If AI can do the work, what should we learn?

[Speaker 2] each other. Seventy years ago a huge tornado came down on Seattle that wiped [Speaker 2] out an entire industry overnight. This industry was gas industry and when [Speaker 2] pipelines started coming in producing gas from the north it wiped out how gas [Speaker 1] was traditionally produced in Seattle. Yes so we're here in Gaswork Parks here [Speaker 1] in Seattle, one of the most beautiful parks that we have in the city and this [Speaker 1] is actually probably one of the few, I think it's the only gas work plant in [Speaker 1] world. I'll tell you how gasworks work. Actually, here's a little bit redundant there, but it's a [Speaker 1] fascinating story because I didn't know gas was actually produced or actually generated from coal, [Speaker 1] and this is actually how they did it. So we're going to show you a little bit more about this, [Speaker 1] but the reason we're here is because, hey, we are in the eye of the tornado, and we like to discuss [Speaker 1] what happens after this tornado comes in and the beautiful things that we're going to see as a [Speaker 1] results. So today we're going to talk about the future of work, the future of [Speaker 1] education, even the future of content and we think this is a beautiful place to [Speaker 1] actually discuss transformative technology like this. So hi I'm Sacha. I'm Shravni [Speaker 1] Wati. And this is the Eye of the Tornado. At least that intro we can keep it. Yeah, no we [Speaker 2] can keep it, we can keep going. Okay so yeah so we are here in Gaswork Park. This [Speaker 1] It's a beautiful, beautiful park here right in front of Seattle.

[Speaker 1] Let me turn around here. [Speaker 1] You can see downtown Seattle. [Speaker 1] Everyone comes here to see all the fireworks. [Speaker 1] Families come here all year long to walk around with their kids. [Speaker 1] It's a beautiful place that they're actually planning to completely remove after they shut it down in the 1950s. [Speaker 1] And it's great. [Speaker 1] I think it's actually one of the few, I think it's the only gas work plant in the world right now that is still standing. [Speaker 1] which is a result of this incredible technology that came out of the American Industrial Revolution in the early 1900s.

[Speaker 1] So it is fascinating that we still have this as a good memory of what technology was like. [Speaker 1] And we can look back and see how, you know, tornadoes can actually transform into beautiful things like this. [Speaker 1] So we hope that you like this. [Speaker 1] We actually have now even a seaplane flying us through here. [Speaker 1] the famous Seattle seaplains [Speaker 1] oh look at that [Speaker 1] I don't know if you can see it [Speaker 1] probably it's right there [Speaker 1] yeah so [Speaker 2] we're going to be talking about [Speaker 1] some really interesting topics [Speaker 2] the first one [Speaker 2] I think we can [Speaker 2] start talking about [Speaker 2] is [Speaker 2] there's this [Speaker 2] interesting video [Speaker 1] or interview [Speaker 1] that was done with Jensen [Speaker 1] I know Jensen is very [Speaker 1] very [Speaker 1] very prominent right now [Speaker 1] in the news [Speaker 1] because all the things he's saying [Speaker 1] AI and the future of AI.

[Speaker 1] But one thing he mentioned that got a lot of people off guard and did not expect him to [Speaker 1] say that, was saying that, hey, he believed that, I mean, they were talking about education [Speaker 1] and he was saying like, you know, kids nowadays struggle with math, with reading. [Speaker 1] And he was saying, why, I mean, why do we need math or reading? [Speaker 1] I mean, do we even need that anymore? [Speaker 1] And I thought it was a very interesting topic to discuss because we all have assumed that [Speaker 1] that is something that is not negotiable. [Speaker 1] Like kids should learn math.

[Speaker 1] They should learn reading. [Speaker 1] They should learn certain things. [Speaker 1] But he's questioning something very fundamental. [Speaker 1] Like what is, what do we need to learn in this new era? [Speaker 1] Do we need to still learn stuff from books? [Speaker 1] Do we still need to go to college? [Speaker 1] Do we still need to get the certifications? [Speaker 1] go to school for? That's a good question. What do you think, Shreban? I think there's still a lot of [Speaker 2] value in just understanding a little bit on where, understanding some of the history of our thinking [Speaker 2] and scientific thinking, a little bit on the concepts, the critical thinking that gets built [Speaker 2] up only through knowledge and also a little bit through practice. There is still a lot of value [Speaker 2] that is to be gotten from education.

[Speaker 2] What is a bigger trick, I mean, what is trickier [Speaker 2] for the education industry going forward [Speaker 2] is how would teachers and universities [Speaker 2] standardize tests going forward, right? [Speaker 2] That's like big thing because people, [Speaker 2] kids these days or anyone these days [Speaker 2] would most likely use AI in their tests [Speaker 2] if they're connected.

[Speaker 2] So how would universities understand [Speaker 2] where the student is standing [Speaker 2] testing, which is, I think, in some, in at least like, for me, personally speaking, it's such a [Speaker 2] core part of my identity, right? And it forms, so many years go into just the standardized test [Speaker 2] preparation, etc. And a lot of value is also placed on the education that we get thereafter. [Speaker 1] So, I mean, does it, what makes me wonder about this is, I mean, you mentioned something that is [Speaker 1] very important in our current society, which is certification or tests. And I think that's [Speaker 1] something that we have assumed as that's not going to change. We still need to get our titles. We [Speaker 1] still need to get our certifications. Our kids go to school and they get their achievements and [Speaker 1] their awards and their approval that actually standardizes. And that's the problem, I think, [Speaker 1] is the standardization, right? The fact that we try to standardize an education process [Speaker 1] that we should maybe question right now, [Speaker 1] should this be standard?

[Speaker 1] Should everyone learn the same way? [Speaker 1] Should everyone learn the same things [Speaker 1] and have the same curriculum? [Speaker 1] I mean, of course, once you go to university, [Speaker 1] we want to diverge from there, [Speaker 1] but pretty much still, you're almost 18, [Speaker 1] everyone kind of receives similar education. [Speaker 1] And I do feel that education might have to need to change.

[Speaker 1] And we might not have to think about titles anymore. [Speaker 1] If what happens, I mean, let's say, I don't know, let's do a mental exercise. [Speaker 1] Like if titles were not necessary for survival to get a job, do you think education would be fundamentally different? [Speaker 1] Do you feel like if you don't need to have a title from Harvard or MIT or any kind of university or any kind of other institution to get a job potentially? [Speaker 2] Yeah. [Speaker 1] do you think we would think education completely different? [Speaker 2] Part of the reason behind education is, of course, you're acquiring the skills, [Speaker 2] you're gaining the knowledge sufficient enough, hopefully, to get a job ultimately.

[Speaker 2] And the idea there is that when you're applying for a job, [Speaker 2] that's also kind of standardized testing in some sense, right? [Speaker 2] So people are actually seeing what your experience is, what your education was, [Speaker 2] did you have any background, et cetera. [Speaker 2] And then they're able to select from a pool of candidates. [Speaker 2] That's just how it works. [Speaker 2] And I think you touched on a lot of topics here, right? [Speaker 2] One is like personalized learning is everybody learns differently, right?

[Speaker 2] But should everybody be tested differently based on what they're learning? [Speaker 2] That I'm not sure about. [Speaker 2] That is a question that... [Speaker 1] Because for me, I guess what I struggle is that testing part. [Speaker 1] I don't like this. [Speaker 1] I never liked tests. [Speaker 1] I hated tests because I felt like it never captured what I truly knew [Speaker 1] new it just and it has to be always a chance maybe there's some part whatever reason you're blanked [Speaker 1] out in one part of the test and you're like okay yeah uh just for this thing that today i forgot [Speaker 1] now all of a sudden this has affected my entire life for me that just just like doesn't make any [Speaker 1] sense but i i want to tell you a story that i was um i was listening uh from one of the i think it's [Speaker 1] one of the leading engineers in in grogbot this kind of new tool from uh xai team and they were [Speaker 1] talking about how they build this thing. Crockpot right now is, if you're not [Speaker 1] hearing it, please go try it out because it's pretty phenomenal agentic platform [Speaker 1] that has made agents very easy to use for anyone. You can install in your [Speaker 1] machine and you can have it running there. It has some virtual [Speaker 1] machine. It can do all sorts of things on its own and it's pretty autonomous.

[Speaker 1] But one of the things that they noticed in XAI, which is a company that [Speaker 1] actually it's actually cursor team which is now part of XAI is that the users that were actually [Speaker 1] using the most within their company was the HR team actually let's walk over there see if we can [Speaker 1] see anything over there so this HR team used to use this crockbot but in a very interesting way [Speaker 1] They didn't go use GrokBot to scan resumes or to get incoming kind of requests for jobs, but they did actually a different approach. [Speaker 1] They actually went out and scanned different forums, locations, researchers that were out in the public that didn't apply for any job in Cursor or XAI, [Speaker 1] but had done work that were really, really relevant for the work they were doing. [Speaker 1] actually headhunting for people. So notice a big difference there. And I think this is [Speaker 1] we're going to see a huge difference on how work is going to be done in the future.

[Speaker 1] Sourced. [Speaker 1] Or sourced, yeah. And it's that probably we're going to have AIs finding the people, right? [Speaker 1] Yeah. [Speaker 1] And finding the people means they need to find you based on what you have made publicly. [Speaker 1] and if that's the case what does that mean right I mean does the title mean anymore no it doesn't [Speaker 1] it said have you written a very interesting paper about something have you created content [Speaker 1] that is really relevant for the company that wants to hire you yeah have you demonstrated [Speaker 1] publicly that you've done a difference that actually is relevant for that company and I [Speaker 1] think that that is going to be the new resume right your public persona your um your your [Speaker 1] social branding right right that is gonna be the new restaurant I mean I [Speaker 1] have my kid always say I think all the kids nowadays when you ask them what do [Speaker 1] you want to be they always tell you you want to be youtubers right yeah that is [Speaker 1] crazy but we think it's crazy but it also feels like well what is not life is [Speaker 1] not crazy what if that is they understand something very fundamental [Speaker 1] that their social presence is what are going to determine their success in the [Speaker 1] future and and it's and everyone can craft that presence right in any way [Speaker 1] They don't have to share their entire lives.

[Speaker 1] It's not like we all kind of like Kardashians all of a sudden. [Speaker 1] But it's more like how do you present yourself out there, right? [Speaker 2] I think I want to break it. [Speaker 2] I want to go back to the education piece just very quickly [Speaker 2] and say that standard university education used to be a composite of like knowledge, [Speaker 2] skill building, networking, and even like personality building a little bit, right? [Speaker 2] You really discover who you are at colleges. [Speaker 2] Maybe the composition of these elements will change going forward [Speaker 2] getting commoditized really quickly just as fast as like intelligence is also [Speaker 2] getting commoditized right maybe universities will help people build [Speaker 2] those profiles that will help them be ready for more work opportunities or [Speaker 2] better work opportunities right so I think that's where I see at least even [Speaker 2] the four degree college degree or even four-year college degree or even [Speaker 2] master's degree they I don't see them going away more moving into more [Speaker 2] is where I see them going.

[Speaker 1] This is something that I said for many years, [Speaker 1] that I do feel degrees, as we know them today, [Speaker 1] are going to disappear as we know them today. [Speaker 2] Yeah. [Speaker 1] I mean, I do agree that definitely [Speaker 1] what is going to be more valuable is experience, [Speaker 1] more than knowledge that you can read in a book. [Speaker 2] Yeah.

[Speaker 1] I think because AI essentially, if it's in a book, [Speaker 1] AI is going to be able to do it way better than you. [Speaker 2] Yep. [Speaker 1] So it's not about that anymore. [Speaker 1] It's more about like, how do you connect the dots? [Speaker 1] And I think one skill that I think is going to be essential in the future is the ability to connect dots. [Speaker 1] And the more diverse they are and the less standardized they are, the better. [Speaker 1] Because if you have a, I mean, essentially you have a set of knowledge that everyone has the same, you're not going to be able to connect dots that are noble for an AI.

[Speaker 2] Yeah. [Speaker 1] If you want to differentiate yourself, it has to be a matter that you have enough set of diverse thoughts that you can connect that you can actually mix them together in ways that are noble. [Speaker 1] And I do feel like we are heading in an era where creativity and being able to think out of the box are going to be super important. [Speaker 1] You don't want to have a training on something that the AI is well trained on. [Speaker 1] So I do feel like maybe university is more like you say, maybe a social experience. [Speaker 1] degree from MIT, but hey, you go to this place that is known to foster collaboration, that [Speaker 1] is known to foster good connections with people, that is known to lead you to divergent thinking [Speaker 1] in ways that is not in the book, right?

[Speaker 2] Yep, yep, yep. [Speaker 1] And the instructors or teachers there are there to encourage you or guide you, hey, [Speaker 1] have you thought about it this way? [Speaker 1] But not to tell you, hey, there's this test that you need to do and prove, right? [Speaker 1] Exactly. [Speaker 1] Probably your reputation, your score at the end of the day is have you built something impactful, which is in your social resume, something that you actually built or done that actually can be persisted in this kind of ledger, let's say like that, that can actually persist over time. [Speaker 1] That's what AI is going to search for, right? [Speaker 1] They're not going to search for your resume that everyone has.

[Speaker 1] They're going to search for the unique things that you've done that nobody else has done. [Speaker 2] Right, but this HR team in GrokBot, for instance, they're very specific. [Speaker 2] Maybe they're an early adopter of the new way of discovering people. [Speaker 2] We don't know. [Speaker 2] But much of the world is not aware that they're being scouted online. [Speaker 2] So how do people even make their presence online? [Speaker 2] What part of their identity should be shown online, even to agents, for instance? [Speaker 2] Or how do we think about that?

[Speaker 1] That's an interesting topic. [Speaker 1] As a parent, I think we're both parents and we have all our kids. [Speaker 1] They all want to be YouTubers. [Speaker 1] They all want to have online presence. [Speaker 1] What does that mean for the new generations, right? [Speaker 2] Yeah. [Speaker 1] That is a very tricky one, but I think it's important. [Speaker 1] I think we're going to see a new rise of content creation, and we'll talk about that in a bit.

[Speaker 1] But I think we need to get comfortable sharing what are we unique at. [Speaker 1] Everyone in the world has a unique talent. [Speaker 1] Everyone has their own talent, their own background, their own history. [Speaker 1] We all need to double down on what makes us unique. [Speaker 1] And I think this is something that we need to encourage this individual kind of capabilities that we all have [Speaker 1] and stop thinking about it in standardized ways. [Speaker 1] By the way, we just got here. [Speaker 1] This is a very fascinating place here in Gas World Park where you see all the machines.

[Speaker 1] they actually have so in order to produce gas they actually had to put all this coal burn it [Speaker 1] for a long time and they had to actually remove the oxygen so you have to be very low percentages [Speaker 1] of oxygen so there was this massive pumps that you see here this really fascinating machinery [Speaker 1] that still of course not useful any of that but it still persists here as almost as a piece of [Speaker 1] art I would say that is fascinating that still this is preserved such a beautiful piece right [Speaker 1] yeah yeah yeah so um so so we talk about education we kind of tangentially talk about [Speaker 1] work and I think we previous episodes we talk about that right yeah so what are your thoughts [Speaker 1] on the future of work?

[Speaker 2] Well, I think more and more we're seeing across all industries [Speaker 2] that tasks are getting abstracted away. [Speaker 2] You can think coding has been a huge one lately. [Speaker 2] And I think even education, like learning, [Speaker 2] we're seeing other trends there as well. [Speaker 2] All sorts of industries, we're just going to see more and more AI permeate, [Speaker 2] But not to replace human judgment. I don't think it's going to ultimately replace one human judgment, second human ownership. [Speaker 2] Because we're looking at intrinsically, it's a non-deterministic software.

[Speaker 2] I think humans still will have to know what's going on in the job, on their job, or they'll have to have the knowledge. [Speaker 2] They'll have to own, they'll have to take ownership. Like nobody will take ownership of your agent. [Speaker 2] Your agent cannot just push a feature and then you'll be like, oh, I don't know who did that, right? [Speaker 2] It's still your agent. [Speaker 2] You need to take ownership of that. [Speaker 2] So. [Speaker 1] That's interesting. [Speaker 1] Like who takes ownership of the results of this AI?

[Speaker 1] That is something that is a lot of debate. [Speaker 2] Yeah. [Speaker 1] About this. [Speaker 1] And I think we, I mean, there is, I mean, when we talk about technology, we rely on technology all day long, right? [Speaker 1] We rely on our cars working well. [Speaker 1] We rely on technology that we assume is going to work. [Speaker 1] But the thing is accidents happen and they're going to happen and then we turn to learn [Speaker 1] from those to try to improve over time.

[Speaker 1] And I think that's why also, and I think this is going to be an important factor in the future, [Speaker 1] there is going to be a whole new space in the insurance world probably to how do we [Speaker 1] ensure, hey, if I run this AI model and something happens, how do I ensure that if something [Speaker 1] breaks or something, how do I, who takes care of pain or if there's any damage done, right? [Speaker 1] And they're probably going to see a new industry of ensuring this AI models in the future. [Speaker 1] So I feel like, you know. [Speaker 2] Do you also see humans giving away their comprehension or their understanding of their expertise, right? [Speaker 2] For instance, in the coding world, I don't think anybody's coding anymore, but people are still responsible to know how things are working.

[Speaker 2] Do you still see in a lot of industries even that comprehension going away or is it going to be abstracted to a higher level? [Speaker 2] How do you think about that? [Speaker 1] I mean, I can tell you, I never, I mean, back in university, I learned about compilers, about micro instructions at the CPU level. [Speaker 1] I understand how they work. [Speaker 1] I never will dare to, for most of the scenarios, to write assembly code or micro instructions for a CPU. [Speaker 1] Get it. [Speaker 1] That, for me, was all abstracted away. [Speaker 1] I'm not concerned.

[Speaker 1] I'm not worried about that anymore. [Speaker 1] I'm not abstracted with a higher language. [Speaker 1] or Python. [Speaker 1] And there's new languages that are very similar to English. [Speaker 1] And when I say today we use that, I guess the new language is English, right? [Speaker 1] Like a new programming language has just turned out to be English. [Speaker 1] So I think we've been abstracting technology for many, many years. [Speaker 1] I'm pretty sure all the citizens in Seattle back in the 1900s, nobody knew how these machines [Speaker 1] worked.

[Speaker 1] They just worked, right? [Speaker 1] And it's just a very unique thing that over time we rely more and more on machines. [Speaker 1] And it's okay that sometimes we don't understand sometimes how they work. [Speaker 1] At least not the vast majority of people understand it. [Speaker 1] Some people might have to go dive deeper into how it works. [Speaker 1] But I think if one thing that we've seen, especially in the last weeks, [Speaker 1] is that actually AIs are pretty good even at doing root cause analysis. [Speaker 1] even better than humans, right?

[Speaker 1] We sometimes, when there's a failure, [Speaker 1] I've seen it many times, [Speaker 1] when something breaks, [Speaker 1] in many occasions, [Speaker 1] you have to spend months [Speaker 1] to try to figure out why that happened, [Speaker 1] try to, you know, decompile code [Speaker 1] to try to find, like, [Speaker 1] the instruction that caused the problem.

[Speaker 1] And it took a lot of painstakingly [Speaker 1] kind of processes. [Speaker 1] Now you can get an AI [Speaker 1] to get you the best analysis, [Speaker 1] the best report, [Speaker 1] so you actually have even better tools [Speaker 1] to understand what's going on [Speaker 1] and improve.

[Speaker 1] think this is something that is very different that we've seen um i think a lot of people are [Speaker 1] talking about recursive self-improvement rsi and i think this is something that we need to understand [Speaker 1] that that's what implications that has for education and work because a system that is [Speaker 1] recursively self-improving is meant to fix itself uh it's meant and i've seen this from the early [Speaker 1] days with open claw and all these agents that's yeah they fail but they also learn really fast how [Speaker 1] to fix themselves yeah and yes we're gonna see issues but we're also gonna see the systems able [Speaker 1] to be able to self-improve like we never seen before in the past it's a i don't know a bicycle [Speaker 1] broke or a car broke yeah you gotta take it back to the mechanic it not did not self-improve [Speaker 1] this is something very different on these systems they are self-improving systems and right now we [Speaker 1] see many of that yeah but now this agents like you know um open claw grogbot and all the systems [Speaker 1] are building uh recursive self-improvement built in and it's fascinating when you see it right when [Speaker 1] you see it actually improving itself when you have a couple systems that i built that all of a sudden [Speaker 1] i i let it run for a couple weeks all of a sudden come back and say wow you actually did a lot of [Speaker 1] things when i was out and you actually improved quite a bit by yourself yeah and i think this is [Speaker 1] a whole different paradigm right on how do we build systems do you think the non-deterministic [Speaker 2] nature of this particular technology uh has any impact on um on the human kind of like awareness [Speaker 2] or how much they should know about for instance if you talk about the uh compressor here uh right [Speaker 2] they'll they'll there are certain valves that determine how much pressure is to be created or [Speaker 2] released or something. So they are deterministic. They work in one direction and they work that way [Speaker 2] all the time, time after time, until they break, right? So which means whoever is operating it [Speaker 2] only needs to know when the signals that they need to realize when it breaks. But in our case [Speaker 2] here with AI, that's not the case, right? We're looking at non-deterministic. How much do you [Speaker 2] think that will impact overall? Or do you think the recursive self-improvement will take care of [Speaker 2] the unpredictability that these systems bring.

[Speaker 1] That is my guess is going to be the big difference, right? [Speaker 1] Deterministic systems do not self-improve [Speaker 1] because they're deterministic. [Speaker 1] And this is the big fundamental shift. [Speaker 1] It's like what happens when the systems self-improve? [Speaker 1] And that's the new thing. [Speaker 1] We'll have to see what that means over time. [Speaker 1] I've seen it with my own eyes and it's mind-blowing [Speaker 1] when you see a system overnight run on its own [Speaker 1] and evolve and get better over time [Speaker 1] to like make it do on its own.

[Speaker 1] So I just feel like it's just a new way of thinking [Speaker 1] that we have to see what happens with that. [Speaker 1] And, but I'm very optimistic. [Speaker 1] I mean, at least seeing the results [Speaker 1] from first experience, [Speaker 1] I feel like this is just going to get better. [Speaker 2] And it's going to be like applicable [Speaker 2] across all industries.

[Speaker 2] So people can just dive in [Speaker 2] and start building all of these workflows [Speaker 2] that can offload some of their mental load today [Speaker 2] and then try to improve over a period of time. [Speaker 1] Yeah. [Speaker 2] Everybody needs just a starting point at this point. [Speaker 1] Everyone should definitely start using these agentic tools that self-improve. [Speaker 1] Because once you try it, you can never look back.

[Speaker 1] It's a completely revolutionary thing. [Speaker 2] What do you think this stands, like if everybody's starting to use these things? [Speaker 2] For instance, like we talked about YouTube, right? [Speaker 2] How does it go if everybody is starting to use these tools and start putting out more and more content? [Speaker 1] Content is going to be an interesting one, right? [Speaker 2] Yeah. [Speaker 1] I think here we're creating content, but I kind of question like this week, actually, we both had this experience with Opus 5.5. [Speaker 1] This is one of the latest models from Anthropic.

[Speaker 2] Yeah. [Speaker 1] I think everyone knows probably OpenAI. [Speaker 2] Yeah. [Speaker 1] Maybe not many people know about Anthropic, but Anthropic is probably one of the leading AI labs in the world right now. [Speaker 1] And has released just this model called Opus 55, which everyone is raving about. [Speaker 1] And for a very important reason, it's actually for content creation, it's quite spectacular. [Speaker 2] Yeah. [Speaker 1] One thing that it does is, and actually you probably, if you saw last week's podcast, [Speaker 1] all the subtitles, all the transitions, animations, [Speaker 1] we're all done with Opus 5.5 with just one prompt.

[Speaker 1] I just thought, hey, have this video. [Speaker 1] This is what we talk about. [Speaker 1] Just make it a more engaging video, right? [Speaker 1] With more interesting details and clarifications. [Speaker 1] We happen to have some whiteboards in the back. [Speaker 1] I didn't actually have to tell him. [Speaker 1] He found that there was some whiteboards [Speaker 1] and he started saying, oh, that's a great place [Speaker 1] to actually sketch some ideas as you're speaking.

[Speaker 1] It was fascinating, the results. [Speaker 1] It took maybe 20, 30 minutes to produce that whole video. [Speaker 1] And it created all the subtitles and everything on its own. [Speaker 1] This is very basic, right? [Speaker 1] But what we've seen out there, and we can definitely share some links on the description of the video. [Speaker 1] There's all sorts of motion design videos that people are creating. [Speaker 1] and it's really fascinating because this is a whole industry that we thought hey well yeah it's [Speaker 1] gonna take a while until it does proper motion design and motion graphics but now it is [Speaker 1] fascinating anyone can do motion graphics any idea that you have in your mind you can transform it [Speaker 1] into a video and visualize and put it out in the world but that made me think something very [Speaker 1] fundamental is well now the barrier of entry to produce content has dramatically go down like [Speaker 2] Yeah, I think about it a little bit differently.

[Speaker 1] So we can go down the stairs here? [Speaker 2] Yep. Particularly two things come to mind. [Speaker 2] The barrier of entry for any content has gone down. [Speaker 2] The barrier of entry for great content has actually gone up. [Speaker 2] What is great content? [Speaker 2] Great content, content that resonates with your audience. [Speaker 1] Resonating, that's tricky, right? Because resonating, you're talking about... [Speaker 1] Reception also.

[Speaker 1] But you can do clickbaits and that resonates really, really well. [Speaker 1] I don't think it's, I mean, if you want to maximize attention, [Speaker 1] and that's a problem I think with content today, [Speaker 1] which is, hey, which engages the most, [Speaker 1] it will lead to maximizing clickbaits. [Speaker 2] Yeah. [Speaker 2] And that's the wave that we're in right now, right? [Speaker 2] We see a lot of AI slot videos that get generated.

[Speaker 1] Yeah. [Speaker 2] This is just the first wave. [Speaker 2] But I think after this is when we'll see some kind of convergence coming in, [Speaker 2] wherein we'll see really great quality outputs, [Speaker 2] AI, but ones that show original human thinking, which is again a little bit difficult to come [Speaker 2] up with because AI tends to come up with repetitive stuff most of the times, right? [Speaker 1] Yeah, anyone that's watched the movie, I'm going to go back in a second, but it's like [Speaker 1] this scene from here.

[Speaker 1] Anyone probably, there's this famous movie from Tom Hanks from Sleepless in Seattle, [Speaker 1] and there's these famous houses here. [Speaker 1] Let me see if I can show them here. [Speaker 1] You see them here. [Speaker 1] those are the floating houses here in in lake union very famous from the sleepless in seattle [Speaker 1] movie so you ever come to seattle definitely should check this out because there's some [Speaker 1] beautiful floating houses there that actually are quite nice here yeah really love it this [Speaker 1] you know we have such a nice weather today right like this is can you believe like this used to be [Speaker 1] the most toxic contaminated place in Seattle.

[Speaker 1] Like imagine like they were burning coal day and night. [Speaker 1] They even made some experiments [Speaker 1] where they actually dug a hole [Speaker 1] and there was so much contamination deep inside the ground [Speaker 1] that they were like, [Speaker 1] okay, we gotta shut down this whole thing out. [Speaker 1] We gotta close this thing [Speaker 1] because the level of contamination was insane.

[Speaker 1] And look, this like beautiful park that we were walking in [Speaker 1] where people come in, bike, they bring their kids. [Speaker 1] They literally build a playground. [Speaker 1] You probably saw when we were walking there [Speaker 1] all the machines were right so i think it's uh it is really fascinating you know how you know [Speaker 1] technology comes and goes and all of a sudden you see complete transformative uh alternatives that [Speaker 1] come out of out of this but yeah going back to the content the human ingenuity behind great content [Speaker 1] yes i i feel goes back to where we're talking about dots right yeah um i know i think probably [Speaker 1] we're probably going to show the this famous comic that many probably you've seen where there's this [Speaker 1] person writing an email and is just saying, ooh, I'm writing this email. I just, I just put like a [Speaker 1] couple of sentences and creates this beautiful long email with all the details that I would never [Speaker 1] be able to write and it makes it so easy for me. Then the next comic strip says, a person receives [Speaker 1] the email, says, ooh, I got this long email. I didn't want to read this whole email. I just got [Speaker 1] the two sentences that I need. So I didn't have to read this whole email. So my AI transformed [Speaker 1] everything right for me that comic is so spot on on what we're gonna see in the revolution of content [Speaker 1] um uh first what does it mean to create content why do we keep on thinking of content as this one [Speaker 1] of thing that you do once and then everyone has to consume the same way maybe you want to see here [Speaker 2] Interesting.

[Speaker 1] Yeah. [Speaker 1] So I think it is a joy here because this thing is heavy. [Speaker 1] So if content, it is something that can be created at any time. [Speaker 1] And it really doesn't matter in what format you create it because on the other side, people are going to transform it in the way they want. [Speaker 1] I don't know many of you, but I know many do this, but I personally, I don't have time to go through all the news websites, all the forums, all the podcasts. [Speaker 1] So I have my own bots that are actually scanning the news daily that actually on the topics that I'm interested in. [Speaker 1] And they craft this really great summaries of all the trends on everything that is happening. [Speaker 1] I would have, I mean, doing that, if I do this manually, would take me a lot of hours.

[Speaker 1] Having an agent that customizes that and provides me in the format that I need, synthesizing the way I want it. [Speaker 1] has visuals in the way I want it now all of a sudden I customize how I create consume my content [Speaker 1] when I say digital content it's not just you know news or or podcast I mean everything that [Speaker 1] can be digital yeah and this music movies series uh uh newspapers even digital software right [Speaker 1] what if both the production and the consumption changes entirely and we just have an AI in the [Speaker 1] If I have an idea, I can put it out in the world. [Speaker 1] AI is going to be able to distribute it and give it to the right person and hyper-personalize it to my audience at the right time and the right format that this audience wants. [Speaker 1] And then how do I make myself different?

[Speaker 1] So we have two dimensions, right? [Speaker 1] One is the consumption. [Speaker 1] I do believe the future is hyper-personalization of content. [Speaker 1] So everyone is going to have their content in the way that they prefer. [Speaker 1] And then there's the production. [Speaker 1] And the production, what is important about production is how do you make things that are truly unique? [Speaker 1] that AI has not seen before. [Speaker 1] That's why I think the capacity to have the capability to connect dots in a way that AI [Speaker 1] was never trained on is going to be crucial.

[Speaker 1] And I feel like the people that are going to be successful in the future are going to [Speaker 1] be the people that are able to connect dots. [Speaker 1] So I think education better be as broad, as generalist as possible. [Speaker 1] It has to be hopefully self-driven by your interests, your passion. [Speaker 1] and that way the more differentiated [Speaker 1] you are from the rest, the more chances you have [Speaker 1] to have succeeded in the future both at work [Speaker 1] in your education [Speaker 1] and in content because at the end [Speaker 1] and we're tying all this together all of a sudden [Speaker 1] content is the way that you're actually going to be found [Speaker 1] by people that are looking for jobs [Speaker 1] assuming we still see companies [Speaker 1] in the way that we see them today [Speaker 1] but I feel those three things [Speaker 1] are going to be very very different [Speaker 2] I also think like on the flip side [Speaker 2] I think there is also going to be another [Speaker 2] time people I mean you just explained that where you just said that people will be consuming [Speaker 2] content also in a very different way so but the the amount of time that they'd be spending would [Speaker 2] be the same right if you have 20 minutes that you could spend on like Facebook and Instagram and [Speaker 2] Reddit all of them combined it's not magically going to become two hours tomorrow because of [Speaker 2] all of the real life constraints that everybody has right so it's still going to be the same or [Speaker 2] more or less the same. It's just the form of consumption is going to vary. And therefore, [Speaker 2] the distribution of powers, even with the social media platforms, will shift more towards creators [Speaker 2] and also more towards AI agents that are actually now new gatekeepers of attention.

[Speaker 2] So I see at least more distribution of power on that side quite a bit. [Speaker 1] Social media is going to be an interesting one. [Speaker 1] Again, we assume social media is still going to be the way it is today with X, with Twitter. [Speaker 2] Bots are everywhere, though. [Speaker 1] Yeah, I mean, you go to LinkedIn and what do you see? [Speaker 2] Oh, my God. [Speaker 2] Don't talk about LinkedIn. [Speaker 1] 90% all AI generated, right?

[Speaker 1] So, you know, social media has to be the same way, right? [Speaker 1] Like when we think about what is the version of hyper-personalized social media, right? [Speaker 1] way. I don't want to have to get this sort of feed that someone's trying to create for me that [Speaker 1] they think that might be relevant. I want to be able to create that and get actually the news in [Speaker 1] the way that I want, right? So I feel social media is also going to change a bit. Web, social media, [Speaker 1] everything is going to change. So I think there's exciting times ahead. We are living in a fascinating, [Speaker 1] fascinating time right now. Yeah. What are you excited for the next week? Let's see. I feel like [Speaker 1] definitely I've seen what Opus 5.5 is doing [Speaker 1] it is quite spectacular [Speaker 1] I'm always excited to see what people create [Speaker 1] it seems like every day there's new skills to build [Speaker 1] on top of that and the videos are incredible [Speaker 1] I'm excited about so have you heard about this [Speaker 1] OpenAI [Speaker 1] the dots right [Speaker 2] that's the one I want to experiment with the most [Speaker 2] is [Speaker 2] OpenAI is essentially saying that it's a proactive agent [Speaker 2] that's always on that understands you [Speaker 2] and personalizes things to your [Speaker 2] context but [Speaker 2] OpenAI as of today is still a single player chatbot, right?

[Speaker 2] So it still doesn't have access to the company-wide context [Speaker 2] or rather family-wide context if you want to think about personal lives. [Speaker 2] So it's going to be very fascinating to see how that thing will turn out. [Speaker 2] And there are also a lot of other companies that are starting up [Speaker 2] in the same concept where you're like, [Speaker 2] oh, we're always proactive. [Speaker 2] Muse is one of those. [Speaker 1] from facebook and uh a lot of people are talking about music i know the automated calls that [Speaker 2] they're making well i think it's easy for i mean it's not easy for facebook because they already [Speaker 2] own like quite a bit of distribution there but i have to i haven't tested muse yet so i'm looking [Speaker 2] to test out both muse and dots and see um how where things stack up make sure you try those too [Speaker 1] uh i think uh the future is an agent uh it's obvious that that is something that is going [Speaker 1] to transform everything. And I think that tied to compound intelligence that you're going to start, [Speaker 1] when you start teaching those bots or those agents to do things on their own, it will blow [Speaker 1] your mind over time how autonomous and how agentic they are. So yeah, definitely encourage everyone [Speaker 1] to try it out. I think for today, that will be good. I think this is a good wrap. Thank you all [Speaker 1] and see you next week. Bye.

[Speaker 1] Thank you.