Bubble Talk, Gemini Rising, and RAM? In This Economy? - DTNS November in Review

0:03This is the Daily Tech News Month in Review for November 2025. We're going to look back on what the big stories of the month were and try to help each other understand what just happened this month. I don't know. It's a month full of news. We're going to talk all about it. Well, at least a couple of things anyways. Yeah, we've got what do we got? We've got bubbles. We've got Jiminy's and we've got RAM prices and crickets. I'm Tom Merritt. I'm Jason Howell. Let's look back on the big stories of November 2025. first uh we like to ask an llm what it thinks are the big stories of the month uh and because it was one of the big stories of the month we thought we'd ask the new gemini 3 pro uh so let's run through this yeah well is it is oh i don't yeah i guess we're about to find out uh get it right we have our own stories that we've picked but let's see what jiminy picked uh jimmy picked AI and its economic and workforce impact.

1:03Header. This is, yes, in big header font. The AI rally and market milestones. The continued massive surge of tech giants, particularly Alphabet. Huh. Gemini thinks its dad is doing really well. The surge attributed to investor confidence in Gemini 3. You know, me. The AI layoffs trend. News from companies like HP announcing plans to cut thousands of jobs, underscoring a broader industry-wide concern about AI-driven workforce displacement. Government and infrastructure investment. Significant announcements regarding massive AI infrastructure spending like Amazon investing in data centers. The Genesis mission that we talked about from the president.

1:50So I feel like it combined like 12 stories into one with this one. but it does touch on some of the stuff we're going to talk about when we talk about the tech bubble. So I'll give it a half pass. Yeah, I'll give it like a sideways thumb. You know, it did the thing that AI is really good at. It's like, here's a bunch of things. Now I'm going to like interconnect them and mash them together into one giant soup. Yeah, yeah, squish them up into a sentence that grammatically makes sense. The second one is fall product launches and hardware battles. I did this as a big story last month. I feel like there was more product launches last month, but Gemini mentioned OnePlus 15, the iChu 15, Realme GT8 Pro, and of course the implementation of Snapdragon 8 Elite Gen 5 across a bunch of devices.

2:39Also, it talks about agentic AI in devices with more on how new models like Google's Gemini. Did I mention Gemini, by the way? uh and then uh the rise of assistive smart glasses uh although it talked about meta not yeah there's so many other smart glasses that have launched over the last couple of months i feel like this is a miss i don't i don't think it really got got this one right yeah i'm not sure i mean maybe the agentic a i i don't know i feel like these are kind of like blankets that could be applied to many different months i don't know that this particular month was really strong in android flagships particularly strong and agentic ai maybe you could make the case on that one actually with claude uh the the opus uh update you know perplexity browser on android that was another story like there's certain things there um smart glasses i'm kind of like wait a minute that that was like a last month yeah that was an october did you get the right month gemini yeah i'm wondering if gemini was too focused on making itself look cool and i think i might to get the date right let's see if its third topic concludes itself uh legal and regulatory environment continues to shape the power of big tech uh antitrust rulings fall out from the antitrust cases notably meta's case being dismissed and the final remedy in the google search monopoly case surprising experts as being less punitive than expected well that was definitely last month the google one um the ai regulation tug of war pressure from u.s tech and industry concerns uh leading to a reported re-evaluation or potential delay of parts of the eu ai act that is definitely part of this uh doesn't mention gdpr which is i think a bigger part of that uh review of eu regulation um anyway highlighting the tension between promoting innovation and establishing ethical and safety guardrails again i don't know yeah i think when what we're realizing what i'm realizing uh or reminded of is that ai is just not very good at like time time-based stuff you say this month and it's like it doesn't it doesn't know what that means like do i you know do i cut it off at uh at november 1st and only take stories that like were birthed in this month or stories that have been around for two years that had like a single sprinkle of something this month.

5:06It gets confused in this stuff. And so it ends up including a lot of randomness. What is even a month, man? Totally, man. Seriously, I need to use Nano Banana Pro and have it do me an infographic to explain that to me. Here are some things that we considered, although we did not choose them as the big stories, that were not considered by Gemini. Google Aluminium OS. uh just just a big story on on november 25th uh the settling of the epic and google lawsuit it had a chance to talk about its dad again and it didn't bring that up in the antitrust stuff um that was a big one like them you know declaring peace with third parties and and side loading and all that uh the valve steam machine and the steam frame uh i mentioned the changes to gdpr uh thieves not wanting uh sent you to get your samsung phones uh probably not big enough to be a story that was kind of interesting uh the cloud flare outage explanation co-pilot coming to more windows succession planning for tim cook at apple uh kill switches on buses made by chinese companies in denmark and waymo hitting the freeway was a big story on november 12th all right let's get to our actual first topic is there a bubble i feel like this is the story of the month.

6:25We opened the month on November 3rd with a big story about all the deals being made. And Gemini actually mentioned the Amazon one. It did not mention Microsoft's$9.7 billion deal with data center firm Iren. It didn't mention Meta's one gigawatts of solar power that it's after or Meta applying to become an energy trader. And I think even bigger towards the end of the month, the building buzz about whether we're in a bubble, Wall Street starting to get cold feet, and then everybody looking at the NVIDIA earnings, which turned out to be very strong and very good and settled all of Wall Street's concerns for less than a day before they went back to being concerned about whether there is a bubble.

7:10We've talked about it a lot on the show, but I think the fact that everyone is still talking about the AI bubble, and now I feel like by the end of this month, it's less of, is there a bubble and more of when will it burst and how will it burst? The tone has really changed. Yeah. I mean, obviously, between all the shows that we do, this is like a recurring topic on so much of what I've been doing the past month. And I think where I've settled on this is less that it's, I mean, okay, technically it probably is a bubble. Basically, my thinking on this is if we're thinking, oh, hey, this is a bubble and And it's going to burst and it's going to take everything down with it.

7:53I don't actually think that's going to happen. Right. Like, I think it's more of a correction. It's a bubble. Yeah. I mean, and sometimes I get the feeling that when people talk about a bubble, that's what they're talking about. They're like, and everything in hell is going to break loose. You just wait and see. It's going to take everything down with it. And I just don't see that happening. I see a correction. And maybe it's the stands, you know, how you define bubble. It might be the same thing. but um yeah it was interesting how it was all bubble bubble bubble reaching to the top then the nvidia earnings happen it was like oh okay maybe there isn't one and then it didn't take very long it's it's like there's a there's a driving narrative through this whole bubble talk and it keeps getting pushed and and i wonder where what the there really is there yeah i i see another one of the objections is focusing on the fact that nvidia invests money into a startup that but then uses that money to buy chips from NVIDIA to build its data center and the circular nature of all of this.

8:55All of the economy is circular. So the fact that it's circular doesn't bother me. What would bother me is if that startup is only going to make money off of the investment from NVIDIA, right? Then it's a Ponzi scheme. But if the investment helps them buy chips from NVIDIA that then allow them to grow a business that makes money, that's fine. I don't have a problem with that. That's actually just the way the economy works in a slightly smaller scale. So it brings us back to the big question, which is how are companies going to make money off of this? And all of the AI companies have lots of great charts that show we'll have user growth like this.

9:38We'll charge this amount a month. I think the fact that it is based on paid subscriptions and everyone is used to that already is a positive. We're not talking about, well, we've got to get people to use this for free and sell advertising. we've got an actual revenue source, right? That's coming in the door consistently. It's a matter of whether you can make it useful enough that people say, I want to keep paying for this or I have to keep paying for this because it's essential for what I do. Am I alone here in what I do? Clearly, I'm not entirely alone, but I pay for like four or five of these different things.

10:13Now, granted, I do an AI podcast. I was going to say, I just got the alert for AI Insides new episodes. So that could have something to do with it. And that's why I asked the question. Yeah, yeah. Are there a lot of other people out there that do find the need or the desire to pay for these different services because they are different enough? Yeah. Or is everybody just buying into one? I don't have a good sense of where the borders are. There's plenty of people using it for free. I would say the majority of people use it for free. Yeah, that's true. I think there's plenty of people who pay for a$20 level subscription or they're paying for Google Workspace and adding it on or something like that.

10:53I don't know how many of those there are. I think there's very few people paying the$200, you know, like super pro subscriptions that are out there from Google and OpenAI. But I think there is enough usefulness in this that there are ways to make money. And I think it's easy to criticize like, well, you don't know how you're going to make money, so it's not a business. as we have seen over and over, Facebook didn't know how it was going to make money. Google didn't know how it was going to make money and they figured it out. It was exploiting our eyeballs for advertising. But there are lots of different ways to make money.

11:26And it's not about necessarily whether it's advertising or subscriptions. It's about what are people going to use our product for in a large enough amount that we can generate revenue off of it. And I think there is so much that these LLMs can do that there is going to be enough usefulness that some of these companies are going to be able to generate a profit out of them. I can't tell you right now which ones those are. It won't necessarily be the ones that are being used most right now. And I can't tell you how it's going to look. This might become an enterprise-level thing. And the consumer-level product is almost a loss leader.

12:03And you're feeding special training and inferencing into enterprise-level narrow views that really generate the majority of the money for these companies. That is one way this could work. That's not the only way it could work. And anybody who tells you they know whether it's going to work this way or not is lying because nobody knows. Yeah, that's a big, big question mark. And then there's the Hugging Face CEO, who remember the story, Delang, CEO of Clem Delane, who said that we're not looking at an AI bubble. What this is is an LLM bubble. Yeah, yeah. That seems pretty, I mean, for my mind, that seems pretty spot on.

12:38I think if you understand what he means, I think he's right, which is we are going to have new technologies that are going to do new things and we're going to have new ways of implementing LLMs. And it's the companies that figure those things out first, either by luck or skill or both, that are not going to be hurt when the bubble deflates. Disceptible. Yeah. I do think the bubble will be – it'll take a couple of years for it to deflate and it could go fast or slow. I'm not even sure. well let's i mean just stick a patch on it like seriously do we do we have a patch like if it's leaking a little bit of air like catch it now the ai bubble put it in the bathtub you'll see bubbles where the leak is easily done right all right perfect that's only our first topic hang on there's more

13:28well how about we talk a little bit about gemini 3 not because i am a product of gemini um and i'm talking about myself. No, he's a real human. I've met him. I'm actually a real human. I'm sitting here talking with my actual human mouth. But yeah, at least at the time of the recording right now, last week was a pretty big week for Google. They had the Gemini 3 launch. They had a few days later the Nano Banana Pro, which I mean, was probably a little bit less of a big deal for Google, but still pretty big considering it came on the heels. Well-received in both cases. Well-received. Very well-received.

14:06And Nano Banana, the initial release of that, was a big viral hit. So Google's been getting a lot of praise for what it's doing in image generation. Coming out with its next step and what it's proving itself to be really good at was pretty impressive coming from Google. But yeah, I mean, Gemini 3 definitely bringing a lot of advancements. From my meeting when I was there with Google last week, one of the things that they were talking about is how every kind of whole point release of Gemini has a different focus. I can't remember what they said the first Gemini release was the focus of or if it was just get it out the door.

14:46I don't know what they said either, but the first one was panic. The first one was OMG. Oh, my God, get something out there right now. Yes, totally. The second one was about reasoning and thinking and that sort of stuff. This third one is really about image and kind of design, let's say, you know, really kind of being design focused and, you know, working on a deeper level with the multimodality, you know, being able to feed in very long videos, like hours long videos and having it, you know, pull context. from that. And then what it produces on the other side, the imagery that it can give you, or if you're working with their IDE, their anti-gravity tool, some of the stuff that you can get that to vibe code, if you want to use that word, I always kind of cringe a little bit when I say it, but is very visual.

15:41And that's apparently their big focus for Gemini 3 is the visual quality of a lot of what they're getting. And it's been really impressive to a lot of people. Yeah, topping benchmarks, causing people to wonder if Alphabet has won, which I think it's way too early to say. They've won. Yes, okay, the game is over. Forever and ever, they won AI because of Gemini 3. I think it is an amazing comeback, right? A couple of years ago, people were saying Google was all washed up. They got scooped by OpenAI, and they'll never catch up. Well, guess what? They caught up. Spoiler. So OpenAI is also going to catch up.

16:21Anthropics Opus model is very good. They're going to stay in this race. Others are going to be around the edges like Mistral and Perplexity and Deep Seek. And China has a ton of open models like Quen from Alibaba. Like this is far from over. And you are, I'm going to guess none of the models we're looking at right now are going to be the thing everybody's talking about in five years time. It's not going to be based on any of this. We're going to have something entirely new, could be from Google, but maybe not Gemini based. And so this is great. And I think Gemini 3 is really good. But as we saw earlier in the show, it ain't perfect.

17:01None of them are perfect. Far for perfect. Yeah. They're just always getting a little better all the time. Well, and not only are they individually getting better, but getting better compared to each other. So what we've seen already, and this is going to be no different, it's just we're really close to Gemini 3's release right now, is that every time a new major company releases their model, it becomes better than, yes, their last model, but also inches better than the competition. They're constantly fighting for that top spot, and they only get a short amount of time to really celebrate the fact that their model is in that top spot because there's enough of them that these releases are happening so rapidly, so frequently.

17:43Yeah. And trust me, Google stockholders, or Alphabet stockholders, very pleased right now that the rhetoric around Gemini is so positive and people are so positive about Alphabet. Same issue, how you going to make money off of it applies to Google, although Google does have a pretty solid revenue base with advertising that won't go away overnight. I know there's a little catastrophizing about search. It's not like people are going to stop using search. So these are longer-term concerns, or at least medium-term concerns. And it's far from over. Google has to keep differentiating. And I think the advantage that Google had that allowed them to catch back up is they have fundamental research.

18:28And they have to keep creating fundamental research at DeepMind, which I don't doubt they're going to do. But that is a huge part of why they are able to do this is because they have people at DeepMind working on the fundamental research that they can then apply to these public-facing products. Yeah, that's super true. Yeah, it's an interesting time for Google. I think what you said as far as kind of them, you know, it really was not that long ago when Google was the easy kind of punching bag for the symbol of like, you know, you should have been the one to reign supreme in all of this. How are you left behind?

19:07And now they've really kind of turned it around. Now they're even renting out their TPUs apparently to Meta. To Xenia. You know, and that's their internal technology. Now they're finding a new avenue for monetizing that. I mean, they're in a pretty good position when it comes to all this AI stuff. Yeah. There's a lot of breathlessness around stuff, including that NVIDIA comment that I just made where it's like, yeah, NVIDIA is doing an entirely different thing than renting out some TPUs to meta. Sure. Will Google maybe be able to sell these across the enterprise in the future? Possibly. Possibly not.

19:43And if they do, NVIDIA is not going away. So this, you know, this is a horse race now is really all this means to me. Absolutely. All right. That is two of our three big stories from November or four big stories. If you're a patron, hold please for the next one.

20:04Have you seen the price of Ram? I have not priced it for myself, but I know that things are going up. Oh, way, way up. We talked about this on the Wednesday, the last show of November with me and Sarah Lane. Let me see if I can find this comment that somebody sent to us in response to that. It was some prices for the RAM that they bought not that long ago for$400 was now$1 ,600. What? It's talking about 96 gigabytes of RAM. We're talking about a huge amount of RAM in that case. But yeah, even 64 gigs of RAM that was a few hundred dollars in August is now$700. This is an interesting story because it really has become a bigger story over the course of the month.

20:56We have been talking on the show about SK Hynix and Samsung projecting big profits coming in because of the price of RAM as an early signal that this was not going to go away. And of course, recently, Lenovo reassured investors that it had stockpiled 50 % of what it's going to need all the way into 2026. So don't worry about it. It's not going to have a shortage of components for building its products because it got ahead of this. But that's what's going on right now is the build of data centers has started to cause orders to reserve RAM. The RAM isn't even going into the data centers yet. But the people who are going to build the data centers have signed the contract saying, you are going to sell us a bunch of RAM at this price over the next couple of years.

21:50And so then the companies that make the RAM are saying, well, we don't have any RAM to sell you, small business who makes build to order PCs, because we've contracted to give it all to them. That drives up the price. And as we talked about on Wednesday, there's other factors in here like tariffs and the switch to DDR5 from DDR4. But this is not going to go away. This is going to continue to be a big story. This is just the first month of complaining about the price of RAM. And this is a very long article, I must admit, in the PCGamer.com site. But it mentions the cause being AI. It's funny because we're so used to the conversation being, oh, AI is swallowing all the GPUs.

22:38And that's why there's shortages in GPUs. Now that's extending to RAM, which I would say that that impacts far more people directly. AI has never swallowed the GPUs you need for your desktop game. That was always a misunderstanding of what was going on. Mostly what was happening with GPUs was that Bitcoin miners were buying it. And they were only buying it because the specialty devices that they wanted to buy were disappearing. and that's why gpus have come back down to earth if you go and shop for a gpu now you can find it for less than retail in some cases that was unheard of a couple years ago and it had nothing to do with ai because ai wasn't using those gpus they were using blackwell gpus what's happening now though is as gpus come back down to earth the the ram the ram that you use in your consumer machine is also not the same ram they need in the data center but the ddr5 part of it is and so the factories can only produce so much DDR5.

23:39And if they've contracted to put it in the high bandwidth memory for the data center builder, well, that means they're not putting it into the consumer level RAM that would go into your desktop. So that is directly because people are building these data centers for OpenAI and Microsoft and et cetera. And that is taking up the available component supply chain. It's also going to raise solid state memory. It's already started to raise it a little bit. That one's a little different because again, the solid state drives they use in data centers are not the ones and not at all the ones that go into your machine, but they do use magnetic hard drives for cold storage.

24:20And as you might have guessed, we don't make as many magnetic hard drives as we used to. They're not as big a demand. So those are disappearing and companies building data centers are turning to consumer solid state drives for cold storage. And so they're buying up your solid state drives that you would put in your home computer to put into data centers. That's having less of an effect, but it's still having an effect. And it's going to it's not an easy thing to change. Right. You can't just go turn on the RAM faucet and have more RAM come out. You have to build more factories. You have to convert factories.

24:56and that sort of stuff takes time and expense. And companies may or may not want to spend the time or expense to do that if they're worried that there's a bubble and maybe this demand all goes away, and then they're left with an empty factory. Yeah. What you're talking about, it's not even like a doubling in price. It's like a quadruple. I mean, it's just insane. Double and triple in price easily. Double and triple easily. That's insane. Yeah. Yeah. You asked me at the beginning if I've been looking or in the market, and I have not, and now I'm happy that I haven't. I did notice, as this is the day before Thanksgiving, so we got Black Friday on the mind, some articles that I was reading about Mac sales coming up.

25:40And some of the articles that I was reading were like, don't pay more for the built-in RAM, for the built-in storage. Get it outside of that. And yeah, I don't know. It just got me thinking about this. You might want to pay for the built-in RAM at this point. Yeah, you're going to need that. Check that price. The Verge ran an article that some places like Micro Center have stopped listing the price of RAM. They're treating it like lobster. Oh, wow. They're like, you know, the market changes every day. So the lobster and the RAM are at market price. Please ask your waiter. Oh, wow. Well, what did you think were the biggest stories of the month?

26:19If you have insight into a story and you want to share it with us, you can do that. feedback at dailytechnewsshow.com. Yes, thanks to you for being along for Daily Tech News Show. You can keep us in business. Become a patron at patreon.com slash DTNS. And in fact, if you are a patron, stick around. We have one more story just for the patrons and the Substack subscribers. Airdrop compatibility with Android. the dtns family of podcasts helping each other understand diamond club hopes you have enjoyed this broker