Oral Exams
01:03:19 Share

Alex, you look good, man. You got a job interview or something? No. I used to have this guy who was on my team, and he came in one time. He was wearing, like, suit pants and, like, a nice shirt, and he got his hair cut. And I was like, you got a job interview? You know, that kind of, like, joke. Yeah, and he did. He's like, yeah, man. And then, like, a week later, he, like, gave his notice. I respected that. ♪♪

Hey Alex, I was thinking about you yesterday. I was watching some nerdy YouTube thing on one of the Y Combinator guys talking about with the Notion founder, just about kind of new forms of interface, stuff that we've talked about a lot in the past and how these AI things are fundamentally changing the discipline and how we think about human-computer interaction. Just stuff I hadn't thought about where

Suddenly designers, and this has got to be a good thing, like,

designers have to care a lot about audio and the relationship, like not just how, say in the Sesame case, how, you know, AI speaks to you, but also what the prompts are on the screen, how you know that a computer's thinking, how audio is represented visually. Lots of new challenges, it seems, for the designer. And also this, the kind of rise of

I guess you call it like the Figma-esque infinite canvas interface and how important that's becoming. Like, I think it will becoming as a kind of a brainstorming surface area inside of... I sort of think of it, that's what I would like to see inside of something like ChatGPT. When I'm talking to you on the phone and you're like, hey, wait a minute, look, and you open up something, it's always that. I was very...

pro getting rid of everything like product design documents and game design documents. Usually when you're making something, you write these long documents with all the features on it. And then six weeks later, it's out of date and nobody manages it. And it's hard to add visual stuff to it. And it's like a Google Doc and everybody's jumping in on it. But instead, we use these kind of like infinite Figma boards where...

You can just drop videos, you can drop images, you can drop text into it, you can scribble something, just make it as... And when you need to find something, you just do the same thing you do in real life. You zoom out, right? And so a lot of... Which is why I think we keep using these old analogs of doing things. And we had a conversation about books last week about people not reading anymore. And...

Maybe I was thinking about why I was less freaked out by it than Brian was or even you. But I think it's just like the way people use things changes. It changes dramatically. And I think we're living through this. The years were where decades of change are happening now in user behavior. It's going to be interesting to see, you know. You sound like Steve Bannon. You're like Steve Bannon on UX. You're great. Yeah.

I'm going to just flood the zone. That's a niche. And then I'm going to embezzle some money or something.

It's interesting that that conversation was with the Notion founder. I've got this thing with Notion, with Notion people, because they're the product managers who have sort of usurped a lot of power because everything's been so process-driven. I kind of remember how you used to have to use PCs and how you organized your documents was incredibly important because you just couldn't just search for things.

You know, that was like a real like there was an entire architecture to it and then that just all sort of went away. And I wonder

Trey, you talked about vibe coding, and now Google is actually coming out with AI mode too. But I'm interested in this sort of, I feel like we're in an in-between state with AI's development, whereas right now, it's very process-driven. I shared with you something that this friend was sharing about how they use deep research and AI.

It was very process-driven in order to get to a good result. And I feel like that's where we are with a lot of these AI tools. Would you mind explaining that process at a high level? Well, I told him I was going to keep it like, I don't want to violate this sort of thing. But basically, he was explaining to me how he was using deep research in order to cut out about 50% of the work of AI.

creating, you know, research reports, basically. And, you know, I've been trying to use a lot of these AI tools to get more efficient. And what I realized is that they expose one of my greatest weaknesses, if not my greatest weakness, which is organization, which I think both of you know well. Not that either of you seem like great at it either, to be honest with you.

Alex pretends to be good at it. Well, Alex can zoom out on his Figma boards and whatnot. I don't even have a Figma board. I think I'm generally organized because my mind is not very well organized and I forget everything. But I don't understand what you're saying. You're not like a Notion person. You're not like, you know. I actually think Notion is a tool that's really...

I mean, it's interesting for certain things, but I think it's one of those tools that is actually built with a bunch of old paradigms. I find it's one of those tools that everybody gets excited about using, but then a couple of months later, it's totally out of date and things get lost. I think anything that requires maintenance is just difficult in a company. But you were saying that it requires... Your friend is doing something that requires process management.

I mean, it's a tool. Everything requires learning and how to use it and a process to get it to do what you want it to do in the right way, right? Well, yeah, but I think that that is usually heightened in the early days of it.

in that you have to sort of start to like think as the tool will think. It's just in the same way that to use internet search, you had to do all these sort of Boolean strings and whatnot. There were these little hacks that you were doing with search. Just like when you were organizing your desktop, you had to have different folders and subfolders and everything. And then

Then you could just search for it. Like, you know, and the disorganized people were like, ha ha. But Brian, I think that maybe what resonated with me is that the knee-jerk reaction in media circles early in AI's development was, you know, gosh, we'll never use this. How dare you suggest you take, you know, humanity out of media circles.

And then guys like Ben Fah went down for the count by using it to write affiliate articles or whatever with fake... I don't know what he was actually doing. And then there was this... I think it was a CNET incident or something where they used it to author some product reviews or something. And then you read kind of what you're doing or we had that conversation on the phone and you're like...

It actually really, really helps me in a process of creating content for the multiple distribution points, both audio and text, that you create in every week. And that ranges from organizing a transcript and turning that into the outline for an email to...

to cleaning up your notes. Like it's, it's kind of cool, right? Like you would. And then I asked you how much time it would save you. And you said probably 50%. Like that's a big deal. Completely. Yeah. A hundred percent. But I also, I also wonder, I think we're, we're in a little bit of a, the strange space with the tool because I,

First of all, even the example that Troy shows are not things you're rooting for. And that's a lot of what came out at first. Which example? Well, the CNET slop and affiliate link articles and stuff like that. I think there's people in certain circles that get excited about that, but the consumers don't really benefit much. I also think that the tools are being sold...

The way the tools are being sold is just you talk to them. They're not tools. Everybody's trying to sell them as things that have zero friction and that will deliver results with zero effort, right? And if anything, when you try that,

You hit a wall pretty quickly, but if you use it as a tool and you say, you know, this is the same thing as using Photoshop or a database or something like that. It just has different inputs and outputs. It's hugely powerful and it feels good using, you know what I mean? It doesn't. Yeah. Yeah. That's kind of the epiphany I had, Alex, when I saw this new creation space that bridged curiosity with research, with creation.

And, you know, and kind of a back and forth process. And that was creating projects in ChatGPT and using, you know, it as a canvas to create inside of. And I think that's a, I mean, for me, it's a big shift from writing in Word or whatever. Absolutely. And it's playful. It's just really exciting. And then that's why I freaked out and sent you guys that video from...

Bilawal Sidhu, who is an ex-Google Map guy, who does, I think, really remarkable videos on YouTube. And he showed a bunch of different use cases that were sort of how he was playing with co-generation LLMs and creativity to make things. And those

It included making a little 3D rendered city, you know, another one was throwaway software to manage shot lists for a video. Another one was applying metadata over top of videos that he had to show things like, you know, how many targets he hit at the shooting range.

to it giving him like pretty professional level feedback on a presentation that he was doing when he uploaded the deck and a video of him rehearsing and it breaking it down slide by slide as to how he could, you know, be clearer, you know, use more gestures, you know, slow down, speed up, that kind of stuff.

And, you know, I just think all of these things together or another little use case I thought was cool. Let's say, you know, there's an academic paper or a longer paper that you want to absorb and you don't have the time for it. Or maybe you're not interested in the deep dive. You upload it to Google Notebook, Notebook LLM, and you have it, you know, digested and make a 10-minute podcast that you take on the road with you. We did that two weeks ago. I sent...

my partner this big kind of report on the state of video games and he didn't have time to read it. So we just rendered a little podcast that he could listen to on the ride home. And he, you know, I gave him the key. It's pretty cool, man. It's awesome. You know, it's like the, the Andre Karpathy thing. It was like co-founder of, of open AI. And his tweet yesterday was, you know, the hottest new programming language is English.

It's a pretty, you know, that's a pretty powerful idea. Hey guys, I got to tell you, I got to tell you, and I think, once again, I think if you take a step back and you start thinking about, you know, the movie Her, and you're just going to tell it to all the examples that these companies give about booking a trip for you or finding a meal for you for your business.

wife's 40th birthday or something are insane. They're just the wrong examples to give. But just this week, I was using a very specific tool to create art for video games and it wasn't exporting things properly and a very tedious process. And it has a scripting language called Lua. And I went into...

Claude 3.7. And I said, hey, I need a script that does this and exports all these folders and it names them this way. And it built it and it documented the code and I put it into the software and it did it. And it took me a minute and a half to do something that would have taken me five hours.

And so you apply that across the entire technology ecosystem. You apply that to people who didn't used to be able to program before. It doesn't mean that they're going to build Google Maps using, you know, you're just talking into a microphone. But all these little things that you can now do

It's profound, you know? It's huge once you start using it as a tool and not try to break it or write a Seinfeld episode. Well, I think that is where we are, and that's the sort of in-between, because you have... It's like Cliff Notes, right? I mean, the bad students use Cliff Notes as a replacement for actually reading the work, and...

The good students use it as a study guide to enhance... Or Chegg, or whatever they're called. Chegg, whatever, you know. But the good students use it as a tool to enhance the learning. Correct. And I think that, I mean, Troy, you had an interesting conversation you had shared. I have a professor, a philosophy professor friend who's been a dear friend of mine for a long, long time. And, you know, in some ways he's kind of glad that, you know, he's late in his career because...

You know, the practice of teaching and the institution is going to go through just like, you know, change that I don't know if he wants to be there to kind of navigate it. But, you know. It's too late now. It makes points like it's impossible to teach kids how to write right now. It's really, really hard to.

because they rely so heavily on the tools. He makes points like, you know, good students use these kinds of tools to be more curious and get smarter. And, you know, a big part of the curve of people that are just trying to get through use it to fake their way through. He said that, you know, there's classic things that you do in teaching philosophy, like ask someone to write a paper on what were Hobbes' arguments and what are the counterpoints and how might he defend them.

Right. And, and, and he said that, you know, like someone who really understands the texts and the ideas can kind of do the, how might, how might he have defended them thing? Well, and it would require a lot of thought. He said, but that's a perfect thing for AI to do. He's like, now he's like before 10% of the people would get something, you know, smart and coherent on, you know, that part of the assignment. Now everybody gets it right. Yeah.

They're going to have to completely reconsider how to evaluate people. Well, you have to move to oral exams, don't you? And on the flip side of that, imagine what it's like to be a student right now. You keep being told all the jobs are going to disappear and or change. College is more expensive than ever before. And the odds of you getting a guaranteed job out of college, whichever discipline you take...

right are lower than they were you know five years ago and are probably gonna keep going down so like you know at the end of the day like a lot of these listening to an interview yesterday and a lot of these gen z-ers are like learning how to maximize like right from the beginning they're just hacking the system you know like if i'm gonna get through college i'm gonna just optimize this i'm gonna do you know three kind of hustle jobs on the side and it

It's very tumultuous out there. I'm not sitting down and reading a book for six hours, man. I got no time. I got to fucking find the next meme coin that's going to pop. I'm kidding, but it's a little like that. I do like your point, Troy, about oral exams coming back. I didn't think we had to discuss that. I actually went to school for a couple of years in Europe. I don't know if you had this, Alex, but they had oral exams versus written exams. They didn't take attendance. There was no tests. I always get better.

It was just all like an oral exam at the end of the... I thought you quit school, Alex. At the end of the semester. Well, I mean, before I quit school, I always did better at oral exam and anything free form. But yeah, I didn't finish high school. It was funny. Wow. That's one of my favorite little Alex facto. I love that about you. A little aside, while in Europe... Did you take a Teal Fellowship? Yeah. Yeah.

No, I didn't during the Nazi youth party. Yeah, in Europe, it would be like I would hide the fact that I didn't finish high school, let alone go to college, right? But I remember coming to America, and I think that might have been when I came to say, Troy, like...

I think on the first day, somebody had heard that, hey, this guy didn't even finish high school. And everybody was like, that's amazing. I was like, all right, these people are fucking crazy, but I like it here. Well, it didn't help that half of them went to Yale. Yeah, that's true. But I think the point kind of is broader than that. Did you see Gavin Newsom's new podcast? Which is now, it's completely normal that it's like, the governor of California has a podcast. And I was like, okay, yeah, sure. Yeah.

But I think to me, like what he's doing with this podcast, and he's the wrong character to be doing it, but politicians are going to have to go into basically the oral exam phase. And that was the sort of knock against Kamala, et cetera. And I think it's going to be broader than that. I think it's with business executives, et cetera.

all kinds of people in that you have to do the oral exam. You have to be able to think on your feet. And that's a different skill set than being really good at being prepared. It's actually like the Bannon and Shapiro principle. They're just like verbal sparring guys. They're terrifying.

I wouldn't want to be in like a debate with one of those guys. No, I'd be chewed up and spit out within the first like five minutes. Yeah. So yeah, obviously they want, I mean, Gavin Newsom had on Charlie Kirk, who I know, I know Alex likes him. And he kept, I fucking hate him. He kept, he kept like, Gavin Newsom kept saying how his teenage son is into Charlie Kirk. Like that is like, you know, like,

It was an interesting way to connect with him, considering, you know, and I thought it ended up pretty softball. And I think that's funny. Like when Gavin Newsom is actually in the hot seat, he's a really good debater. Like when he was on Fox and things like that. When he becomes the host, he kind of,

I'm going to say he still acts like a Democrat. He takes on the rules of being a host and not being too much of a jerk and listening and not interrupting and being the one that asks questions. Well, that's not how these guys do it. I think if you listen to Steve Bannon's show or whatever, they'll full-on make sure that even as the host, they get their points across, not the guest's point.

And I think that's why it's maybe... They're way less encumbered, for sure.

Yeah. Yeah. I mean, he's incumbent because he's a governor and like, it's a different skill set at the end of the day. I mean, I think he's got the skill. I think it's just the rules of the game. Yeah. I don't know. I, I kind of thought it was a good idea. I agree with Brian on this. I, I think that this idea that as Democrats, you take the high ground, you don't engage, you insist on telling, you know, the truth and you worry that, you know, full engagement means platforming someone with toxic ideas. I,

I don't, it just doesn't strike me that you win that way. Get into it.

you know, figure out what your weapons are and just get into the conversation. Like that's the modern sort of digital Socratic way. Get into it. But the thing is, he's not really able to get into it for whatever reason. I mean, this is not like... He did get into it though, Alex. Did you listen to it? I didn't listen to the latest Bannon one. I listened to the Charlie Kirk one and I felt that that was like frustrating because he was just... He was too polite, did you think? I mean, he was full on like...

Yeah, I don't think he was fighting back. If you're listening to somebody on the extreme end of this, if you listen to somebody like Destiny, I don't know if the audience knows who Destiny is. He's all sorts of a character, but this guy is full on.

Like this guy kind of, you know. He's rabid. He's feral. Exactly. So like you're going to get into a bare knuckle fight, you know, go bare knuckles. And the thing is like, I thought Charlie Kirk got Gavin Newsom to say things that he wanted. He got to, you know, say all sorts of, all sorts of things without being really truly refuted. Like Gavin Newsom was being very polite. I didn't hear the Bannon one. That's a lot. That's a lot. Brian makes this point. And there was a fellow who was kind enough to respond to the

last week's PVA weekend newsletter and talked about how we are living in this kind of post Gutenberg world where you know the logic and formality and structure of you know of of a society driven by written media is evaporating into a new world of of that looks more like it did you know

pre the, you know, before the printing press where it is a kind of oral culture and it is about, you know, who can, who can move groups of people most powerfully with their words. And in, in that, you know, truth is fungible and, and, and kind of personality is an important weapon and, you know, audacity is important. And that's what kind of,

Trump does so well at and anyway the guy the guy wrote it all up in our in in the Comment section of the email and did you read that Brian? Yeah, no I thought I was right on but I think that it all goes back to that idea that like this sort of skill set needs to shift like I always think I've always run into Executives that I need to talk to that I know like on a personal level to be thoughtful and interesting and able to like spar with ideas and back and forth and

And then if we go on stage at an event or they show up for some podcast interview because they want to have an interview, not a conversation, and they got a PR team in tow, okay, they're different, right? And that was always the rules of the road is that you needed to be media trained. And media training to me –

It seems to me you've been media trained, right, Troy? Maybe unsuccessfully, but I've never been. Who was the media trainer for you? I've been media trained many times. It's like being potty trained. Yeah, and I still soil myself. I think you can smell it now more than ever because you can see when...

people are pulling the discussion back to talking points. You can just see when they're not answering the question and they're saying, look over here. You see the wires these days. Yeah, you see the wires. I think we always saw the wire. I think it's one of those things that we do as people. I think we always saw the wires and we never felt entertained by it. If you ever saw a corporate speak

like a meeting, it just washed over you. Or an interview. With markets being where they are, I just turned on CNBC and there are so many... They bring some of these CEOs on that just say fucking nothing. And your eyes glaze over. And people need to remember everything's entertaining and if you're not engaging, if you're not charismatic, great, that's fine. But then you shouldn't be on TV. That's it, right? Time is finite. We want to...

we're ready to consume seven hour podcasts if the people are saying interesting things that feel true. But like PR press releases or spoken or written are just like never something people read and they have never read them. Nobody has ever cared about this shit. We haven't changed. It's just people started doing differently and people go like, oh wait, you can do this?

Like there's this great joke in the, in the show party down where somebody like this, this guy who's a caterer goes to this rich person's house and goes like, let's go. I got a karaoke in the back room. He goes like karaoke in your own house. Is that legal? Is that possible? You know, it's just like, oh yeah, you can do this. You can do this and we don't have to endure it. And I think that's what we all opened eyes to. Well,

Well, that's why I think that the whole Silicon Valley go direct thing, which was kind of corny when it first came out, but I think it's accurate, right? And I think a lot of things that happen in tech wash down through the rest of the economy. And basically, go direct is...

Don't outsource your communications to a PR staff. If you're running a company, you're in charge of that. And to me, this is kind of like how a lot of the CEO class there wants to take back power from HR that they believe that they gave away during the DEI era. I think the go direct is a take back power from being overconfident.

overly scripted from being overly managed with talking points. And it used, that used to be a competitive advantage. And I think now it's a liability, right? Like if you can't think on your, on your feet and you can't go into, look, it's corny, all the stuff that people like Charlie Kirk say,

about like he's like long form podcasting is like going into the arena and it's a test of your manhood and stuff like that. That's so ridiculous. These guys got to stop jerking off to the 300 movie. Like it's fine. But that said, I think underneath that there is a good point. I guess I'll change my plans tonight.

It's easy to say that, but it's also like, and it's fresh to me because we just recorded some videos and we, you know, we have a few cuts and in the end, the stuff that's coming out is as stupid and hopefully as real as we are, but it took us a while to get there. It's vulnerable to do this. It's vulnerable to be yourself. It's easy to hide behind MBA speak, you

You know what I mean? Like it's, it's, it's the same way. Like it's, it's, it's hard. It's easy to let chat GPT do your work for you rather than use it as a tool for you. So you can do better work. That's what I'm saying. Maybe we have a theme. Yeah, I know. But I think, I think that a lot of these CEOs and these executives, they're not performers. They might not be extroverts. And, and,

And the way things were done was actually everybody was kind of happy with it. But it's going to be interesting, right? We talk about Apple a lot. Their stuff is so polished. They never say a word out of line. But I think that's starting to show up as brittle. Well, that's where I thought we would go next. The world sits somewhere between. I'm segueing us.

Yeah, between, you know, the Twitter responding to any PR inquiry with a poop emoji to, you know, to Apple on the other side, who's, you know, the polish of it feels gross, actually, to me. It doesn't feel right. Well, it feels, especially now when they literally, I mean, for... They shipped a turd emoji with Apple intelligence. Yeah. I wanted to hear Alex talk about this turd that they shipped.

This turd that they shipped. I'm not sure that I'm qualified to talk about it. Siri? It's a podcast. Yeah. No one ever said you had to be qualified. You talk about media all the time. No, I mean, I think the AI stuff has been, like, right from the beginning, right, I think the ads were pretty tone deaf. It was showing horrible people doing stuff that you know that the AI is not going to be great at doing. The tools are probably one of the first time I ever turned off a feature that

on iOS. Like I said, I'm an early adopter. And the whole thing was that they promised a better Siri. And hopefully we can cut this clip in, a little bit of this clip in from Curb Your Enthusiasm of the guy yelling at Siri. Hilarious. It's so good. Siri, directions to Wolf's Glen. Directions to Great Wolf Lodge.

- No, Siri, Wolf's Glen Restaurant. - One option I see, Woods Garden Supply on Benedict Canyon. - No, no, no, you're not listening. I said Wolf's Glen Restaurant in Westwood. - A wolf's den is a habitat that provides wolves with protection from weather and-- - No, you stupid fucking idiot. - Because that is all of us. And Siri actually feels like it's gotten worse.

And they've delayed it and it's been delayed indefinitely. And the trouble with Apple is that their polished way of communicating is great because when it lands, right, it's just, yes, we're late, but here we are. Here's what we're doing. We're integrating it into your life and it's going to work really well. But this time, not only did they not have the stuff ready, they started advertising stuff that wasn't ready. And the advertising now feels completely misleading and it's

If you look at what Apple intelligence is on the iPhone, it's nothing. And they didn't have to do this. They didn't have to do any of this. They could have just sold the phone. They could have just made us wait another two years. Everybody would have been fine. And they didn't have to boil the ocean. So it was avoidable. Well, you know, it hit me yesterday, Alex, because I was...

I had a computer somewhere. Oh, no, on Shelter Island that I never used. And I turned it on. Wow, tumble brag. And I turned it on. What, having an extra computer? That's a big deal. Yeah, I have a computer I never use on Shelter Island. I use all my computers. I actually only own one computer. You guys, let's move on. I only own one jacket, too. I'm like Bernie Sanders. Yeah, so I turned on, it was an iMac, actually. I turned it on, and it updated.

And the first thing it did after the update is it put me into Apple's new, I don't even know what the software's called, but it's a new little application connected to the OS that allows you to generate images with AI. Image Playground. And it was so stupid and so slow and so nanny. It was like nanny wear. Like it was just like, oh, you can make a goat that looks like an octopus. That's cute. And, you know, it just felt like,

I mean, when you compare that to the feelings that we were talking about of being completely emancipated by connecting LLMs to coding, to writing. You having a conversation, you had a conversation with a computer last week, right? That we put at the end of the podcast. Oh yeah, yeah. Sesame. Yeah. That was incredible. And you compare that as this with this little emoji creator or whatever that thing is.

I just think what it said to me that Apple's got a lot of work to do and I'd be concerned about the company's performance, to be honest. And one of the things, it just struck me thinking about it, that sort of AI is this probabilistic thing, right? So it's

It's kind of naturally wrong a lot of the times. It's very human. And Apple is a very kind of deterministic company where they like, you know, a real kind of tightness in how they launch products and features.

And all of these, and Apple is, if nothing else, I think you would agree, is an interface company. And at the heart of the AI revolution is, you know, completely new thinking per the beginning of this podcast and how we think about interface. This is a real threat, right? And I think that Google is

It's just going to be fundamentally better at it, not just because, you know, they have deep, deep bench in AI. They have real, you know, an ability to manage the search index and connect that to AI in all kinds of great ways because they can launch things.

Android across multiple devices and make AI the heart of it and need to invest in being super good at that, you know, Apple, because of the path they've chosen and because who they are as a company is going to have to let others come closer to the metal of their software and their hardware to create product like open AI, right? To create products that enhance the marketability and desirability of that device. Right.

And they don't want to do that. So you're sort of like half you're half pregnant, right? Like you're not an AI company or you're Trump. Maybe you're trying to be, but like there's so much innovation happening around the periphery that you have to let that in or you have to be great at it. And they're not good at either of those things. And I find that really I'd find that really concerning, to be honest.

Let me ask you a question. I get a sense that people in those companies, if they had a choice, would prefer if AI never released because I think everybody's acting desperately. Even OpenAI is constantly talking about what's coming next, completely distracting us from what you can do today. Microsoft's doing its thing. Google's not sure how they're starting to integrate AI into their search, but

Some research is showing that, hey, surprise, surprise, 98% less people click on links when an AI answer is given to you. 91%, according to Tolbid, although Tolbid does have a dog in this fight. Let's cut it by half. Let's be generous and cut it by half. That's still less than today. Apple is acting desperate. Apple made a lot of desperate decisions, which they didn't have to do. Well, you're starting to hear calls for Tim Cook's head and people calling him Grandpa and such.

I think there is a generational challenge there because when you see the way they've treated gaming and the way they are not trusting developers to try to solve some of this. Vision Pro was a huge fail. Siri's a huge fail. These are big, big chickens. I agree. I agree. And I'm agreeing with you. I'm agreeing with you. John Gruber brought up John Scully. That's never a good sign. Sorry.

I think Tim Cook was the right person for the right time, but it's moving so fast. It requires someone that's not as worried about tarnishing their legacy.

You know what I mean? I think if you're on your final five years and you have to make such big, risky changes that it might kind of make you look bad, I think that's where they are. I'm actually having dinner with some Apple people. I should be more positive. Yeah, you should. They'll be fine. Speaking of truth, man, this is the new world. We're all just jamming. Are they fruits? On that note, Troy, I want to get your take on Google's AI mode.

It's coming. As Alex referenced, obviously, AI in search is

is a looming disaster for a lot of publishers. What do you think the impact of this approach that they're taking now with allowing people to switch on AI mode will be? You know, let's go back to a thing that Professor said that surprised me a little bit. He writes, in fact, I designed his last book cover. He writes these, you know, just academic books that have an audience in the hundreds or maybe low thousands.

political philosophy and stuff like that.

And he said, increasingly, we need to write our books for AIs, not for people. Because most of the people that are going to get information about the arguments and thoughts we create in our books are going to get them abstracted through an AI layer. And I just bring that up because I think that's a profound... Or at least enough for it to matter. It's a profound notion that your audience is automation or a robot or...

you know, some kind of mechanical brain. But does that mean you have to change the way you write, Troy? I mean, I think you have to always be conscious of who the audience is and how, you know, what the economics of that are for sure. And that, I think, takes us back to the media argument. I'm pretty convinced that, you know, we've had a lot of discussions, Brian, about how content needs to change when the friction is taken out of

the places where people used to make money in the connection between search and content. Largely, you know, evergreen content is a good example of that. I think, you know, you see stuff become more human, more personal, more point of view, you know, more sub-stack, more niche, more expert-driven, etc. I'm pretty convinced that someone like ChatGPT, it's funny in kind of consumer-networked or consumer technologies like

like chat to VT, when you get an early lead, and you become the Kleenex of the category, you can usually maintain it. Now they're trying to maintain it on two fronts, right? They're trying to evolve. There was announcements this week around how their API's are going to become more sophisticated, so that, you know, people that build things on top of open AI can do more and more things with their core technology. But really, I think the power is in them being the default for consumers.

And in order to do that, I think the winner has to have a free offering or at least a subsidized offering where advertisers differ the cost to the consumer of the product. And so, you know, they will maintain and be and kind of realize their market value aspirations if they can create a product that everybody uses. And for that, they by necessity have to have a commercial layer connected to it.

And so I don't know, you know, I could speculate where the ad product kind of manifests inside of new interfaces. How do you turn it? I mean, it's, there's an insane amount of economic opportunity inside of a response, but how do you do that? I think the next challenge in advertising is when something is designed to give you the best answer. How do you weave advertising into that when it's not the best answer?

It's an answer. There's no truth. It's just the advertiser's truth and then the AI's truth. So the competition is going to be weird because AI uniquely, if the inference cost, and this is very hypothetical, but if the inference cost can be brought down low enough...

then there is value in just ingesting all that data of all those queries and the conversations with people. So there is a world where potentially one of these players manages to generate value without advertising. I think advertising is going to be really hard, and if it works, it's going to feel really insidious.

Because it's going to just mean that like you're biasing answers towards advertisers, which, you know, sure. I think it's, you know, I think there's a big opportunity for commerce. Like it's such a great tool to find a thing that does a specific thing. Well, maybe Brian makes a good point, Alex, because that's kind of what I was just saying. But maybe there is no best. And maybe you, you know, in a follow-up question, you suggest that someone looks at this electric toothbrush, not that other one.

ChatGPT gives me multiple answers all the time and asks me which one's better.

They don't know. They say which way they put them side by side. But even humans don't really know because it's preference, right? Yeah. It's trade-offs and preference. But that's what I'm saying. Like, I think it's, I don't know, maybe there's a whole new commerce layer to that and maybe it just goes back to affiliate links. But I think chat GPT is actually really good at finding what is best for me. And I think that there could be like a commercial transaction there. Let me give you a very specific example. You know what we do in those cases, Alex? No.

interstitials yeah yeah but i think it's it's not it's not traditional it's not advertising but but i was looking for a mixer right to record music on right now specifically i could look for best mixer and i would get the answers that the internet comes up with but for me i needed a mixer where all the plugs were at the back not at the top because i need to fit it in a tight space

That nobody had ever written that article, but Chad GPT managed to find me the best mixer with the plugs at the back. That will fit in my space. What are you making? What are you making with that mixer? I'm making music.

I'm a musician. Okay. But I mean, you're pretty far down the funnel at that point. You are. You are. But there's no reason why. But I think this is what these tools allow us to do. The funnel gets compressed. The bottom of the funnel is going to probably collapse in many ways with the actual commercial transaction. That makes complete sense. But people are going to still need to understand...

the world around them and what all of these amazing products that capitalism creates are all about. They're going to need that. I'm long on advertising. It's an odd relationship. Right. Imagine, back to Joe's point, if he's writing for the AI. Now, you're the guy that makes the mixer, and you have a guy or a person that's really good at analytics, and

And they figure out that really what people want is having plugs in the back so they can put it in compressed spaces. Then they start creating information that gets fed into the AI, which is the modern, you know, the AI equivalent of search engine optimization.

to make sure that their product features and the virtues of back plugs are fed into the system. Sure, that's already started. Yeah, that kind of commercial optimization flow is going to happen immediately. But the thing is, I do feel like that these AI companies, because you have to think about it, that there's so much competition. OpenAI is not...

you know, is not sitting pretty, right? Like Deep Seek came out, you know, and just shot up to the number one spot for a second. The challenge with- But that's an interesting statement because Deep Seek come and Deep Seek go.

Like, no one's using deep seek this week. Absolutely. That's fine. But if you start injecting advertising or if there's a sense that the content that I'm getting from this machine that I'm talking to, there's a different relationship. When I'm looking at Google, I can...

Even though the way they do it is a little cheeky, I can parse out the fact that there's advertising and there is not advertising and I'm doing that research. When you're having a conversation, you're building a relationship with this chat interface, right? It's going to be really hard to sell me the idea that some of this stuff is going to be paid advertising, right? And in a large competitive space like that, it opens up.

an opportunity for somebody who's running like a cheaper inference or whatever to say like, hey, our answers are not paid for by advertisers. Our answers are true and real. Like this is a competition that Google never had to...

Google never had a competitor that came out and says, look, we're nearly as good as Google, but our results are ranked. They're not pay to play. And it's still free because the data could be worth it. That's going to be really interesting to watch. I don't have an answer, but I don't expect ads...

right at the answer layer to be like an easy solver, one that can even be solved, just that the market supports. Okay. As they say, I want to put a pin on that one because...

I think this one, you're right on a lot on like 9 out of 10, Alex. Hey, man, I said that. This might be the Vision Pro. You know, I was laughed out of the room when I said like, you know, the AI spaceship is over the White House. Or when I said that Elon was crazy. Oh, shit. I knew that was going to come up. By the way, Alex used to be the best designer of my ad products.

Alex and I created some of the most dastardly ad products on the internet, I would argue. Yeah, yeah. I mean, anyone out there has ever been annoyed by a YouTube overlay ad? Thank these two. You're welcome, guys. We made it and got no credit for it. Also, the expanding banners. Oh, the lower third. What did we used to call that? That thing that existed at the bottom of the browser that I despise now. The drawers of your phone?

That never goes away. Yeah, the tray. Yeah. I want to talk about franchise value because I should add to everyone. I should have done this at the top. I keep forgetting. You should sign up if you don't already get the PVA weekend editions at peopleversalgorithms.com. Troy's been very diligent. He's leading. Alex and I have been less diligent. Annoyingly so, I would say. Yeah. I write a lot.

I get it, but you know what? It's not time because I sent it out early this week. I got nothing. Crickets. Alex will send me a text tomorrow. It'll say, where was the link for that again? I sent it like two days ago. He's not gone in there. Crickets.

And, you know. I have to understand what a WhatsApp thread looks like if you don't look at it for two hours. Yeah, it's a bit of a mess. But AI is going to solve for that. Anyway, you have a good piece in here about franchise value. I didn't see this. Where did AB, an anonymous banker, share this media valuation spreadsheet? I didn't get that. I'd like to see that. AB and I have a private thread. Is it a back channel? Yeah. Oh, wow. Yeah.

Well, I want to get into that one, too. So anyway, let's talk about franchise value, because I start to wonder whether franchise value even exists in media anymore. Maybe you should just build these media businesses, you know, just arbitrage it and, you know, have one foot out the back door because, you know, your operations could get shut down. Well, it just struck me as it's a Buffett idea. And he used to describe local newspapers as having

franchise value. Because, you know, if you were, you know, a media consumer in those years, you remember that someone that owned, in my town it was called the Leader Post, and

and you know there was you know you know newspaper delivery kids and you know they would come and collect at your house and it was a absolutely essential part of living in that community and it had franchise value because it had market power it had pricing power had a moat it was essential and i think that other types of media have different types of franchise value like

Something that, you know, Cheers had tremendous franchise value because of its IP and longevity and distinctness. There were obviously lots of programs on network television that had tremendous franchise value. Vogue magazine had tremendous franchise value. And I always, you know, everybody always envied Vogue because...

It had, you know, whether it was realized because of their cost structure or not, it had just like way more profit potential and way more pricing power than the second, third, fourth, fifth competitors in the fashion space. And, you know, Vogue had franchise value and Marie Claire was a survivor.

And there was a big, big difference between the two of them. And so the rarefied world of franchise value is when your kind of media brand or media business gets lift off. And it could happen for reasons that are local monopolies, could happen for reasons of just...

you know, perceived quality or real quality or the quality of the talent, or it could happen for reasons I think of operational complexity and mastery of that app. Right. And so he was showing me the, the banker anonymous banker was showing me kind of valuations in the market and,

And you see up in the high valuation, like 15x plus, like enterprise value, you see at the high levels that there's a couple of media companies in there, like New York Times, which I would argue is tremendous franchise value. But most of the others are kind of platform media hybrids, right? Whether that's kind of Netflix or Amazon or...

I mean, as you move down, you see Disney. And then as you move down further and you start to see kind of old media that doesn't occupy the middle anymore, because the far right to me was always someone that was like an ad network, pure monetization, arbitrage, right? People that just make money in the kind of Byzantine world of digital ecosystems. Right.

And so they trade around like four, five, six, seven X, whereas franchise things can like a, like a Google or Amazon or whatever that have tremendous modes trade, you know, 20 and up.

Right. 15 and up. So you're saying arbitrage, the arbitrage plays have the lowest, like... And it's on EBITDA. It's no longer the magic book. They did, except you're now seeing the middle, what I call the survivor media companies, like would have been maybe, you know, local television like a Sinclair or...

you know, scaled newspaper businesses like a Gannett, they're getting pushed down, you know, into kind of arbitrage territory in many cases worth less as a multiple than arbitrage. And the poster child there is my friend, the VEX company, Ziff, that's trading for like four times forward enterprise value, but multiples, which is really kind of crazy because

company has 30 to 35 margins but you know the market is hammering their stock price either because they think they have i guess tremendous platform risk because of the changing nature of google or you know they they don't understand the complex is a really complex business there's data businesses there's cnet there's so there's you know pc mag there's you know

coupon businesses. So it's a complex story to understand. But it just occurred to me that in the world today, call it media, media tech platform, there's there's a continuum. And on one end, you have things that have kind of franchise value slash market power. And on the other end, you have the kind of arbitrage. And that was really the the genesis of that little blurb. I'll throw it in the email this week.

So just to put like a number on that, Ziff Davis is in the last year, their market cap is down 41%. That's kind of rough. I assume that a lot of that has to come from just, they're so exposed on search.

That's the essence of that business, right? You would think, yeah. So tell me, how do you build franchise value now and is it possible even? Because a lot of the examples you use- From scratch or from people around? Well, yeah, because I feel like it's harder than ever to build franchise value. If you look at the strongest parts- Why? Of the media ecosystem, a lot of them around-

personal brands, niche, and I don't know if you can build as much franchise value. I think it's hard. It's either platforms or personal brands, and personal brands are fleeting. You even see it in... But you do see, I mean, I would argue straight from valuation perspective, Mr. Beast raising with a basically no-profit business at $5 billion, there's got to be franchise value in there. It wouldn't...

Now, whether that's real or not, I don't know. It is interesting that Mr. Beast is like, you know, he has like pretty much an unprofitable media business with a candy business.

So even the new guard apparently are not making money. I mean, yeah, and he's definitely building franchise value. He's going to probably build up other influencers, but it's all the commerce stuff. It's like selling capital. Well, because he's a brand. And that's the thing. It's like these older companies, a lot of people like to give grief to Forbes, right? I've given some grief to Forbes over the years. But look, they've got a really strong brand that still means a lot to a lot of people in a lot of different contexts.

I always said, I was meeting with a magazine executive yesterday, and I was like, yeah, every single magazine company needs to decide how Forbes are going to go. Because there's only a few of them that can be on, I guess you put it on the left side, on the left side of that continuum on the franchise side. And you can go, you can have a direct business, you can be the New Yorker, you can maybe be the Atlantic, but most are going to end up either...

in that middle or most likely you're going to get pushed to the right. It kind of goes back to a point though. I think you can absolutely build franchise value today and maybe for certain people it's easier than ever, right? Like Kardashian, Skims or the folks that did A24. These are companies that have huge franchise value. A24 is a great example.

So, I mean, and these guys just, what do they do? The thing is like, what you cannot do anymore is get a bunch of people in suits together with $200 million and say, now we're going to build something big. That is no longer the case. You need to either have like,

creative vision or really kind of understand culture a certain way or be able to be like such an influencer that you can build brands out of them. This is how it's all starting, right? The fact is the moat is no longer do we have a bunch of money and MBAs? The moat is like are you able to place yourself in the world in a way that stands out so much that you can build franchise value around it and hey, the chocolate is going to outlive Mr. Beast, you know, like

that,

That's what's going to happen. But I think it's at the inception point that all this stuff is changing. Yeah. And we have to get AB back on. They are... Anonymous backer. We're already not saying their name. Our pipeline to regime media. They have ties to regime media. AB wrote about this like the other week about how Newsmax, that deal for Newsmax does not get done without Trump in the White House. We've gotten this whole podcast without mentioning Trump. Yeah.

But it'll be interesting to see how you can build franchise value if you go on the regime media side. Because you know what? Nothing is forever. But then I guess you just go into the opposition. That's the way to do it. Yeah. Should we make an elegant transition to a good product? Yes, that sounds great. Okay.

I have two things on my mind. I always start it some way, like something like that, but should I get the easy one out of the, it's just that I'll start with the, no, no. Okay. We can cut that out for sure. What a tease. Yeah. Yeah. No, no, no. It's just that I mentioned this video that kind of changed. I want to get a little more specific about why I saw this guy annotating his workflows, doing a

a range of things from making throwaway software to making little graphic things that he would use in his social media to annotating a video to having the AI coach him and making a presentation at TED. And I just started thinking about how

Crazy this was and then the notion of vibe coding was was launched by this guy Andre Carpath Carpath II and then you know There's a funny meme going around because people have turned Rick Rubin into a meme for vibe coding Because Rick Rubin loves music and makes music but doesn't play instruments and the notion of live coding is that you can make code

you know, without being a coder, right? And the quote from, I don't know whether this was made up or not, tools will come, tools will go, only the vibe coder remains. It's like Rick Rubin, right? And so I was like, well, what does it mean? And I started to bullet it out. And I would really encourage you to watch. We'll put a couple links in PVA Weekend that you can subscribe to.

or download, or maybe we can put them in the show notes. There was also a video of the Y Combinator people, the smart Y Combinator people talking about vibe coding. But I was just thinking about it because it's big in my household because my son is artistic or is a musician and he is like deep, deep in this stuff. He wants to be a coder. He's deeply inspired by it.

So the things that I would say about it, four or five things to consider. One is art encoding or merging and that a new technical literacy is requirement in all creative disciplines, whether you're Brian Morrissey or Alex Schleifer or my son. Software applications increasingly become fit for purpose. We make tools as we need them.

And I think that that's at a micro level, like creating a little tool to help you make set lists for a shooting schedule for a video to SAS tools that annoy you, Brian, like the stuff that you want to do, like you can make that stuff increasingly. I think this is really, I'll get to the point of where, where I think this leads us in a second, right? Coders don't need to spend nearly as much time coding. Like I'm hearing stats where like 80, 90% of code is written by AI, right?

So they become product people by necessity or systems designers, right? Because they're not there writing the code. So they have to get closer to the consumer or closer to the kind of meta framework, right?

I think that it changes, like I said, how we think about SaaS and packaged software. I think the way we string tools together is in total flux and a new industry of sort of tool and productivity gurus emerge. People like the guy that I put in that video, the video I think you should watch. I think back to the professor, we have to rethink education, role definition, and career paths. These are huge things.

Technical moats disappear, right? You can't just win by technology, right? Taste, experience, and attention, as everybody always talks about, growing importance as differentiators. And finally, this is what struck me. This was my visceral feeling in watching this video, is we begin to think about our world differently as an ever-evolving kind of virtual and physical stew that we collectively create.

And I just feel like it's a real transition point in, in, in everything from kind of how the winners win and, and how we make things as creative, you know, beings to,

And, you know, what we teach the next generation and what will be required of them to make their way in this world. So that, I don't know if that's a good product. I guess that's a bit of a speech. It's not almonds. I mean, this is a different. It's not almonds. It's not almonds. And if you want to get into almonds, I used to sit, Brian, on my, you know, in many households, there is a bifurcation of TV watching between the father and the mother.

I don't know if this happened. It's happened in my household. I'm embarrassed to say. I have my zone and Jillian has her zone. They meet occasionally. In my house growing up, my dad would watch one thing and my mother would watch TV in the bedroom. And I would lay on the bed with her and watch Carson. And so for me, talk shows were a warm memory. I love live television talk shows. And so John Mulaney's out with his new one on Netflix. Yeah.

And it's called Everybody's Live. And it's weird. And it doesn't quite work yet. But this week, like live on Netflix, I find very interesting the concept. This week he had on Michael Keaton, Joan Baez, a writer named Jessica Roy, who I don't know, and Fred Armiston.

And, you know, he's got this co-host that, that weirdo, Richard kind. Do you know who that is? Yeah, he's great. He's, he's the voice of bing bong. Yeah. And they, it's, it's totally unscripted. There's a theme. This, this show's theme was finance. I mean, it's a talk show, but it's like a theme, like a podcast, right? But you got to change the format. I mean, it's, I think it's, it's right there. They had Tracy. What's his face from SNL as King Latifah who died in the audience. Yeah.

And then they had Cypress Hill as a musical guest. It's a strangest show. It's really, really awkward. You know, there's Alex, you'll love this. One thing I loved about it is they, they do these little interludes, these shots from LA and they just have this shot of an abandoned sideways hot tub on the side of the highway.

That's amazing. And it's just like so great, right? You'd never see that on any other kind of produced network television. So what is the live element? It's live. It's live. I know, but how does that add to it? Unpredictable. People say weird shit. Because I do wonder whether live will come back. Cypress Hill is playing in Jort and Richard Kind and Joan Baez are dancing to Cypress Hill. Like weird shit happens. I think live is a...

different experience knowing that it's live is a completely different experience and and it's where i think like you know where youtube and tiktok are interesting oftentimes is that live stuff like go on live tiktok and then just like you could experience it completely differently because you don't know where it's going live tiktok is a terrifying place i will tell you it's fucking insane i've never been it's bizarre it is insane i mean if you want to get in touch with the culture go in there

just explore it yeah I think we should see more live stuff I'm glad that binging is over it seems like binging is over

Well, I mean, because the problem with binging is that it wasn't creating these cultural moments, right? Severance has this thing that where every week, you know, they're dropping an episode and a lot of people talk about it. This thing? You mean the way it's always been? Yeah, no, absolutely. I mean, I think people are just adjusting, but at the same time... I like it. I'm going to vote for it. Binging?

Anti. How good. Yeah. When Netflix dropped the whole season all at once, they became the talk of the town. And, you know, everything ebbs and flows. You can't just stay static.

I love that White Lotus comes on once a week. I never liked it from the beginning. That's the thing. If you take the Andy Rooney thing, you're going to be right 100% of the time on new stuff that doesn't pan out. This world is not for you. You read books. I do. You read books. You only have one computer. I was talking to one computer and I read books, but I failed because I've been trying to slog through, which is not even a long book, that Chris Hayes attention book.

That's summarized. I keep losing focus, which I think is kind of funny. But then I went to visit my parents and I left it there because I didn't.

I'm sure they've devoured it. I lost my attention. Yeah. But are you watching White Lotus, Brian? Yeah, I am. And that's perfect. That's like once a week. It's a vibe. It's perfect. It's perfect. And that's all I need. And people can complain about it not being as good as the previous season. I'm like, I don't care.

like i don't it's not a piece of art it's a piece of art it's a piece of art it's a piece of it come on it's fine that it's different and okay well maybe we can bring that up in a later episode let's uh let's wrap it up i've got a hard stop

Okay, Alex has a hard stop as always. Everyone check out peopleversealgorithms.com for I'm going to be submitting my response. It's an interesting format. We're still working on it. Yeah, are you going to get in there? Are you going to get in there, Alex, and make some comments? I always call you out and then I'm like, oh, I guess I better edit that because he never came by. What is it called? What is it again? Stop it. Do not ask for the link again. Put in a voice memo from Alex. I think that would be better. I like that. Yeah, put in a voice memo.

He's Whisper or whatever that is. All right. Take it easy. See you guys. Bye. Bye. Bye. Bye.