backed some sort of data center moratorium. And just to file this away in your brain, Axios called Colorado's Manny Rutanel and Minnesota's Matt Little AI safety advocates. So I know that's not data centers, but, you know, it's tangential to it.

6:15Huyen Tue Dao: And we might want to put away Little and Rutanel as people we may be hearing from on the AI issue. Of course, tech companies are now hiring more community liaisons and OpenAI is trying to use the name Compute Campuses instead of Data Center. Maybe they should have just gone with cute campuses, just shorten it. So you'll probably hear that bandied around as well. Uh, when I thought we would run through some of the things that might happen as a result of what seems to be a very clear public sentiment to slow down and restrict the construction of new data centers in the United States. Sounds good.

6:55Huyen Tue Dao: All right. Obviously the, the, the biggest one would be the competitors in other regions that don't have the restrictions may catch up to us companies. All of the makers of large language models and agentic AI and all of that say that compute is their biggest impediment. They have mostly solved the equipment problem and procuring the equipment. They have gone a long way to solving the power requirements. And so now the big thing in the way is political. So if you can't build the data centers and increase your compute as fast, you're giving folks in other regions a chance to catch up. I mean, how can I phrase my feelings on this?

7:38I mean, generally, that does make sense. Yeah. And the entire world is, whether we like it or whether you personally like it or not, investing in data centers. And certainly, I mean, obviously, the biggest competitor or the biggest overseas player, China, probably, you know, I don't actually personally know what their current situation looks like. But obviously, we think about folks overseas like that. I would ask you. Just real quick.

8:03Huyen Tue Dao: We mentioned yesterday on the show that China is proposing a plan for from development to worker protection to regulate the entire AI industry. Don't know exactly whether that's going to slow data centers down or not, but they are doing something. So, you know, we know. OK. All right. Sorry. I didn't watch yesterday's show. So, OK. So I suppose that this is an argument and this is an argument for. I mean, what is this an argument for? I don't know that it's an argument. It's just a fact. It's a trade-off. It's like, hey, if you think data centers have negative externalities, which a lot of people do, one of the trade-offs is your companies may not go as far.

8:46Huyen Tue Dao: Now, there are people who are like, I don't want these companies to succeed at all. So I don't care. That's on one extreme end. And other people closer to the middle may say like, well, they have enough of a lead. I'm sure they'll be fine. Let's be sensible about how we build these things. But it is an externality here. Another one would be the uncertainty over RAM and other component prices might continue longer. If we slow down data centers, that means we're stretching out the point at which the RAM crisis is going to continue to be needed. On the other hand, the capability to build new manufacturing capacity, as long as that's not slowed down, would come.

9:25Huyen Tue Dao: So it could actually have a benefit in that it slows down the purchases of RAM until the capability to build it can catch up. I mean, that would be like that part would be my assumption, I think, as a regular person who likes having their RAM. Thank you very much. And uses it quite frequently. Use mine every day, too. Yes, I use mine quite a bit. I would say that I suppose that for those of us that do care about RAM, that that would seem the obvious conclusion that I and maybe that's incorrect. and that maybe there's more nuance in the supply chain than we expect and of course that's why i phrased it as uncertainty yeah kind of just don't know right it could go lots of different ways but i mean again i i i generally feel like for the average human uh average person especially the average person in this country the per person benefit of ai businesses technology and data centers is uh contentious at word contentious at best or undetermined at best yeah um although and i would say that perhaps um the the the need that we have had whether you're you're realized it or not whether you're actively needing a ram or you're just a consumer of tech of of gadgets and technology um i don't know i i just can't help but feel like sure maybe we'll have more to spread around is my simplistic thought of thought about it but it would be interesting to see again, like if the demand does go down, and obviously we've had so many stories over the last long about manufacturers kind of refocusing most of their efforts towards data centers.

10:58I mean, and that's fair. If all of these companies have basically regeared their production and their entire business to data centers, would that actually have a negative impact? But at the same time, we're already feeling the squeeze. How much more can you squeeze out of us? Yeah.

11:13Huyen Tue Dao: I don't think the demand goes away. It just stretches out. stretches out okay so then maybe like delays and even more like like uh less uh less promised or rather uh under delivered shipments because another part of this is data centers start to get built elsewhere right they're just not built in the places that they would have been built otherwise could be other states in the u.s that have fewer restrictions could be outside of the united states um also the cost of ai services might stay high uh one of the things that has been happening is as companies add compute, they are able to lower the cost because they have the compute to meet the demand.

11:51Huyen Tue Dao: They don't have to use the cost as a restriction on the resource because it's a constrained resource. That could happen. Frontier model makers will work for more efficient models and therefore have a competitive advantage if you have a more efficient model. That could be some of the open weight models, but a lot of people consider the open weight models to be less efficient. So the frontier model makers might have an advantage because they also have open weight models. People forget about that. And if they are working to make them very efficient, I think that is definitely a thing we'll see is, well, if we can't get the data centers as fast as we wanted to, we're already seeing that.

12:30Huyen Tue Dao: Let's make them more efficient. So they make better use of the compute that we have. This is the thing about this entire AI boom. I won't call it a bubble yet uh maybe we should but um but it's probably a bubble but probably a bubble this is the one thing that has been fascinating and i i feel like i shouldn't be surprised but i am um or that i still feel that i shouldn't feel shocked but then i am in that generally speaking um the way that we've approached ai is just a more is better um and i know i've mentioned like token maxing and just companies kind of using using ai especially even when it was cheaper um kind of without any real, like most, like a lot of companies, I wouldn't say all, that's not really fair, but I don't have numbers, but it feels like the vibe was that, okay, we're going to kind of maximize our usage of this.

13:17We're just going to go kind of all in on this idea that if we, like, as fast as we can scale up AI, we're going to like more or less exponentially expand our productivity. And somehow the cost of everything wasn't discussed like first and foremost. and that it's so strange to me because I feel like in so many other facets of technology, even manufacturing and kind of commercial endeavors, is that efficiency and cost effectiveness is kind of just part and parcel, more or less. I mean, I suppose you could argue maybe like at the beginning, like kind of like in a study phase or kind of innovative phase that may be like less of concern.

13:54But it's so crazy for me for all these months to watch all of this kind of like more is more and this, you know, not the space race, but like the AI race to build these frontier models. And again, having like, you know, a relatively small pool of companies that can afford this much, like this much compute and encourage people to use compute as much as possible, racking up huge bills. It still is like, it's still astonishing to me that only just now are companies starting to kind of turn around and be like, maybe we should use this more efficiently and maybe we should determine the good places where the spend on compute has the most ROI and then understand where, you know, I guess the way we did things like not analog, but like pre-AI, AI analog.

14:44It's so astounding to me. And from this perspective, I can't help but see, like maybe if we squeeze the data centers, we can actually kind of enforce some of that, some of that introspection and kind of focusing on that efficiency. And if it makes frontier model makers more efficient, I can't help but feel like that would be at the very least that could be like the smallest good thing we can get out of this.

15:08Huyen Tue Dao: Yeah, I think there's two separate things that are easy to mix there. The token maxing side of it seems to be more of what you're you're talking about, which is companies saying just use more of it, which I always thought was ridiculous. And thankfully, we have seen a turn against that. Yeah, took a while. On the actual LLM operators and training and inferencing, it is just a fact that when you put more compute against them, they get better. All of those hallucinations and mistakes and the things that people complained about a year ago, those have been improved because they have been able to put more compute against it.

15:44Huyen Tue Dao: So I think I have less of a criticism of that part of it where they're saying, look, we will have the return on this investment because it will improve these models. these models and they will be more useful i think that makes more sense to me than the other part of it which was like let's just use them as much as possible we don't know why because there was a lot of that and i think i think you're right i think having uh the the focusing on more efficient models can probably have a better effect although we have we have seen companies start to turn against those as you've seen the llm companies start to price them appropriately and companies start to realize how much they're paying for those tokens.

16:25Huyen Tue Dao: So I wonder if this is really any kind of effect on that, but maybe it will be. Yeah. I don't disagree with you that the raw fact is more compute equals better models. I would still say that my intuition, and I don't have numbers, is that if there was a way that we could quantify the benefit of better models across all of the different uses across society I still doubt whether the ROI holistically which is again not for company not per person but holistically is good but that's kind of like my fluffy granola take on it granola I mean it's a question it's yeah yeah and whether you like your granola hard or fluffy doesn't matter it's still a very fair a very fair question to be answered absolutely I think that plays into another one of these effects companies maximizing inference on the edge like Apple does, local device inferences may benefit.

17:22Huyen Tue Dao: Again, because that's just more efficient. You don't need a data center for it if you can do it on your device. You mentioned the bubble earlier. I think the bubble might deflate faster or burst if there's pressure on companies to compete more and not the rush to like, we'll just get more compute and then we'll be fine. That starts to put put a little pressure on immediate returns. And maybe a bigger bubble would be prevented because of that. I also think big companies will have a benefit as this may make it harder for smaller, newer, more agile competitors to get going because they can't get into the data center.

17:59Huyen Tue Dao: So a lot of the NeoCloud stuff that people have talked about may go away. Now that could be a good thing too. A lot of this NeoCloud stuff felt a little fluffy, like some types of granola. I'm going to. So, but it, but it could have a dampening effect on that. And it certainly has already caused speculation that other ways of doing data centers, I think the one that got all the headlines was orbital data centers, but you know, offshore data centers that are ocean cooled and other ways of doing it will get more attention as well. It would be interesting. I have to admit, I feel ways about things when you say orbital data centers.

18:36Huyen Tue Dao: I would be more interested. Well, yeah, let's just call it what it is, because everyone thinks Elon Musk immediately when they think orbital data center. Then however you think about Elon Musk, for good or ill, is going to color that. But there are lots of people working on orbital data centers, including Google. And there's also, you know, people working on orbital data centers that have nothing to do with either one of them. So I think there's something there. It's a hard problem to solve, though. That's not a thing that's going to be solved and be like, well, we don't don't worry about the data centers anymore.

19:11Huyen Tue Dao: We'll just put them in space. It's going to be 10 years before that becomes a practical alternative. If it becomes a practical alternative. It would be interesting because, I mean, like it feels like a pipe dream or something out of like a near future sci fi. But at the same time, I have to admit, though, now that you're talking, I was like, well, that would solve a lot of problems that normal people have with it in regards to, you know, light light pollution, noise pollution. yeah water pollution the water usage and all that other stuff that is like a that are that are that are poor terrestrial mortal selves have to worry about um that is interesting uh and if we can take like kind of personal things out of it yeah that that would be interesting i just wonder can we wait 10 years if the pace i mean like i guess what it would be so fascinating to see what the story would look like if that becomes the future what the intervening 10 years or however long it takes to do that would look like but i i think it it ends up there's just a confluence if it works out it's just a confluence of, oh, we can now put, you know, we will have solved a lot of our feelings about data centers by now, because there's a lot of emotion mixed in with a lot of legitimate concern in there.

20:15Huyen Tue Dao: And emotion usually burns off faster. And then we get boiled down to the legitimate concern, which I didn't put it down in our notes, but I think that's another effect that this will have is encouraging people in their communities to speak up and go to meetings and ask their local elected representatives who are in situations where these are being put in, like, are you taking the proper safeguards? Are you holding them to account? Are they using evaporative cooling? Please don't. Use a closed system. That's solved. Let's not use our water for this uh are they getting their power have they locked in how they're going to get her power so that it doesn't affect our stuff uh i i've worked with one listener to kind of help her arm herself on like these are the things they need to do to do it responsibly and she actually got a data center stopped because the person who wanted to bring it in was not wanting to do it responsibly they were wanting to do it on the cheap and i think that's the most positive aspect of all of this is the laws that are saying like no let's let's have hearings let's get the community involved If we're going to do them, let's do them right.

21:24Huyen Tue Dao: That's what Governor-elect Hina Hossa is talking about. That's what Governor-elect Acton is talking about. Let's not have them, but let's figure out how to have them so that we are limiting the negative externalities and making the companies who want to profit from them pay for those negative externalities. I think that's a fantastic, positive note that obviously, you know, all of us that have concerns, some people did take the time to speak and definitely some representatives in government have listened. So, yeah, it was always worth speaking up. You know, I guess one could say that for for governing, for people who are in politics, politicians and people who work in government, it is their privilege and responsibility to listen to their constituents.

22:05And for us, it is just a privilege to be listened to by you because ZT Nest is made possible by you, the listener. So let's take time to thank Joni, Hitek Oki, Chris Zaragoza, and Judge JLo.

22:17Huyen Tue Dao: Yay! Thank you. With ChatGPT, you can do more than get answers. You can get things done. Whether you're leading a big project, planning an elaborate trip, or trying to balance work and life, ChatGPT can help. Give it a goal, plus things like files, notes, or project plans, and it can help you build a templated deck, develop a trip itinerary with daily updates, or create a prioritized schedule so you can spend less energy sorting out the details and more on bringing your biggest ideas to life. Get started at chatgpt.com. Being a serious business owner takes hard work, patience, and a drive to keep moving forward.

23:01Huyen Tue Dao: So it's important to find a serious business card that goes the extra mile. With the VentureX Business Card from Capital One, you earn unlimited double miles on every purchase. And with big purchasing power, you can spend more and earn more. This is your sign to take your business to the next level. The Capital One VentureX Business Card. What's in your wallet? Terms apply. See CapitalOne.com for details. Cash Now, More Later from Opendoor gives you cash up front for your home, plus all the profit later. That's no chaos now, no cash left behind later. Skip the showings now, pocket extra profit later.

23:36This is so simple now, this is so awesome later. Or sell fast now and pop the champagne later. Cash Now, More Later, now available nationwide.

23:46Huyen Tue Dao: Start your offer at opendoor.com slash ACAST. Profits calculated after fees and costs, eligibility and offer price may vary. Grainger knows if you're a procurement manager at an industrial manufacturing facility, you're balancing inventory and budget while sourcing critical supplies across the operation. Too much inventory? Costs climb. Too little? Production slows. Grainger's$2 billion in on-hand inventory helps you get what you need when you need it, so you can help keep budgets balanced and lines moving. Call 1-800-GRAINGER, click grainger.com, or just stop by. Grainger, for the ones who get it done.

24:23Huyen Tue Dao: Do you snore? Did somebody hand you a CPAP and did you quietly stop using it? The mask that never quite seals? The hose tugging at you every time you roll over? A machine sitting abandoned in a closet? Excite OSA works a different way. It treats a root cause instead of managing the noise, and you use it while you're awake. 20 minutes a day, while the coffee brews, while you set your fantasy lineup. an FDA-cleared prescription device for primary snoring and mild obstructive sleep apnea. Then you go to bed like a person. Nothing strapped on, nothing plugged in. In clinical studies, 90 % of patients saw a reduction in snoring.

25:07Huyen Tue Dao: No sleep study required. Prescribed online, shipped to your door. ExciteOSA at springsleep.com slash sleep20. Code sleep20 for 20 % off. Excite OSA is an FDA-cleared prescription-only device for primary snoring and mild obstructive sleep apnea. Not for everyone. Individual results vary. Risks and benefits at springsleep.com.

25:38Huyen Tue Dao: There's more we need to know today. Let's get to the briefs. Okay, well, days since Tom asked an Android developer to talk about iOS is now zero. because we're going to talk about the latest betas of iOS 27.3, iPadOS 27.3, and MacOS 27.3, and as well as VisionOS 27.3s. Developers may now bring any content or action from their app into Siri AI. Previously, apps could only access Siri AI if they met certain Apple-defined categories. Apple also added a skill that can teach Siri AI to combine several actions into one request. So people can just ask for help in their own words, according to what Apple wrote.

26:21Users won't have to name the app if it's on a screen, showing a live activity or already the subject of the conversation. Again, kind of more contextual. This feature is in preview and won't reach end users for a few months. And of course, you know, if you're in the EU or mainland mainland China, rather. It will take longer since Siri AI hasn't launched there. But possibly shedding some light on Apple's vague reference to new revenue in its earnings call this week. Apple intelligence features that run on Apple servers will face daily usage limits. Apple wrote that, quote, expanded access to such features will be available for a fee in the future.

27:01Huyen Tue Dao: Yeah, that's my best guess for what they were alluding to. Future fees. Yeah, they're going to start charging the biggest users. That's usually how Apple does things. So, you know, it'll probably be part of your iCloud subscription or something like that. But one of the reasons I wanted to make an Android developer talk about this is because I thought it was really interesting to see Apple just go straight to like, yeah, it's open. This app is now open. You don't need a particular schema just to access it. I mean, that's got to, I feel like that sounds like a good thing. Yeah, and it's a little shocking because as much as one would like to complain about it, and I'm sure my iOS brethren probably are less have a less good time about it.

27:46The kind of stringent control that Apple kind of maintains over iOS and iOS ecosystems. This is kind of shocking. And it feels just for one thing, actually, I went to the Android Dev Summit last week. That's why I wasn't with y 'all. And they had announced, it's been around, but a similar feature called connected apps where you can bring app functionality into Gemini and Search. And similarly, there's just an API called app functions that developers can open up functionality. Like it's just there. And so I can't help but think that part of what iOS is doing or Apple is doing rather is that, again, maybe from the perception, of course, that they are still behind, that it only makes sense to, if anything, have parity with Android.

28:31I'm not saying they looked at Android and said, oh, we should do this thing too. But the current state is that, especially with app developers in particular, that we're very aware, everyone is super interested in how they can tap into what is kind of becoming, both with Siri and with Gemini, the intended central way that you are interacting with your phones. and this is like a concept we see all the time right like there's there's tons of companies out there working on a kind of app you know sort of getting towards that appless world you just interact with a singular thing and it does all the things regardless of whether you're actually ordering from doordash or doing something else and so it just feels like okay this is the state of things um and so while it's kind of surprising that they are letting control it it also kind of makes sense i think with with with the the the tone of the room and where like these different platforms are going.

29:25Yeah.

29:26Huyen Tue Dao: It is the kind of thing that makes Siri AI more useful. Yes. A hundred percent. Yeah. So, uh, one of the, one of the examples, uh, was like, you've got the airline app up on your screen and you say to the app, uh, can you mail my flight information to Eileen? Yeah. And it knows, oh, you're talking about the airline app. You're talking about the flight information that's on screen. I'm going to extract that. I'm going to give it to Siri AI. Siri AI is going to look up Eileen. It knows, oh, that's one of your top five contacts. It's obviously your wife. I'm going to take that information. I'm going to message it to, or I'm going to mail it to, or whatever it was I said, it will remember.

30:03Huyen Tue Dao: And I won't even remember. So, you know, that's why they want it to be this way so that, you know, you don't run into like, oh, I can't do that right now. Yeah. So yeah, I think that's, I think that's going to make Siri AI a little more useful. That was a big problem before we had LLMs when, again, sorry, Android developer, but when we had Google Assistant and just the ways that, you know, there was always the idea that, oh, I can talk to my assistant and I can do things that are not Google related in it. And before it was, without LLMs, which is a lot harder and I'll give LLMs and kind of like more conversational interfaces a lot of credit, is that it was very kind of rote, Like you had to like pick like your key phrase kind of like, you know, hey, G, hey, A, whoever.

30:52And they were very rote. And that the fact that we didn't have this more smooth natural language like interface kind of just made those features kind of die in the way. And also it made it hard for us to invest in these things. And it's like one of those like chicken and egg things, right? Sort of ish, where the value of these kind of centralized assistants like Gemini, like Siri comes from the ability, like comes from their capabilities. And the incentive to produce these capabilities and make them available to the assistants is kind of also predicated on how much they're being used. So if I'm an app, if I know everyone's using Gemini or Siri, then of course I'll have to do that.

31:29So it's a really interesting time. And I think it's almost like someone has it's like a weird game of chicken where someone has to kind of give and say, here, you can have the thing so we can all get to the same place. So I guess in a sense, Apple's like, OK, fine, you guys do it. Do the thing. And then everybody will. I don't know. I love the chicken analogy. The game of chicken didn't quite work out, but maybe this will.

31:48Huyen Tue Dao: It's the chicken and the egg and a game of chicken with whichever one it is. With an egg and a chicken in a car. Anyway. Something's getting scrambled. I'm just saying. Also in the iOS 27.3 beta is Selfie Sing. This lets you use the Sing karaoke feature that's been in Apple Music for a while. That one removes vocals and shows you lyrics so you can do karaoke with any song but now you can record a video of yourself singing now it doesn't support the whole catalog right now it's got a limited playlist includes songs like since you've been gone by kelly clarkson what makes you beautiful by one direction like a virgin by madonna baby one more time britney spears nine to five dolly parton uh etc they tried to pick the big karaoke hits for this uh and then you can add effects to your voice after the fact, megaphone, echo, concert hall, studio, before you share it with your friends.

32:44Huyen Tue Dao: As two people who have publicly done karaoke in front of other people this year, I feel like we had to comment at least a small amount on this. I don't know that I would ever record this and send it to any of my friends. What about you? No, because while nominally we were recorded doing our karaoke eight maybe like only a handful of people seen it and everybody just remembers very fondly that you got up and did a thing now there's evidence and unless there's the best auto tune ever i don't know maybe i i know uh no i guess if you're having a party with friends you could you might be recording it so that then you could share it later like oh yours was so good can you share it with me i know that there's a whole population of people that can sing very very well and that this will be absolutely fantastic and fun and so i'm very very happy for them do not bring it anywhere near me that i'm saying that for you for y 'all like that's better for y 'all if it doesn't come anywhere near me yeah show me yours but i'm not showing you all right well if you want to show us any feedback that you that you have on the show you can certainly get in touch with us.

34:00I mean, if you want to do a selfie sing, you could probably send us that. And you can find us on the socials at DCN Show on X, Instagram, Threads, Blue Sky and Mastodon. And for TikTok and YouTube, you can find us at Daily Tech News Show.

34:18Huyen Tue Dao: Being a serious business owner takes hard work, patience and a drive to keep moving forward. So it's important to find a serious business card that goes the extra mile. With the VentureX Business Card from Capital One, you earn unlimited double miles on every purchase. And with big purchasing power, you can spend more and earn more. This is your sign to take your business to the next level. The Capital One VentureX Business Card. What's in your wallet? Terms apply. See CapitalOne.com for details. Running a small business means wearing a lot of hats. Between managing employees, handling payroll, and keeping up with paperwork, there's a lot to juggle.

34:54Huyen Tue Dao: Luckily, Gusto makes it easier. Gusto is online payroll and benefits software built for small businesses. It's all-in-one, remote-friendly, and incredibly easy to use. So you can pay, hire, onboard, and support your team from anywhere. From automatic payroll tax filing to simple direct deposits and benefits options for nearly every budget, Gusto helps simplify the process. Plus, enjoy unlimited payroll runs for one monthly price with no hidden fees or surprises. And with automated tools for onboarding and offer letters, you can focus on what matters most. Try Gusto today at gusto.com slash tech and get three months free when you run your first payroll.

35:31Huyen Tue Dao: That's three months of free payroll at gusto.com slash tech. One more time, gusto.com slash tech.

35:41Huyen Tue Dao: And now the quick headlines that will definitely make you smarter. All right. Well, the review embargo for The Legend of Zelda Ocarina of Time remake has been lifted. The original is the best reviewed game in history with a 99 on Metacritic. The remake got close with a 96 based on 80 reviews. That makes it the best video game review of the year on Metacritic, though we know GTA 6 isn't out yet. So we'll see how it holds up. Yeah, still pretty. It gets in the top 10. Pretty good. All time. So it ain't bad. uh macronix one of the longtime makers of roms for nintendo cartridges announced it's getting out of the rom business uh it's going to shift to making nor or flash and emmc memory chips which apparently these days make up 95 of its business there are other people who make rom for nintendo but uh you know it's just not as profitable as it used to be well there we go well thanks to lurker b for noting this on the subreddit attacks on connected refrigeration have expanded from grocery stores to refrigeration systems used by French blood banks.

36:51The French Blood Establishment, or EFS, was able to regain control or take units offline, and no samples or blood products were affected.

36:59Huyen Tue Dao: That makes this a good story. It's a good story. They went after the grocery stores who weren't prepared, and they spoiled a bunch of meat. They went after the blood folks who are like, no, this is important stuff, who were prepared, and they did not spoil it. Same plan with the blood. Thank you. Binance says it has normalized outflows after a theft of$6.3 billion worth of cryptocurrency and that the attackers who conducted the theft appeared to have started working on it in June. This was not a one-time like we fished it out of one laptop. Binance says its investigation indicates the attackers used a range of techniques, including social engineering by fake recruitment interviews.

37:38Huyen Tue Dao: They would go after Binance employees, do an interview, You get them to install software during the interview that would then be used to gather intel to improve their ability to do the attack. There were, of course, malicious software downloads, compromised internal AI tooling, endpoint compromises, and an advanced transaction authorization method. So this was very sophisticated. Wow. Wow. Okay. Sorry, as I get over my shock on all that, Tencent released a new foundation model called Hi4. Kind of wish it was called Hyrule, given our earlier story. So Hy4 is comparable to recent releases from Zy, Moonshot, and DeepSeek, but still behind the major foundational models from Anthropic, OpenAI, and Google.

38:23Huyen Tue Dao: You know, it's so odd when, since the frontier model makers put their frontier models in a restricted area where these companies can't access them and distill them, their models have not been improving as fast. It's just a weird coincidence. Oh, have they checked behind the TVs at all? Yeah, and since they've removed all the things behind the TVs, their models haven't improved as well. Under pressure from the United States government, Oracle canceled a contract with Tencent to provide data center access in Southeast Asia. All right, well, Alibaba, Tencent, and ByteDance are expected to receive around a whopping 1 % of the NVIDIA RTX Pro 5500 chips they ordered.

39:06as NVIDIA has difficulty meeting demand. The three companies have ordered a combined 6 million chips and over the next four months will receive a few tens of thousands.

39:18Huyen Tue Dao: Yeah, so can't get them from Oracle outside of the country. Can't get them into the country. What if we just go with Huawei, which is inside China? Well, Huawei is telling customers that they may need to source their own RAM if they wish to use the next flagship AI chip from Huawei. Chinese memory makers cannot produce enough high bandwidth memory themselves. And Chinese companies are restricted from buying from SK Hynix and Samsung. The Next Web describes some very elaborate smuggling techniques to get the RAM into the country because the import restrictions are based on processors. So they'll buy crap processors with good RAM, carry them in luggage across the border, strip out the RAM and then use them.

40:03Okay. waymo co-chief executive officer takidra mawakana said the company will end its partnership with uber in austin and atlanta in january 2028 and operate the hailing service for waymo cars itself

40:19Huyen Tue Dao: as it does in other markets yeah so waymo waymo wants to compete with uber not be a partner and in the category of names to be aware of uh file away india's elo elo story tv it's spelled elo elo like electric light orchestra twice elo elo story tv is a micro drama app in india and the news today is that it's got a bunch of funding including from nbc universal but it also has 33 million active users and growing uh micro dramas are on the rise amazing don't get me down okay Right, how does it? Mr. Blue Sky. Yes, Mr. Blue Sky. While we end every episode of DTNS with some shared perspectives, Joe works in IT at a college and has some thoughts on the Google policies on AI and search for schools.

41:08Huyen Tue Dao: Yes, thank you, Joe. Joe wrote, hey guys, the output of AI is obviously critically important, but only one side of the equation. I'd be curious to see what inputs from a normal Google search go into Gemini. This is, by the way, in relation to the story yesterday that Google is turning off all Gemini access for schools by default. And then the schools can turn it back on rather than having it on by default and making the schools turn it off if they don't want it. So he's saying, I'd be curious what goes in terms, locations, tech profile, search history. There could be a concern from colleges or schools in general of what students would be searching for.

41:44Huyen Tue Dao: And if those searches would end up in Gemini for the general public, this could be a concern from a data privacy standpoint. and I'm not familiar with the laws governing grade school, but in higher education, there are laws like FERPA, which regulate how data is shared between systems. And it's something that we have to evaluate when purchasing or bringing on any new software. Another aspect of this could be from the procurement standpoint. I believe many states are enacting procedures where any use of AI has to be fully documented during the purchasing process. This would include any pieces of software that are using AI to deliver a service.

42:15Huyen Tue Dao: From the standpoint of someone in education, having AI completely decoupled would make the purchasing decision a lot easier and a lot smoother. Not sure how big of an issue any of this is, but thought I would share a different perspective. Keep up the great work. Thank you, Joe. Wow, yeah. And just, I'm pretty sure Joe knows this, but for anybody listening, Google does not use search history from educational Google accounts to train its generative models. And it has additional data protections for Gemini and Google Workspace for education, but it doesn't have those on personal Google accounts.

42:47Huyen Tue Dao: Search history can be used to improve models depending on your settings. You have to go in and say, don't do that. And that doesn't eliminate the need for schools to evaluate what information is being shared between services when FERPA is involved, like Joe was saying. So it also doesn't apply to students using personal accounts for their own equipment. So thank you, Joe, for sharing your insight or knowledge on that. That's very helpful. And I'm still, it's funny, on our subreddit, a lot of people are like, well, the school can just put a Chrome, like I suggested yesterday. They can put something in that restricts the AI from showing up for students, which they could.

43:25Huyen Tue Dao: But I think a lot of administrations are like, but why should we have to do that? Shouldn't you give us an option that's AI free? Yeah, absolutely. Well, thank you so much, Joe. What are you thinking about? Do you have something that will help us better understand or just have some other insight into a story? We would love to hear it. Feedback at DailyTechNewsShow.com. Thanks to Joe for contributing to today's show. Thank you for being along for Daily Tech News Show. Didn't have a new patron today, so it could be you tomorrow. Become a patron right now. Patreon.com slash DTNS.

44:01Huyen Tue Dao: The DTNS family of podcasts. Helping each other understand. Diamond Club hopes you have enjoyed this program.

44:13Huyen Tue Dao: with the spark cash plus card from capital one you earn unlimited two percent cash back on every purchase and you get big purchasing power so your business can spend more and earn more steven brandon and bruno the business owners of sandcloud reinvested their two percent cash back to help build the company's retail presence capital one what's in your wallet find out more at CapitalOne.com slash Spark Cash Plus. Terms apply. The quarter pounder with cheese is made with 100 % fresh beef cooked when you order it. Then comes the melted cheese, the slivered onions, the pickles, the ketchup, and the mustard.

44:51Huyen Tue Dao: Add hot crispy fries and an ice cold Coke, all for$8. And if things go well, one loose fry at the bottom of the bag. The quarter pounder with cheese meal. Now$8.