The World’s Smallest Autonomous Robot - DTNS 5192

0:01This is the Daily Tech News for Monday, January 26th, 2026. We tell you what you need to know, give you some important context, and help each other understand. Today, Apple released a new AirTag, but it's not as small as the world's smallest autonomous robot. Oh, yeah, I'm excited about that robot. I'm Tom Merritt. And I'm Rob Dunlouin. Let's start with what you need to know, however, with the big story. Apple releases a product in late January, and it fills the news hole. So the big story, if you look anywhere on the internet today, is new AirTags. It's an update on the Apple AirTag, which was first released five years ago.

0:40First update in five years. It doesn't get a new name. It's not called AirTag 2 or anything, just AirTag. But it does get some upgraded internals. So new ultra-wideband chip that will give you longer range. If you're unfamiliar, ultra-wideband is also a wireless signal like Bluetooth or Wi-Fi. But it's better at directional location because it can send more pulses, measures time of flight better. doesn't use as much power. What you really need to know is that it's better at that precise location stuff and it doesn't use a lot of power. There's also an upgraded Bluetooth chip in the new AirTag though.

1:15So that also, as you might've guessed, gives it a longer range. That added range from both UWB and Bluetooth means Apple claims you can use precision finding. That's its feature where things beep and rumble to guide you to the lost item. You can use that from up to 50 % farther away. Precision finding can now also be used on your Apple Watch Series 9 or Apple Watch Ultra 2 or later. So you can use that to find the AirTag device. AirTag speaker is now 50 % louder. So when it's beeping at you, you should be able to hear it better. And Apple said it improved privacy tracking, but it didn't give us any details on that using some security through obscurity, I guess.

1:57It did remind everyone that these are not meant to track people or pets. Now, the pets part is important because the privacy tracking feature could accidentally trigger if you've got it on a pet. You know, someone else is walking your pet and then they get a notice like there is an AirTag near you, stuff like that. It's just not designed for that. It's$29. Same price as the old AirTag. $99 for a four pack. Same price. You can order today. It'll ship later this week. And it looks the same. So that means it doesn't get a notch for a key ring or anything. But if you've got an old AirTag holder, it should work with this.

2:30And if you're buying a new AirTag and you want to put it on a key ring, you're going to have to buy something to hold it. So I am all for companies like Apple coming out with new stuff and not charging you more than they charge you for the old stuff. Yeah, man, I'm with you. The only issue here is that it looks the same. So my concern would not necessarily be directly from Apple, but eventually when these things hit Amazon and Best Buy and this and that and the other, I want to make sure that I'm buying the new Apple AirTag as compared to the old Apple AirTag. because unless it doesn't tell you, there's no new name.

3:04So I want to just make sure that it says 2026 somewhere in the box so that I'm spending the$29 to get the new hotness and not to get the old hotness that literally looks the same and costs the same. That is a great point because old AirTags will be in the supply chain somewhere, you know, for the next year or so. Eventually, they'll go away and it's not as big of a deal unless you're buying something used. But even then, if you're buying something used AirTag, Although buying a used AirTag is a chancy proposition. You might not want to do that anyway. But yeah, if you're buying used devices, you got to watch out for this.

3:34Otherwise, yeah, I'm all for like, let's just keep it simple. Call it an AirTag. Make it the same price. Just give me a better version of it. And I'm into that. Yeah, we can imagine if Microsoft made these things. Or even if Google was making them directly, it would be the AirTag 7-9 YFAC Special Plus Edition or something like that. So I do appreciate that. Like I said, my only concern is that I just want to make sure that when I'm buying the new hotness, I'm actually getting the new hotness and not the old hotness. Yeah, look for that part of the description and make sure that you got the new one.

4:08DTNest is made possible by you, the listener. Thanks to Michelle Sergiu, Kirk Stephenson, Miranda Janelle, and our new patrons, John and Will, and returning patrons, Cress and Jesse. Whoa! Monday is when we always welcome the new patrons and the returning patrons and everything. So good to have John. Good to have Will. Welcome. Everybody welcome them in. Say hi on Patreon in the chat or something or in the Discord. And welcome back, Chris and Jesse, as well.

4:36All right, there's more we need to know today. Let's get to the briefs. Wired reports on a study from scientists at the University of Pennsylvania and the University of Michigan published in the journal Science Robotics, describing an aquatic robot 0.3 millimeters on its longest side. It's considered the smallest autonomous robot yet created. It moves by creating an electric field and pushing charged particles into the liquid around it, which causes drag in the water molecules. Wire describes it as if the robot itself is not moving, but the ocean or river is moving. It can move a distance of its own body length in a second and can change direction to follow complex paths or move in groups like a school of fish.

5:13Because it has no moving parts, it can swim for months at a time. Solar panels generate 75 nanowatts of power and a special circuit can operate at extremely low voltage. The solar panels take up a lot of space, leaving little room for circuits, so the software was condensed into a special instruction that fits into a small memory space available on the device. It also has a sensor for measuring temperature. It communicates like a honeybee through movement, which can be observed with the microscope and be interpreted. Different robots can have different instructions in order to perform separate roles in more complex tasks.

5:46Applications could include monitoring individual cells or assembling tiny devices. production cost is estimated at one cent per unit that's actually not terribly expensive one cent per unit no that's incredible units at a time yeah yeah i mean obviously these things are the size of a grain of salt literally like they show it fitting inside the zero on the date on a small coin uh so so they're they're incredibly small uh you're gonna want a bunch of them but hey for a buck you can get a hundred of them so uh you can you can start doing a lot of stuff with them and they're practical. They actually work.

6:21You can control them. You can program them. It's a very simple program, obviously, but there's a lot. This is the kind of robotics that is probably less flashy because it doesn't have any moving parts. It doesn't have pincers or anything, but you'll be able to use it to push things around and monitor things. And I thought the idea that it does a little honeybee dance to communicate, and you do have to observe that optically with the microscope but i thought it was fascinating just the the ingenuity and getting around the problems at that size we we have been talking about and you know observing nanites in movies for decades now it's been it's been since the you know 50s and 60s people are shrinking stuff down and injecting into your bloodstream and this isn't there yet but it's getting there it's getting to where you know we're going to be able to do some really effective things um to very very small things in liquid with things that you know might be called nanites at some point yeah yeah I mean this this is the first time anyone has been able to make an autonomous device now granted again it doesn't have moving parts but it still can move and it can follow instructions and do things and sense things uh first time you've made it at this size um you know I I don't know if we're we're bumping up against the limit of what you can do at this size because they they had to really work on the programming to bring it down to one instruction that will be able to have complex behaviors.

7:51But man, every time you say, I bet we're at the limit, somebody figures out how to get past the limit, as seen by this, it's been 40 years that people have been trying to bring this down below a millimeter, and these folks finally did it. Finally did it. If you listened to our discussion with Andrew Main on the weekend about AI accelerator chips, today's chip news is going to make a little more sense to you. So go back and listen to that if you didn't. It was a really good conversation about Cerebris and Grok with a Q, the chip maker Grok, about, you know, why are they doing this? Why are they more successful?

8:25Why do they help? And it puts in a different light this announcement today that Microsoft has announced its latest in-house chip for inferencing. Inferencing, of course, when you're running the model, answering queries, doing things like that. the Maya 200 is Microsoft's new in-house chip. Microsoft says it should work faster and at lower power. But when you compare that to things that Cerebris and Grok are doing and a few others, it's interesting to see how much it is not exactly at the forefront of that. However, all these companies are wanting to get control of some of the chip making for their in-house stuff.

9:03So this is an example of Microsoft continuing to proceed at that. I also have a bunch of other news here. NVIDIA invested another$2 billion into CoreWeave. That will help CoreWeave purchase land and power for data centers. They think up to five gigawatts they're going to go for. CoreWeave will also implement NVIDIA's Vera CPU in its data centers. Not a GPU, CPU. That's a move by NVIDIA into the CPU space, which is currently dominated by Intel and AMD and data centers. The next one less about accelerators is Reuters reporting that Samsung is going to start production of its next generation high bandwidth memory chips.

9:38High bandwidth memory is all what these data centers need. Samsung's HBM4 will start production next month to supply both NVIDIA and AMD. This gives Samsung a chance to catch up to SK Hynix, which has been dominating in the high bandwidth memory space. And the chip acquisition fad has reached quantum computing as well. IonQ is a company that makes trapped ion quantum computers. They offer greater accuracy. They have some contracts with the U.S. government. They have agreed to acquire U.S.-based Skywater technology, which is a chip foundry. They make semiconductors for quantum computing, aerospace, and defense.

10:19IonQ intends to make a, quote, vertically integrated quantum platform business. They want people to think of them as the NVIDIA of quantum computers. We'll see if anybody ever does. So what's interesting about all of this news is that you're seeing the shift away from companies saying we have to have the fastest, most high end processor that man can make to we need to have processors that man can make. And maybe they don't necessarily need to be the fastest because we need to control our own supply chain for our own data centers. And that is that is really a difference. If you were to go back four or five years ago, they didn't necessarily look like that.

10:55It was, can you get me the fastest thing that you can make quickly? Now it is, what can you make? Okay, let's make that. It's not the fastest. Okay, we're good with that. But I can get a data center full of them. Yes. Okay, we're good. And that is a market difference in the market, just in a relatively short amount of time. And it's not like they don't also want them to be fast. You're not saying that. It's that they are balancing the supply and the cost and the vertical integration in there. And it's not just about the chip being fast. like cerebris's chips aren't necessarily the fastest but they are fast at inferencing or they are better and more efficient at training power efficiency is almost important as speed when you're talking about these chips so so yeah your point is well taken there's it is much more complex than just looking at the gigahertz and going like that's the best one i mean it always was but even more so now than ever the fastest car in the world doesn't drive a family across cross country very effectively at speed.

11:53So, so I think that that's how they're looking at this. We need speed, but we needed to do what we needed to do to get the job done. Are you going to be doing a lot of windy mountain driving? Are you going to be, you know, flat on the ocean, different car, different purposes. The register reports on French artist, Theo Champion, AKA RootKid's new piece of art called Spectrum Slit, coded in Python and running on a Raspberry Pi. It translates radio signals in the 2.4 gigahertz and 5 gigahertz bands into lights on an array of 64 led filaments a hacker rf one software defined radio samples the signals and the program running on the raspberry pi converts them into signals that control the lights the printed circuit boards and lights made up the bulk of the cost at around one thousand dollars now unlike the robot earlier uh this is not necessarily practical this is just interesting uh and i i hope that uh root kid is able to get this installed somewhere where people can look at it.

12:50He's talking about he might be willing to make these for people if they want. He's got some other ideas for other pieces of technological art that he wants to do. But I am fascinated that this sort of like is quiet during the day when no one's around in his apartment and then lights up as soon as everybody comes home and they have their phones on and their laptops on. Because again, it's not taking the data and it's just noticing that there's activity in these bands and then reflecting that in the lights pretty cool and it probably looks wicked awesome i mean it probably looks really really cool uh just as you know to see it because because it is art so i would not be uh i would not be ashamed to take a look at one of these myself you know i'm always into like cool abstract type art type stuff so this sounds pretty cool morgan stanley surveyed businesses to ask them whether they have cut jobs uh or reduced hiring because of new generative tools and what increases in productivity they have seen.

13:49The report surveyed firms that have been using tools for at least a year. And this is, I think, important to understanding this, in five industries, consumer staples in retail, real estate, transport, healthcare equipment, and automobiles. So we're not talking tech companies. We're not talking AI companies that have a vested interest in saying, wow, these tools are great. We're not even talking about tech companies like Netflix or Google. We are talking about companies that have a non-technological oriented business. Were they able to see productivity increases and that they decide to either eliminate jobs or not hire more people because of those productivity increases?

14:29Across Australia, Germany, Japan, the UK and the US, productivity increases were reported at at least 10 % and close to 14 % in Australia. So they are seeing productivity increase with the implementation of these tools. Most of these nations had job losses, net job losses. UK companies reported the biggest effect, an 8 % net loss. They reported that they were adding jobs because of these tools, but also eliminating jobs. And so the net was 8 % reduction in the UK. So for example, they did not fill 23 % of jobs while adding 15%. So you get a net loss of 8%. Japan reported a net loss of 7%. Germany and Australia, 4%.

15:14The US reported a net addition of 2%. So in the US, they cut 17%, but added 19 % because of, again, according to the survey, because of these tools. Now, granted, this is self-reporting. So there's a lot of room for error and bias in here. But I thought it was interesting to see even when all other things are equal on the self-reporting side, US companies seeming to take more advantage of like, ah, this tool, let us do something we couldn't before. Yeah. So the interesting thing here is that, as you said, these are not necessarily companies that we think of as tech companies, but the truth of it is that this is where most people in tech actually work.

15:57Most people don't work for Google or Microsoft or Apple. They work for the department store, the hospital, the insurance company that you don't think about it being a tech company and they are implementing these tools. I also think that the United States has figured out very quickly, it's like, it's not necessarily in our interest to just say we're laying off a ton of people because we're putting in productivity tools. That doesn't really help us. It's, you know, if anything, it might hurt us and maybe get us regulated in a way that we aren't, you know, right now. So I think that you might see some of that start to appear elsewhere as well, because telling people they're losing their job because AI can do it is not a story that people want to hear.

16:40And, you know, when those things start to happen, you start to get politicians involved and that's not necessarily where these companies want to be. So I think that that's part of it. But the other part of it is that, and folks have been saying this for a while, that, you know, we're not necessarily talking about only AI, but AI is a big part of this. AI isn't really replacing people's jobs. It's the people who can use the AI tools who are replacing the people who can't. So is there a link? Absolutely. And will you see companies like an Amazon where Andy Jassy comes out and says, yes, we are letting these people off because AI is replacing them.

17:13You will get a company here and there that will say that. But for the most part, that's not necessarily what's happening. And this is self-reporting, but even though the numbers are down, it doesn't look like it is as bad as what we once thought it might be. Yeah. And this survey also noticed that a lot of the job reductions tended to be in entry level. So automating things, a lot of times you have entry level, somebody does the tedious job first, right? And so the automation gets rid of the tedious job, you don't need the entry level person. And so we're going to have to develop a new way of bringing new people on board.

17:47And what do you give the new people to get them up to speed and gain experience? I also think the U.S. might be adding jobs because there has been more exposure to these tools since ChatGPT started here. There's been more time for people to get used to them. I'd be curious to see if Morgan Stanley were able to survey Chinese companies. Would you see a net addition there? Totally different economic landscape there, I know, but that's the closest equivalent to a country that has had a lot of close experience with these tools to the United States. I hope and I imagine that you will see these other companies like figure out like, oh, we don't need this.

18:30But then realize if you're a smart company, hey, we can get a jump on the competition if we hire people to use these tools, like you're saying, Rob, to do things that we couldn't do before. And that will give us an advantage. And then you start to see addition. And that's what we seem to be seeing in the United States right now. Yeah. And this is a hope of me personally of what I think happens, you know, coming in the future here is that the companies are saying we don't necessarily need to lay people off. We need to just retrain people to use the tools to get more productivity so we can just do more stuff with the same number of people or maybe even slightly increase.

19:06But I don't think the companies just want to say, hey, we're going to replace you with a bot. That's not something that they want to do because that doesn't work out for them in a long term future. Yeah. Well, the smart companies won't do that. The companies that want to stick around won't do that. Some of them will try and maybe they'll last for a while. What do you want to hear us talk about on the show, folks? There's lots of ways to let us know. One way is our subreddit. Thanks to Motang and R.W. Nash for submitting stories that we're talking about in today's show. Always good to see what's happening on the subreddit.

19:36You can find it at reddit.com slash r slash daily tech news show.

19:43finally some quick headlines that are good to know might make you look smart in the future if you know this stuff tic tac usd that's the brand new u.s version of tic tac has said that a power outage at one of its u.s data centers caused issues with tic tac and other services in the u.s over the weekend including delays in uploading logons and other issues yeah remember we said whatever happens whatever weird things happens after this is announced everybody is going to Nothing has to do with the change in control. They're saying that Murphy's law was absolutely in effect this weekend. Absolutely.

20:16Absolutely. EU announced a formal investigation of XAI over people using XAI to create non-consensual deepfakes. And WhatsApp has been officially designated by the EU as a very large online platform under the Digital Services Act. that brings extra reporting responsibilities for its public messaging channels. This is not the interoperability. That's the Digital Markets Act. This is the, if you have public posts, you have to be transparent. You have to moderate them. You have to limit exposure to children, all that. WhatsApp has enough public messaging channels that they now fall under that. Microsoft is investigating reports that a limited number of Windows 11 devices failed to boot after installing the January 2026 security patch.

21:02This comes after Microsoft released an out-of-band update Saturday to fix freezes in Outlook when PST files were stored in cloud drives. Yeah, that's the second out-of-band update already this year. We're not even out of January. It looks like we might be on our way to a third. We've got a week to go. So, yeah, it's kind of crazy. Benchmarks for Intel's Core Ultra Series 3, a.k.a. Panther Lake, are out, and they're pretty good. They show the X7 358H and the X9 388H beating the Apple M5. They do not beat the M4 Pro. They beat the M5. Still good. However, later this week, Apple is expected to release new laptops with the M5 Pro, M5 Max chips.

21:43So Panther Lake's reign in this particular niche may not last very long. YouTube music now syncs. You're now playing cue across devices, including Android and iOS. so you can keep listening to the queue right where you left off. Bloomberg's Mark Gurman reports that his sources say Apple's going to release that update to Siri in beta in February. That would be the first version of Siri to be powered by Google's Gemini. And data from domain name stat reported by Sherwood News says that there are now more than a million websites using the.ai domain name assigned to the island of Anguilla. Three years ago were around 100 ,000.

22:21Yeah, apparently 22 % of Anguilla's budget comes from domain name registrations. And there are a lot of really itty bitty countries that that's how they make an enormous amount of money because they've got cool. Tuvalu with.tv. Yeah, exactly. We in every episode of DTNS with some shared perspectives. Today, Anthony has a word of thanks. Yeah, Anthony wrote, hello. Just wanted to say a quick thanks for the info on the Gmail Pop3 story. When I first heard about it, I assumed that wouldn't affect me in any way. Then my mother texted me that her at MSN.com address that she's had for decades and accesses through Gmail stopped working and she was very upset.

23:04Thanks to that episode, I knew exactly what had happened. Got her set up on the Outlook app. So thank you. Anthony, that's great. I'm glad that that helped avert a crisis and got your mother back on her email fast. That's great to hear. That is good news. And I would imagine that there's a lot of folks that are in that type of situation. They're using Gmail as an email client to access mail somewhere else. And they've just been doing it for 15, 20 years. It's like, oh, it's not working out. Let's go figure it out. It's probably this. It's probably that they got rid of the pot. Warms our heart to hear that helped.

23:38So, folks, we want to know what you're thinking about. So if you've got some insights into a story, please, please, please share it with us at feedback at dailytechnewshow.com. Yes, please do. And big thanks to Anthony for contributing to today's show. Thank you for being along for Daily Tech News Show. You keep us in business, folks. You can do it by subscribing to our Substack, dtns.substack.com, or becoming a patron, patreon.com slash dtns. Talk to you tomorrow, everybody. Have a good weekend. Or good Monday. What am I talking about? What day is it? Stay warm. It's no solar.

24:19The DTNS family of podcasts. Helping each other understand. Diamond Club hopes you have enjoyed this program.