Episode Transcript
[00:00:00] Speaker A: Yeah.
And welcome to another episode of the Examining Podcast, the technology focused podcast that dives deep. Good afternoon.
I'm trying to think of a good nickname. I just say, chris, I need something else to say.
You don't have a nickname for you.
That's what we need on this podcast. We need nicknames.
[00:00:23] Speaker B: Exactly. Yeah. I won't tell you what some of the. Some people have called me before, but.
[00:00:28] Speaker A: Really, you have nicknames from other people? Oh, that's. That's. Yeah. You don't have to share those. It's okay.
I'm trying to think of them. I don't have any. That would have any. It wouldn't make any sense to people. I think that's part of the problem with nicknames. Right. They're so.
[00:00:41] Speaker B: That's true.
[00:00:42] Speaker A: Specific to the people who know you. So people would be like, what, you think it's really embarrassing because you know the context. Other people would fly right over their head.
[00:00:49] Speaker B: Well, I don't even know if they're necessarily embarrassing, but it's just, you know, like, for example, maybe I'll just tell you.
I'm getting delirious now. But like, one of them is Hansel, so.
[00:01:04] Speaker A: Oh, yeah.
[00:01:05] Speaker B: Like, remember there was that one movie with Ben still and. Yeah. Zoolander.
[00:01:11] Speaker A: Hansel.
[00:01:12] Speaker B: Yeah. So that was one. What else do I have?
Christopher Columbus.
Kristoff.
[00:01:19] Speaker A: Kristoff. I like Kristoff. I can use that. Christoph. Hansel.
This is irrelevant to this podcast, so I hope people are interested in this intro. Nicknames. That's what we have to come up with.
But we're going to talk about. We have two things to talk about today. One of them is big news coming out of Apple. It's not just an Apple podcast. We want to talk a little bit about the fact that Tim Cook is going to step down in September 2026 as CEO and John Ternus is taking over. John Ternus is a product guy. So we're going to talk about what that means and kind of why that matters in the current tech industry climate and just kind of discuss it generally. And then we were going to talk about all the rage. Anthropic is all the rage right now. They're having a moment, but we want to talk about their Mythos model, which they're like, just kidding. Not releasing except to special companies because it's so powerful that they're somewhat concerned about it. So those are the two topics today. And then we were going to end with a tip about. We're going to talk about Notebook lm. Since we've been using that quite a bit.
So maybe we just get into the first one. So Tim Cook, who has been, I think he's been CEO of Apple since 2011, he was famously operations officer. So COO. Yeah. So Tim Cook is one of the reasons why the supply chain for Apple is so tight.
I don't know if, if he did the offshoring or some of it or just tightened up supply lines or. But he's just very good at that stuff. And so that's one of the reasons Apple, you know, was able to kind of dig itself out from its problems because you didn't have like a lot of excess inventory. He just really, really cleaned up that system. And Apple runs well, they're going to run and they might be running into problems as a result of all that offshoring now. But you know, for many years he was a operations person, less so a product guy.
And so he's been CEO since 2011.
He's still going to be executive chairman. But John Ternus, who has been at Apple I think since 2001, product hardware engineering. So hardware engineer is taking over as CEO. And what's interesting is that John Ternus was actually one of the people, if not the person or one of the main people who oversaw the transition from Apple moving from intel to Apple silicon.
That's 2020. I can't believe it's been that long.
But you know, Intel, Apple Silicon is so good.
It's, you know, it's gotta be tough to be a Windows person right now and looking at these computers because it's just remarkable what they're able to do and that the integration between their devices.
He oversaw that transition.
But it's an interesting discussion because Ternus is a hardware product person.
I thought maybe we could just talk about, you know, what we think that means. And kind of this era of more of the tech industry has moved so heavily towards its focus on services that this is almost going the opposite direction.
[00:04:34] Speaker B: Yeah.
[00:04:35] Speaker A: And what that stands, right. Apple often either is the last to a common thing that everyone is working on or they go in some other direction. Like they went into retail when everybody was getting out of retail.
And so this is very like them.
But I wonder if we'll start a trend. I wonder if it says something about the subscription tech mega tech company, business model.
[00:05:02] Speaker B: Yeah, I mean I, you know, I look at it like with, if you look at Apple, putting a hardware engineer in the top job feels like a quiet admission that eventually the product has to matter more than the spreadsheet.
And so you know, the, if you look at it, you know, you mentioned the subscriptions, but the, those services margins, they're, they're quite addictive.
[00:05:26] Speaker A: 76% for services versus say 40 or 25 on products, depending on what product it is.
[00:05:32] Speaker B: Right? Yeah. But at a certain point like that, the business starts optimizing for extraction over delight. And so the customers feel it long before the analysts do.
[00:05:45] Speaker A: And so, yeah, that's kind of what Cory Doctorow talked about with his book and shitification. His idea, right, is that you, you just, you constantly squeeze your consumers, you squeeze your business customers to kind of get more revenue out of your existing base, but you're not, you're not creating like a kind of that exciting experience that got your original base there in the first place.
[00:06:07] Speaker B: Yeah, yeah, for sure. And I mean the subscription, it's supposed to support a product, but too often the product starts serving the subscription. And you know, while Wall street loves that recurring revenue, the customers love, you know, getting something that's actually worth returning to. And so I think this is good. Like, you know, the interesting part is that it isn't that Apple pick Turnus, but it's that the, it, this pick, it actually reads like a statement to people and you're signaling that, you know, the, the operation excellence wins for so long that, you know, you start to crowd out the, the product courage.
[00:06:54] Speaker A: Yeah, yeah, I agree. And it kind of, it does go in the face of kind of everything that's happening. So like we see know other companies, so like, you know, Microsoft, you know, it's a software company, but software is still a product, but it's really moved into a subscription model and it's very difficult. There's one, I think there's one tier of Office you can buy as a standalone for a computer. Everything. They've pushed people to Office365.
Continual price hikes. You know, when they added there was a 30% price hike to a bunch of these plans because they just auto included AI, even though nobody asked for it. Right. So this is the problem. Once you get people on a subscription, you can raise prices. Happens in the gaming industry. You know, Microsoft has a gaming side, Xbox. Right. Hasn't done as well as Sony, but it's still a very profitable business. It's just not, it's not as impressive under the umbrella of a company like Microsoft that makes much more money on its other divisions. But it's not an, it's not like it's a not profitable business.
So then they went to Game Pass and that's like the subscription model, you can sample all these games and it used to be a great deal, but then you know, they buy a big company like Activision, they want to roll that stuff in.
Well, if it's going to be part of your subscription and people aren't buying it day one for, you know, $70, $80 a game, it's going to cut into that day one pricing. So then they got to raise the prices to make up for it. So I think until they just lowered it because they had been doing so bad, it's amazing that they just lowered the price of that. But it was like $35 a month Canadian or something. It went from 25 to 30, $10 a month increase. It's ridiculous. So they had to lower it more on the Microsoft side. You know, they've, they've really tried to push AI into everything. I think that's the, I think that's the reason that Copilot, which could have been a great brand, has been kind of sullied because they've tried to shoehorn it into everything even though that everything is called Copilot, but that doesn't have the same functionality. So the co pilot plus PCs, you know, especially those ones that run arm based chips from Qualcomm, they have comparable performance to Apple's then they're very nice devices but you know, they, they're pushing this like constant like AI pressure.
All these companies, whether it be a streaming company or a desktop company, they're all trying to like squeeze their customers for this like recurring revenue.
And I think, I think where people get tired of it is that they have a lot of these subscriptions as you and I have talked about in the past, but often they're just not very good.
So when I look at Apple's revenue for their services business, a lot of that comes from the money that Google gives them for being the default search engine. That's a huge percentage of their services business. So that this isn't just coming from Apple Music and Apple TV plus. But I don't, I don't find that they really market that stuff super aggressively.
It's there.
[00:09:56] Speaker B: Yeah, no, for sure. I mean this is where like I feels like, like if your most profitable business model is slowly making the, the product more annoying, that's not really strategy. That's kind of like you're just drifting. And I mean this is where you can see like look at with Copilot, increasingly their business model is looking for a use case.
[00:10:20] Speaker A: Yeah.
Because it doesn't make any money.
Yeah, yeah. I agree.
And you know, maybe we go back to, I'm trying to think of if probably I saw an article recently. I didn't, I didn't link it in our notes, so I apologize. But it just occurred to me we could probably find it. It said that, you know, Tim Cook's biggest legacy was actually they thought was a product which was the Watch and Health, turning Apple into like a health conscious personal product. But even then, you know, the, the Apple Watch is like the best selling watch on the planet. But it's a, it's a physical product. Right. The software, the software alone as a health service on a phone wouldn't be the thing. It's, it's part of that product. They're interlinked. Right.
I think about things like then the MacBook Neo, it's like, you know, out of the box you can get all of the Apple Office equivalent apps and they all run really well and it kind of does everything you need with a minimal amount of RAM and, but it's a very low end, low cost laptop with some compromises, but it's a really sturdy product.
I don't see a lot of headlines for companies anymore based on their, on their services. You know, when you, when you look at just what people are interested in, it's, it's like a product that gives you something in, in return. You could even say that a subscription service that's like an AI subscription like Claude or Chat gbt, especially if you use it as a programmer. That is a compelling, delightful product because it does all sorts of cool stuff. Right. It is a product in itself rather than trying to like.
So maybe it's not just services that's the problem, but maybe it's like this, this increased bundling of services and then raising the prices to justify the bundling even though people didn't ask those things to be looped and glued together. I feel like that's where we're starting to see a lot of problem. It's with the bundling of these services.
[00:12:25] Speaker B: Yeah, yeah, exactly.
[00:12:28] Speaker A: I don't know. Do you have high hopes for John Ternus?
[00:12:32] Speaker B: Well, I mean some of the, the coverage if you look at it, I mean there, there, there has been some failures like the Vision Pro for instance, or even imagine putting in. And we've talked about this before. A billion dollars, 10 billion into a car project and nothing came out of it. And they were hiring people from like Mercedes and BMW and all sorts of companies.
You know, it's, it's not an easy thing. I mean now apparently some of the rumored product development.
They're actually going to take that Vision Pro technology and put it into something more like a meta Ray Ban type of glasses, you know, might be good. I mean, who knows?
It's. Again, it's. I think this is where in some ways like the, there's maybe too many products coming out and the product lines are getting like, you know, there's going to be the rumor or the hype or anticipation is that there's going to be an ultra version of the iPhones
[00:13:40] Speaker A: that are foldable for this fall and some sort of Ultra MacBook Pro with a touchscreen and yeah, not that I really look for a touchscreen on a laptop. I still don't understand why people want this.
[00:13:52] Speaker B: Yeah, yeah, exactly.
[00:13:53] Speaker A: I've never touched the touchscreen.
[00:13:55] Speaker B: I don't know. It's. It's funny like I was just meeting with somebody for a client meeting and it was an Apple laptop and they, they constantly were touching the screen and just to explain things. So I, I can see people I don't know. I personally don't. But yeah, I could, I could see the, the behavior like that. It seems like it might be something that you, you would intuitively just touch the some aspects.
[00:14:20] Speaker A: I also wonder too if you know, despite we can criticize Apple for Liquid Glass, a lot of people really don't like doesn't add anything that introduces some accessibility problems. I feel like this will get refined over time and Liquid Glass will just become fine like every other interface that's been refined. I think in the interim it's difficult but it is important that they, the rumor is that they are going to.
Their next operating system is going to be. Well, for people who don't remember, it's going to be like Mac OS X Snow Leopard, which was famous for just kind of refinement and kind of cleaning up everything. It was very rock solid release and the rumor is that that's what it's going to be and it's. And it's interesting that they're doubling down on that rather than focusing on features and design changes. I think perhaps there's a lesson from Windows about what happens when you neglect your operating system. So like Windows 10 had, you know, all sorts of problems but was finally refined. Windows 11 still has tons of problems. It's notoriously slow. Things like the File explorer on Windows 11 are incredibly slow.
Things like the taskbar is really glitchy, you know, shoving AI into everything, even though it's like the window to their whole company.
You know, pun not intended. It's the lack of delight and kind of usability, the expect the expectation that it would run smoothly and perform. It has kind of driven a lot of people away from the Microsoft platform. You know, we talked about this in the past, but even like Linux growth, because you know their hardware is abandoned, Windows won't support it anymore. I wonder if this is if Apple just got lucky releasing the MacBook Neo because it's such a low end, inexpensive entry point into an operating system that tends to function much snappier, much more consistently.
[00:16:21] Speaker B: Well, and you know, like speaking on the AI side, I just remembered what we were chatting about earlier. But you know, this, this timing, you know, you look at it, Tim Cook, I mean he's taken this company, took it over in 2011 from Steve Jobs, which at that point it was like a $300 billion company and now it's like a 4tr trillion dollar giant.
There's nobody who's going to have a legacy that will be able to match something like that.
But if you look at IT leaders, they, there's a certain pattern at that level.
They want to leave when they're still on top.
And so this timing matters. And you know, Apple right now it's, it is trailing on AI, that gap, people are noticing the attention.
And so maybe that this is part of that narrative, you know, in terms of Cook's tenure, it's like you're passing on the baton to this somebody who is a product guy that might be able to figure out these issues. And so either way, either it'll work out or, you know, he ended up on a high note.
[00:17:33] Speaker A: Do you think that people are not buying? Like my understanding of the conversion rate is they've had their best quarters yet in China. China's had a struggling smartphone market.
I have a iPhone 15, so I don't have the AI features. I'm the last generation that doesn't support it.
But I don't, I don't find, you know, if I, if I have the Claude app, if I want to do something, I don't, I don't find it to be lacking. Are people really buying these devices because of integrated AI features? Like my understanding on the Windows side is that it's just all over the place. People are tired of it.
[00:18:11] Speaker B: Well, I mean again, some of the things that they've come up with, like one of the anticipated changes is going to be a totally revamped Siri and which may be able to go and work with. And again you look at it like they were. Apple was one of the first to come up with this type of voice assistant.
And so for the longest time, whether people are buying it for that feature or not, but at least it worked and it was, it was made fun
[00:18:38] Speaker A: of for a long time.
[00:18:39] Speaker B: Yeah, well, but I mean for the time it was still, you know, it helped with the overall experience. I, I would think. Right. So.
And for simple things, maybe it's still fine, but now it's, it's gotten to a point where, I mean, maybe there's, it's a missed opportunity and I think that's where that gap is coming. I mean imagine these companies that, I mean, well, they're not profitable when you look at OpenAI or anthropic or any of these companies, I mean they're burning more than they're actually bringing in given the development. I mean it'll be interesting to see like I look at like Microsoft and I think we talked about this a while back but you know, they are actually going and making, severing their ties with OpenAI and slowly this was, what's that?
[00:19:29] Speaker A: Very slowly, Slowly, slowly.
[00:19:31] Speaker B: But I mean the, the one thing is, I mean if you recall there was that Mustafa Suleiman that they took.
It was kind of the most strangest thing because it wasn't, it was almost like a buyout that we talked about. It was like a buyout but not really. Like you basically had him come over and took everything. It was almost like what happened when Steve Jobs came back into Apple and so he became the head of Microsoft AI. And, and they are planning to go and launch their own frontier models that are going to be on par with whatever is out there in terms of,
[00:20:07] Speaker A: it's really been stepped back from the product side of the AI. So he's not, anymore, he's not overseeing the co pilot side anymore, but he's more focused on the back end, the models which, you know, seems like a bit of a demotion I think from a public view, but perhaps that's what he wanted to do.
Yeah, I mean they're going to have to do their own thing at some point. They don't want to be dependent on the next Netscape, which is what OpenAI looks like increasingly.
Perhaps that's a good segue to the second topic. So Anthropic previewed its new model Mythos, which is its new model.
And so it unveiled this on the 7th of April. So not that long ago was called Project Glasswing, you know, and it was, I think it was leaked.
It's, it's, it's, it's Code name was re leaked that they were working on this frontier model. It's a large model, it's not a specialized model. It's another frontier model that supersedes kind of Opus. Opus 4.7 is the new, the newest version of Opus, which is their thinking, their, their best model.
And it said that it has autonomously found thousands of zero day security vulnerabilities across tons of software, including a 27 year old security flaw in OpenBSD, a 16 year old FFmpeg flaw, a chained Linux kernel privilege escalation. I actually don't know what that means.
And so there. It's amazing to me that this has happened because Mythos is not trained specifically for coding for bug fixes. It is a just another general model but a much more powerful one.
And so they were concerned about releasing it because of course then you, you know, you unveil this model that can find zero days and you know, hackers can subscribe to Claude Pro and Prod Cloud Max too. So they, they allowed 40 vetted organizations, so like Amazon's, AWS, Microsoft, Apple, Google, CR Strike, you know, JP Morgan, Vinnix foundation to have access to this.
And so it was the only time, a second time since they said, according to scientific American, since GPT2 in 2019 that was a myth, a model was withheld because they felt it was too dangerous.
They think GPT2 is dangerous. They, those people would be just floored by what we're releasing now.
So it's interesting so that we have these decades old bugs just, just incredible. Now in some ways it's a really good, it's a really good news story because we have these tools that can find zero day vulnerabilities. So you know, good people and bad people have access to the same models but apparently over 99% of the vulnerabilities that found are unpatched.
[00:23:12] Speaker B: Yeah, well and you know if you look at it like the Mythos is impressive probably in the the same way like nuclear technology is impressive, you know, the capability is real, but so is the discomfort.
[00:23:26] Speaker A: Yeah.
[00:23:27] Speaker B: And like you mentioned like you know, with who's has access and I think that's one of the, you know, the most revealing parts of this isn't what the model can do, but how tightly access is being controlled.
[00:23:40] Speaker A: Mm, it's really funny. Oh and it's funny too because so we had that all hoopla in March about the Pentagon labeling anthropic as a supply chain risk under the Trump administration and they've of course taken them to court and I don't know if that's been resolved yet. But now, you know, these people all want access to this model.
The people who've deemed them a threat to national security.
And so it's, it's kind of interesting about face.
[00:24:10] Speaker B: Yeah, well, and again, like, I mean, in one of the organizations is reportedly the NSA, but if you build something and decide only 40ish organizations should touch it, like you're, you're making a strong statement, maybe even stronger than anything that a press release would ever, ever could. And so, I mean, look at all those people that canceled their subscriptions from OpenAI to move over to Anthropics. Claude. And you know, they're basically doing maybe the same, if not, maybe worse by. And controlling and giving access to only these 40 companies.
[00:24:45] Speaker A: Yeah, they, they could be. Do you get the impression that. So, I mean, there's been some discussion about Mythos if, if it's, you know, as, I mean, this is a huge marketing for them. They're in the press, they're talking about it makes it look like they're leading the pack. Do you think that they've lapped open AI at this point?
I mean, it certainly sounds like it based on, you would think, from press coverage, but I'm wondering if that's true.
[00:25:11] Speaker B: Yeah, I mean, I, I don't know. I, I guess certainly like from that press coverage. I mean the promise that they're kind of making is there's like faster defense and. But the, the fear is like off faster offense and so, and maybe both are probably true.
But yeah, I mean this machine, if it, if it helps patch the hole, it can also maybe find the next 10 as well. And so, you know, that is pretty interesting how you could maybe go and take that exploit window, taking it from like months to minutes.
And so, yeah, I mean, this is, it's interesting kind of development for sure.
[00:25:56] Speaker A: Yeah, we'll have to see. I hope I get access to it on my plan. Probably not.
[00:26:01] Speaker B: Yeah, but you know, again, it's, it's funny how like this, like how you mentioned too, this capability emerged as a side effect.
So.
[00:26:10] Speaker A: Yeah.
[00:26:10] Speaker B: Isn't that like. And this is where again, like all these AI companies, I mean, it isn't really. This is like the most unsettling part is that it's not what the frontier model can do, but it's, you know, what it does on it without you even thinking of it. And so there's these like, unintended consequences.
[00:26:33] Speaker A: Well, I think it speaks.
So it does a couple of things. It tells. So for the last year, there's been a lot of coverage talking about how these frontier models are too expensive, they've hit a wall, they're not going to get that much better.
Our good friend who we've never met, Cal Newport, has talked about this. He actually has a podcast episode. I haven't listened to it all, but asking, is this marketing? Is this real?
He's had the same. The same argument.
But if it is as good as they say, clearly the frontier models a have not hit as big of a wall as been predicted. So that's one thing.
And so even if a model that's not tuned to find security vulnerabilities specifically is able to do this, imagine what this could do if all it was designed to do is find security vulnerabilities. I think that's what's. And you know, you and I have spoken in the past about small language models or perhaps more specialized language models. Are we going to have these mega models that use up tons of electricity?
It's hard to say. But you know, this, this to me is like scratching the surface because it wasn't at all intended to find. Like you said, find zero days. But what if they had like the Mythos Security Vulnerability Edition that just strips out everything that doesn't matter and is targeted just to find stuff like this? I mean, I think that's more the interesting thing.
[00:27:57] Speaker B: Yeah, yeah, for sure.
[00:27:59] Speaker A: So we'll see. But I'm. I'm very impressed with Opus 4.7 currently. So that's, that's what I'm working with.
But I do miss Chat GBT a little bit.
[00:28:11] Speaker B: You're officially off.
[00:28:13] Speaker A: I.
I don't. I'm not letting. I haven't decided.
It gets expensive, Chris. You know, there's, there's little things I like from a model perspective and a writing and a summarizing. There's a lot of things I like. I do think that deep research, I think for engaging, for creative brainstorming, I think OpenAI does a better job from a tonal perspective for writing. Anthropic doesn't sound as much. I've talked about this in the past as much like an AI, so it matches your tone if you give it a sample.
I find deep research is a lot more thorough with ChatGPT. So as a research tool versus an analytical tool, that's kind of where I'm seeing the difference. If you're using ChatGPTs primarily as like a general tool and you know, one stop shot images, that's probably the best. Everyday man, our colleague Chris, I think said that AI, but really does a really good job of kind of that as a research tool.
Or Claude is more of an analytics tool. That's how I see it.
So it kind of depends what you're doing.
But we know with that. That's a great segue into our tip.
So you and I have been playing with this. I have an example to speak of Notebook lm. So that's a Google product.
I'm assuming it uses Gemini in the back end. It has to be. Yeah, must do the same thing. And so it's there.
How would you describe Notebook LM to people? I mean, I think we've talked about it in the past, but is it, would you say like a learning tool?
[00:29:51] Speaker B: I. I would categorize it as a learning tool.
You can upload a bunch of resources. There's different options where you can go and have it create quizzes or infographics. Slides.
[00:30:05] Speaker A: Yeah, make a podcast, put us out of business. I've been doing that a lot, so it's really amazing. And one of the things that I've been doing is that, you know, uploading resources.
So I, I subscribed to a quite a few newsletters. There was a whole bunch that I wasn't able to read by Jacob Nielsen, one of the founders of the. The NN Group, the Nielsen Norman Group on design and ux. And, and so I. You. What you can do with Notebook LLM is that you can basically compile a notebook of. Based on sources. So I uploaded all the links and I had it generate summaries. It generated an audio podcast between two robot interviewers, slide decks. I was trying to. I was trying to generate some flashcards based on Gestalt principles. So Gestalt principles are, are really common principles of, of design.
And I was just, you know, refreshing myself. I. I recently got an. The second edition of the Laws of ux, which, which incorporates a bunch of the Gestalt principles.
Gestalt principles are from like, kind of the Berlin school of psychology. These ideas of, you know, cognitive processing and how people, you know, see interfaces and inputs, it all comes from psychology. But I kind of wanted to create some, you know, refreshers, some infographics. So I've been using that to create slides and quizzes.
There was an article I was reading that I took a bunch of notes on recently on information literacy, also library instruction in the age of AI. And it generated this really cool infographic that kind of outlined, you know, the different tiers of literacy, of AI literacy. And you've seen Some of this stuff, it's quite remarkable what it can generate, especially if you were, you know, an instructor and you maybe have a lot of your own content, but you're looking to generate some, you know, infographics or images. I think this could be really powerful, especially for illustrating concepts that you're trying to learn. Right? As a.
It's. It's just a great.
Yeah, just. Just a great kind of not summarizing engine, but almost like a reinforcement tool. Like, I have this slide, you know, that's designed for Gestalt Principles on common region. This idea that, you know, if you group things together on a screen, the brain sees them as belonging together. But even more powerful than that is if you put, like a box around it, they're in a common region. So a form that you fill out for your driver's license is a good example. One of the reasons why, you know, evenly spaced things are not a good idea is that even spacing doesn't group things to make alike things belong together. So you don't want even spacing. And so it's actually made this entire slide deck based on my notes on Gestalt Principles and stuff like that. And it's really quite remarkable. And so, yeah, really cool.
I just wanted to point it out. It's a great learning tool. I've been using it a lot to catch up on articles that I can't read or get to. I can, you know, have them audio. Audio, audio podcastified or whatever you call it, so I can take them on walks. And it's just a really amazing, amazing tool.
[00:33:18] Speaker B: Yeah, no, absolutely. I mean, I first came across. Actually, it was some students that told me about it and then, yeah, I just kind of just thought I'd experiment with it and see what happens. But again, there's a. There's this is. Isn't it funny that, like, it's. It's kind of hidden within the. The suite of apps that you have within Google. And so but if anybody ever has a chance, like you should, at the very top there's this icon and it launches the studio.
And so that's where you get access to a bunch of these things within NotebookLM.
[00:33:58] Speaker A: Yeah, it's just remarkable. So just as something I point out, I would encourage people we'll come back and maybe report back based on other things that we do with it. But it's a pretty cool tool.
But with that, that's pretty much all I had for today. Did you. Anything else you wanted to chat about?
[00:34:16] Speaker B: No, I think it's good.
We've We've had, you know, we talked about product and AI and. Yeah, I think it's.
We'll see what more is to come over the. The coming months.
[00:34:32] Speaker A: Yeah.
See which company burns through its cash faster.
Okay. Well, take care.
[00:34:40] Speaker B: Yeah, take care. Goodbye.
