How Packy McCormick Finds His Next Big Idea - Ep. 29

[00:00] I feel like we could make a Cod project for you that has a really rigorous set of rubrics for what essays are best. Hell yeah, let's do it. I'm going to impersonate you. I apologize. I'm the Ravaki McCormick from Not Boring. I want to create a detailed rubric that explains what goes into my best writing. Do you have any pieces that you've written that you think are bad? Huh, no. [00:30] our knowledge base. And I'm going to say, here's an essay draft, please grade it. [00:48] Are you looking to get the most out of ChatGPT? Everyone talks about prompt engineering, but what they don't tell you is the best way to get ChatGPT to follow instructions is two tickets to the gun show. [01:00] - 2001, 2002. - And if you wanna maximize your gains, both with your body and with Chachapiti, I highly recommend you try Creatine Gummies from Create. I actually eat these every single day and they help me maximize my lifts and give me more energy. They also taste pretty great. If you use the link below, you'll get 30% off. Just use every 24. - Are you getting this? - Paki, welcome to the show. [01:24] Great to be here, Dan. Excited for this. Really good to have you. So for people who don't know, you are the founder of Not Boring. You write some of the premier essays and analysis of tech. You have about 200,000 email subscribers.

[01:42] And I think you've just been one of the main voices of this era in tech. And in addition to your writing at Not Boring, you also have a fund, Not Boring Capital. [01:54] You've been a good friend in this whole journey, and so it's really fun to have you on the show. [01:59] It's I mean, thank you for the for the interest of you here. I still remember that I don't think there would be a not boring if it weren't for that group chat that we had on on Telegram with all of the right. The first day that I walked in and like saw you in there and saw a few other people in there, I was like, I can't I'm like, I've hit the goldmine of Internet writers. This is amazing. And I really I don't know if there'd be not boring without that. So fun to bring a full circle here. [02:21] Yeah, I love it. And I want to get into, we'll get into some of your like, your AI use cases, you have some interesting ones that I really want to talk about. But I think before we do that, like, I just want the update, like, just almost as a friend and like an admirer of your work, like, what's, what's the latest? Like, what's going on with Not Boring? Where are you right now? What are you thinking about? [02:39] Oh, I'm kind of like, kind of have come to reuse the phrase full circle, full circle. Again, one of the first things that I ever wrote before there was not boring, just like when I started writing was this piece on like, natively integrated companies, like internet slash vertically integrated businesses. And I've really just like gotten obsessed again with, with vertically integrated businesses, calling them kind of techno industrials. I think we've talked about kind of AI, and how it applies to analyzing some of these, because they're just like these big complex piece. [03:09] things that if they work turn into really good business models and businesses that make an impact. And so more and more on both the writing and investing side, I'm trying to converge a little bit on really just diving deep into some of those companies, which means that I'm writing about a crypto company next week and that was a complex one that just nerd snipe me and continues to nerd snipe me. But more and more looking at I've written about Anderil, a company, Fuse Energy, that's doing

[03:39] fusion energy a mining company earth ai and like i just love this excuse to dive into these industries that i should know more about and don't when i start the process and so trying to orient as much as i can around uh you know the investing and writing in those types of companies [03:54] That's interesting. What characterizes a techno-industrial company? Is it like Frontier Tech? I don't love the name techno-industrial, but I needed something that's different than... [04:06] deep tech or frontier tech, because I think there's a lot of deep tech that is incredibly cool kind of science project. [04:14] type companies where they're trying to make something that has never happened before happen and there's scientific risk there and it may work, it may not work. And then there's a stage beyond that where some of that stuff has already been invented and then you integrate it into a company that can deliver a product. This is the thing that I look for. Deliver a product that there's a ton of demand for already, better, cheaper, and at higher margins, which then you can reinvest back [04:44] cliche one i think andrel is like by far the best example out there right now like spacex i think kind of kicked off the the hard tech revolution both in terms of as an example and the talent diaspora from that company but like the the space launch competitors kind of sucked and so like obviously zero knock against uh against it's like the most impressive company in the world right now maybe but i think that to me the interesting unlock about andrel was that they said like we have this different approach to defense and we think we can actually take on these like

[05:13] you know, huge hundred billion plus dollar defense primes that kind of just operate in this cabal, these five companies that dominate everything. And we think that if we make, you know, smaller, cheaper, more suitable hardware, but we infuse it with software, then we can deliver a better product to the government at better prices and cheaper margins and take on some of that risk ourselves as opposed to doing it cost plus. And it's like kind of working now. And I think that kind of thing is going to happen in a lot of other industries. And so that's, that's the thesis. [05:43] Like, what is the why now of that? What's making those companies' shapes potentially important or impactful right now? [05:51] Yeah, I mean, I think one is like they're fundable again. And so like these are some there's still like a bunch of them that we look at. I think still investors like I don't know, that seems like a lot and a lot has to go right for this to to work. But they are fundable now. I think AI is one of the big why nows. I think, you know, if you look at something like that. [06:10] nuclear, for example, there's public sentiment shift, there's [06:18] There is this shift from construction, this belief that every nuclear project should be a gigawatt reactor that is a construction thing to a manufacturing process. It's starting to see enough companies do that, that the fuel supply chain is strengthening a little bit. There's just a bunch of things coming together. The why now, I think... [06:37] Yeah. [06:38] It's hard to do like a general why now on these. And like, one of the things I like about it is that there are specific why nows. I think one of like the big why now is you just feel like in a bunch of different categories, there is sclerosis among the incumbents and there's the opportunity for a company to come in and do something better. I mean, like someone should replace...

[06:57] Boeing as one good example. That mining company that we looked at, the interesting why now is that it's just gotten a lot more expensive and harder to discover [07:06] new rare metals and new critical metals. And so you need kind of like this happens in the history of oil development, the history of mining, where like you kind of get everything that you can at one technological level and you hit a wall and people like, oh, shit, we're going to run out of this stuff. We're going to hit peak oil. We're going to run out of critical metals or China is going to get all of them. And then someone comes in and innovates and then you can kind of like get to the next level of something that used to be more expensive that now you can make economical. And so it's different per category. But I think, uh, [07:34] Infusing software and hardware is probably the big overall trend. That's really interesting. And then just selfishly, what should I think about my crypto portfolio right now? I'll say I'm holding. I've got all the Bitcoin and the Ethereum. But obviously, there was a huge bull cycle, I guess in 2022-ish. And then it seemed like it was almost coming back six months ago or four months ago. [08:04] traveling in which is like only crypto adjacent um like what's your current crypto take [08:10] I'm also hodling Bitcoin and Ethereum and Sol and, you know, a few other things and our portfolio. Is it hodl or hodl? I don't. [08:21] HODL? HODL is closer to HODL. I've only ever read it. Yeah, me too. I've never said it out loud. And I understand why people don't say it out loud, because it's kind of tough.

[08:32] But I mean, obviously not financial advice and blah, blah, blah, but I have held on. I've bought more in the bear market. I think I was certainly excited when I got to see the prices of the things in my portfolio going up in that kind of last mini little bull. But I think so much of that was exogenously driven and driven by the Bitcoin ETF and the anticipation of the halving and things that don't actually matter. A lot of the people even then that I talked to, I was like, oh, this is awesome. Is this the bull market? Is it happening? [09:02] This isn't the bull market. And I think a lot of people smarter than me have said, you know, 2025 is when it happens, it'll probably happen in 2025. But, you know, I think we're still waiting to see what those breakout things are going to be. This company that I'm writing about for for next week, Blackbird, I think, is a really, really interesting product that is kind of crypto and and more just kind of a restaurant app. There's a company that I wrote about last week called OnCyber. It's one of our portfolio companies that. [09:29] makes it easy to build 3D worlds in web browser. It's another interesting one where they wrote their own kind of scripting language to build worlds on top of the game engine they had created to build this NFT gallery. [09:41] But now, you know, the founder, I told him on a call what website to make, and he like kind of described it. And it wrote something in their scripting language in a chat GPT, kind of in a chat GPT bot. And then he plugged it in. And then he just had this kind of like game slash world that got made. And so I think that's going to be really, really interesting. And they have crypto kind of payments just... [10:02] baked into the site. So you can turn on, you know, stable coin payments, you can turn on NFT minting, a bunch of easy stuff. So I think hopefully we see a bunch of things where crypto is infused, but it's not the whole product in this cycle.

[10:15] Do you buy the argument that we're just starting to get to the place with L2s and other crypto infrastructure that we can really start to see actual breakout consumer apps in this next cycle? Where do you think the next breakout comes from? [10:30] If I knew I would be richer than I am. So I don't know where the next breakout comes from, but I do genuinely believe that. In the middle of the bear market, I talked to a bunch of people who kind of re-inspired me on crypto generally and wrote this piece that essentially says like, [10:48] crypto gives you these things that you can't otherwise do now like up until now you've had to make some like real performance trade-offs in order to use those things and so if you need those things like if you're launching a meme coin or if you need to own an nft that you can't own other like there are specific things where that trade-off and performance is really really worth it but over time the performance trade-offs [11:10] drop and the capabilities you have either remain or get stronger and so at some point you almost get like a free option on all those other things that crypto enables i don't know if it's this cycle or the next one when that like kind of flips but i think just over time as performance gets better as things get cheaper you just get stuff for free with crypto that uh that will be useful to include in in different products [11:31] And what are your thoughts on the... [11:34] I think there's this sort of galaxy brain take. I think it might be Fred Wilson, who's one of the main people behind this. But just generally, there's a thought that as AI proliferates, enables the proliferation of lots and lots of content that may or may not be trustworthy, it creates more of a job to be done for crypto, to verifiable identity, to verify who's saying what.

[12:04] Like what's, what's, um, [12:06] It's a medium part of my thesis, and I think it's one where it's a big part of a lot of people's thesis. I think Fred has a very nuanced take on it. I think there's a general kind of crypto AI thesis that is like, oh, my God, like crypto and AI, like, of course, it's going to happen. And crypto, you know, the crypto AI coins have taken off, like whether or not there's an actual kind of real product behind them and all of that. So it is like a real narrative in a meta. Yeah. [12:32] But the thing that I'm still trying to figure out is like, there is a tipping point where at some point, like, [12:40] To sign, to have to sign something, if you're like the only person who's signing something and proving that it's real, it doesn't really matter. Like if no one is signing their transactions, uh, [12:51] cryptographically, then there's no expectation that other people will. So I don't know where that tipping point comes. I do wonder if an X integrates it at some point or if platforms start to integrate. Maybe. I do think we wake up in a world in 10 years and that's probably something that just happens for people to prove it. Obviously, WorldCoin doing the proof of humanity thing and others doing proof of humanity or decentralized ID. I think all of that is interesting and will be valuable. I just think there's a lot of stitching work that needs to [13:21] It's just like kind of default that, [13:23] If you're sending something as yourself, you're signing that thing. [13:27] That makes sense. So I want to start getting into the AI, your uses of AI. And the way I want to do that is like just sort of taking a step back. You as a person, you have this sort of like...

[13:40] incredible amount of curiosity. You love thinking about really, really complex companies and spaces and writing long form essays about how it all sort of fits together. And so, [13:51] Your job is almost to be really curious about the world and about companies, do a shit ton of research, write about it, and use that to crystallize a worldview that then you go and invest on. [14:21] and using all that writing to create a thesis that then you use to invest. [14:26] Yeah. So I think it's, it's evolved a little bit. So with GPT-3 and ChatGPT, it started out just as I was writing something, you know, like I wanted it to give me like facts and all that. And it didn't do that, you know, predictably well enough that, that I could use it. But if I were looking at, you know, writing about fusion or doing this, this podcast that we did Age of Miracles with Julia DeWall and I, where we looked at nuclear fission and fusion, there'd be a lot of things where I'd be writing a script for that. [14:56] getting X, Y, and Z about vision right? Or like explain what I'm getting wrong here? Or is there a simpler way to say what I'm trying to say here? So it was a lot of going back and forth and doing kind of like... [15:07] do I understand this right? Or like, here are the different types of kind of fusion fuels. Am I missing any? Or like,

[15:14] Different things like that as I was exploring. And I still do that for sure as I'm using the chat tools. [15:22] It evolved into, I started using ChatGPT as my editor. And I'd say it's like an okay editor, not a great editor. It's like not going through and doing copy edits or line by line. It's not like really tightening my pros. What I've started using recently though, I've started using Cloud 3.5 Sonnet and now Projects. And so Sonnet, I just think is at least like it sounds nicer. [15:44] as an editor and it makes me believe that it's giving me better feedback than ChatGPT. I ask for feedback in a grade just because I'm vain. And every time, pretty much without fail, the grade starts at an A-, and then we'll get to an A the second time I send something, and then we'll get to an A-plus the third time I send it. I could also not change something, and it'll still probably go through that same progression. If I say... [16:10] If I say when I'm like asking for feedback, send it the same exact thing. And I'm like, all right, give me a grade on this one. Somebody else that I sent it to said that like, this was one of the issues with it. [16:19] Almost invariably, it'll really emphasize that issue and then give me a B or B+. So it's not the most reliable editor, but it will say, this section is too long and it drags, and I'll go back and read it, and that section is too long and it drags. Or I do think it's too... [16:37] defensive or like wants you to like really look at all, like examine all of the other sides and counter arguments to what you're saying, which I think is a valuable thing from a valuable exercise for me to do.

[16:48] do as I'm writing. And if there's something that like really does kind of, uh, [16:53] poke a hole in my argument, like then I'll go back and kind of try to incorporate that. But like, it really, I think, overemphasizes the defensiveness in writing, which I, [17:03] It's okay. But I do think it's useful. And I think it's just become like almost a habit now where if I'm stuck for a second, I'll just throw it in into Claude. It's like a thing to do to do something while I'm feeling stuck. Projects I've started using for the past couple of weeks, and they've been a revelation. And that's both on the writing side and the investing side where... [17:24] Take a piece, the piece that I'm writing on... [17:28] on Blackbird, like, [17:30] dumped all of my notes with the company in there, dumped the materials that they sent. I was studying, you know, like when I write one of these pieces, I'm trying to understand the industry that they're operating in. So try to understand the restaurant industry, put analyst reports in there, put blog posts that people have written in there, like just dumped as much content as I could. Everything that I'd normally put into a notion as I'm going to research, I put in the notion now, but then I also just dumped the text into the [17:57] in the clod [17:58] And then I'll ask you questions as I'm writing. I'll be like, cool. Here's the section that I'm writing. Here's what I wrote. Is there any interesting data anywhere that I should be using that I haven't thought about using? Or are there quotes from X, Y, and Z investor that I talked to that might be useful here? And that's been really cool to be able to do. Or like, here's my full investment memo on this company. [18:17] Am I missing something like really important? Or if you're, if you're an LP is looking at this, like, where are you poking holes in my thinking? So that giving it that context and then asking it to analyze the work given that context, I think has been really, really valuable.

[18:30] That's really cool. I want to see that. Can we look at your project for Blackbird? Ooh, I don't know if there's anything secret in there, but all right. [18:39] Thank you. [18:39] So... [18:41] Here on the right-hand side is like... [18:44] A bunch of different stuff that I... [18:48] was reading [18:49] about Blackbird. And so that could be Fred Wilson's original blog post, a couple of those. Jay Drain at A16Z wrote a blog post. It could be their press release. It could be [19:01] blog posts that the company's written an interesting one that they did recently. They, uh, just dropped their, uh, their white paper, the, the fly paper. And so that one's really interesting to say like, all right, like I'm going to read this and there's actually a little bit more approachable because it is more of like a consumer and restaurant facing product. But for a typical crypto white paper, to be able to have something where I can just ask it questions as I'm going through and ask it different things is super, super valuable. Um, [19:30] And so like one thing that I did here was ask, [19:36] you know, please explain black bread economics in a style. That's a crossover between not boring and the diff or like, you know, I was, I played with doing the diff recently. Um, [19:46] And I always have to tell it to... [19:50] to not do it like a caricature and then it'll do it like a caricature and then I'll go back and be like you did it like a caricature and then it'll write something kind of like good in my my style and I don't take that but I'm like just looking for just inspiration for different angles that I might not have thought of and I feel like putting the diff in there like in this one in particular they had a couple of explanations that I was like yeah I don't know if I would have explained it like that but that's actually a really good explanation um and so just you know use this

[20:20] over the top. [20:22] And then use it to go back and forth, kind of almost describing... [20:26] the thing that I'm reading back to me in my own voice, in the own way, in my own way that I would write about it. And it's not again, perfect, but it's like close enough that it does make it sink in more than just kind of looking at a regular white paper would. I just want to stop you there. Just stop you right there. Like, I think it's so, um, important and misunderstood how much summary is a part of good writing. Um, and in particular, like how much summary is, uh, [20:56] even know what they're talking about and then for writers to give to readers to like set up the point that they want to make um like that's it's a huge huge huge part of writing and for something like that okay if you scroll up like [21:11] A summary of tokenomics, for example, it might... [21:16] take you like a day or whatever to like really deeply do that even if you even if you generally get it like even if you've read all the stuff and you have it in your head like to actually put it into like a crisp clean format could take a day or two at least and one of the beauties of cloud projects but like just generally this this generation of ai is like [21:38] particularly for complex ideas that you already understand, it just saves you that summarizing time in your writing and in your thinking so that you can get to the more interesting stuff. And yeah, like you're probably not going to just copy paste this exactly, but it's, it's surprising how much just having it in that form where it's like basically in the context that you need it, just like speeds you up to like, just get through that part of it so that you can do the actual

[22:08] And... [22:09] And like, if I do summarize being able to say like, cause I'm always debating, I'm [22:14] Should I do more in? [22:15] probably too often do more than less. But should I do more here? What am I missing here? Did I get anything wrong? A common one is you're like a Hacker News commenter. What is wrong with how I just summarized this thing? [22:31] And I do think there's a danger to that. And like, same with the way that they give feedback when I asked for feedback in a grade, which is like, [22:38] I think there have been a couple pieces probably that got a little more watered down than I would have wanted them to because I have this critic there that otherwise wouldn't be kind of involved in the writing process. And so I think it's a balance between just like making sure that I avoid – [22:54] really obvious errors and then like kind of watering down or like pulling back my argument too much because I have a hacker news commenter kind of all of a sudden in my in my ear. And so that is like an interesting balance to try to play with. I do that all the time. I mean, like anytime I'm writing about something that's technical, I just like have this voice in my ear that's like. [23:15] you're not really a great programmer. [23:18] Do you really understand this? You don't have a PhD. And I love Claude for that because it just gives me... [23:27] enough confidence because I can just put what I'm saying in there and just be like, okay, pretend you're like a PhD computer scientist. Like... [23:34] What are the holes? And it might not get everything, but it gets enough that I just feel comfortable that like it's I'm not saying something completely stupid. And I feel like that's so important for writers to like be able to like have that resource. And previously it was only available to people who had like big editing teams or a lot of resources or whatever. And it's just like it's the coolest thing.

[23:54] Totally. Um, but there's, there's one. So, I mean, so this is, I wrote a piece, uh, [24:00] a few weeks ago called the American Millennium. And I had this, uh, I had this idea that I wanted to play with. That was one of the interesting things about, uh, [24:11] America is that the system works so well and the entrepreneurial drive works so well and all these things that kind of no matter who's in the White House, as long as rule of law and the Constitution are maintained, we might not suffer the fate of the Romans because people can just build things and prevent the decline. And so I wanted to show that. [24:32] First of all, play with the idea, but then show that. And so I asked it to code me up a little thing where like... [24:38] I could show what happened to the, the vector sum of kind of like government and entrepreneurship on progress. And it just coded that up for me. And I went over to, and it's like, you know, very, very simple thing, but I like put it in the, I put it in the essay. Um, [24:53] And then I went over to Replit and just like dumped the code in there. And I had a little website that you could like click on in the essay and, and play with it yourself. And so it's like fun. Like this is not going to, you know, this is not going to hold up to peer review by any stretch of the imagination, but it's just like another like interesting thing where I never in a million years would have, uh, would have made this. Maybe I would have tried to draw a shitty version of this, but to like have this little interactive thing that I could put in there, uh, because of, because of Claude, like that was, I thought, a pretty fun thing to be able to do. That is so cool. [25:23] how you made this. So press continue chat. Like I want to just go up to the top. Like I want to know how you, like where you started this. Cause I think this is like just such an underappreciated thing where you're someone with, um, with so many ideas in your head and suddenly you can just like make something that's this, um,

[25:42] new form or new representation of something that's been in your head that like maybe previously like you could have done but it would take a lot of work or it would have you had to pay someone and now you can just like get something out of your head in this totally new way that's not writing but is like it's a visual representation or it's a it's an interactive demo and share with your audience like I just think that is the coolest thing. [26:04] It's really fun. And I've done this a couple of times. This is like a slightly different one, but I used ChatGPT for this one when I wrote about America's tolerance for risk or risk appetite. [26:18] I had it judged through a bunch of different periods, like what America's risk score between one and a hundred was and just like asked it to make me a graph on that. And you have to disclaim the hell out of the fact that like, this is not accurate. It is just guessing like all of that, but it does give, I think a pretty good kind of like. [26:36] rough guide and it's a quick way to make a visual where I don't know how I would have [26:41] have come up with those numbers otherwise. And so just to be able to quickly be like, "Here's a dumb idea. Can you give me these numbers and then turn them into a chart?" It's something that I probably wouldn't have done, but if I had, it would have taken a few hours and would have been equally imprecise. And so it's a fun little thing that I've done a couple of times now. So this one actually started with, "Here's a thesis I'm playing with that I'd love your help thinking through. Since technology compounds and government doesn't, technology should overtake government and importance at some point." And so it gives me some ideas here, [27:11] What are the, can we read through some of the ideas? Like I'm sort of, I'm sort of curious. Yeah.

[27:16] So you're right, the technology tends to advance exponentially. Each new innovation builds on previous ones. Government evolution, while governments do evolve, they generally do evolve at a slower pace. Relative importance, the idea that technology could overtake government in importance raises some questions. What do we mean by importance, political power, impact, economic influence? Is it a zero-sum game or can they both increase in importance simultaneously? [27:37] What are some scenarios if that happens? What are some, again, always, what are some counter arguments to this point? And then historical perspective. And then. This is really cool. I think that's so cool, actually. And I know, like, I totally get your point that, like. [27:52] It's sort of annoying when you're just trying to figure out how to make this essay better or whatever. And then it's like, here are 15 counter arguments. One of my pet peeves is Claude always apologizes, even though I didn't say you did something wrong. I'm like, okay, make this better. And it's like, sorry, you're totally right. And I'm like, you don't have to say sorry. It's fine. But I also think it's so cool that it's even presenting some of this stuff. [28:22] question like that which like the question presumes a particular worldview um there's a lot of assumptions and it doesn't just like say here's the answer um it like says here's like let's assume what you're saying is right but here's like a bunch of ways that it might be wrong or some things you need to think about which is like completely different i think from the like sort of filter bubble world that like we come from with google where it's just like showing you like

[28:49] the latest Breitbart screed when you search for something that like sort of indicates that you're like, you want that, you know? And I love that. I think that that's really cool. It might be for you. For me, as someone who doesn't ever get anything wrong, it's not particularly valuable, but... I love it. I love it. Yeah. I mean, I just don't know what it's like to have such godlike power. But yeah, that's good. [29:19] And then you can go in and the other thing I don't want to do is spend all this time like, oh, my God, I have this really cool idea about technology and government. And then like 75 people have written that exact thing before. So I'll ask it. And, you know, sometimes it's it gives me things that are real. Sometimes I'm like, oh, my God, that sounds like the exact blog post that I want to write. And it's a completely made up thing and it doesn't make any sense. And so I'm like, all right, cool. Yeah. [29:45] So there's that. Did any of these help you? Let's read through them a bit. The Sovereign Individual. Are any of these actually, you kind of get it from, or it helped kind of shape the argument for you? None of these actually, yeah, none of these did. Yeah, that's interesting. Because I do think that is a sort of underrated thing for Claude, especially for writers or anyone who's trying to do something new, is like... [30:12] if you're truly writing something new there isn't a pre-existing website that like talks about this idea in terms that you're talking about it um and that's that makes the question effectively ungoogleable it's like who else has thought about this like you just can't google that um you can like try a couple things but like it's just unclear whether or not you're gonna find it and because claude or chatty or any of these ai tools like

[30:36] kind of get the underlying concepts. They can like move laterally through like the entire space of like human knowledge to like zero in on people that might've talked about similar things, but in different terms than you. And that is like the most valuable thing for people like us. I just love it. I agree. Uh, yeah. And these, these ones, I hadn't read everything, but I like, you know, I'm familiar enough with the ideas from sovereign individual. I've probably read most of the things that, that Mark Anderson and Peter Thiel have read. And so like, I had a lot of it and so [31:06] of thinking about this thing that I'm trying to think about. But yeah, there's been plenty of times where I'll be like, oh my God, I have this novel idea. And I'll be like, philosophers have been talking about this for 17,000 years. And I'm like, all right, cool. You saved me a lot of time. Thank you for an embarrassment. Thank you for letting me know. And it is also helpful because I do think there are a couple of types of writers brought up, but like, [31:33] Ben Thompson and Byrne in one category where like, [31:37] they just have bigger, like more structured brains probably. And like anything that comes out, they're like, Oh, let me hang it on. Like, I'm like, [31:44] This piece of my framework over here next to all of this other relevant information. And my brain is like, that's an interesting new idea, even if I had thought about it yesterday, because I just have like no memory and no whatever. And so having something like this, where I can just check those things and maybe in another brain just kind of exist, I think is really helpful. [32:01] Yeah, I agree. I feel the same way about my writing. Like that's something I've been trying to change is like, I call it like, um, it's sort of like, um, sort of like a sitcom. It's like every day is new, you know, every episode is new. Um, whereas like, yeah, people like burn or like Ben Thompson, like they're much better at like having like a plot arc through the whole season, you know? Um, and I, I do agree, like having this tool as like a throwback, like I, I often will put in custom instructions, like things about like what I'm thinking about or what are the, what are my key ideas right now?

[32:31] Like, [32:32] harken back to those things or keep me on track is like really helpful. Oh, that's really interesting. [32:38] Yeah, and then we actually get right into the... [32:41] Government and entrepreneurship vector graph. Yeah. So out of there, like, and I think this shows how kind of messy my process is. Like I'm asking about these different books and asking it to summarize. And I'm like, all right, great. Nevermind that. How would you show a graph where government and entrepreneurship are vectors that can either work together or against each other? And then says, that's an interesting visualization concept. I can help you create a graph to represent this idea using a react component. We use a simple coordinate system where the x-axis represents government influence and the y-axis represents entrepreneurial activity. [33:11] The angle between the vectors will show whether they're working together or against each other. Let's create this visualization. And so then it... [33:17] goes ahead and actually creates the [33:21] visualization and it has a direction in this one but not a magnitude. [33:30] And so this is great. I also want to be able to adjust the magnitude of each and look at what the combined effect, let's call it progress, looks like. [33:38] And then, you know, it says excellent idea, which, you know, I appreciate. That was a pretty good idea. At least it's not apologizing. Yeah, I'm so sorry. And then it gives me something here to kind of play with and... [33:52] You know, then I just like, this is a really long chat. I'll ask it for feedback on different things. Um, [33:59] like I'll ask you this particular thing about the graph that I think you got wrong.

[34:03] Uh, ch-ch-ch-ch. [34:05] And it... [34:06] explain something, [34:08] to me and then makes an update to the little visualization based on kind of like almost it thinking through. [34:16] That makes sense. I think the preview isn't loading. Then it does. I apologize. Then it does apologize for the inconvenience of the preview not loading. [34:25] And then I asked for a simplified version because it gives me a choice between [34:29] a simplified version, code explanation, alternate visualization, conceptual explanation. So I just want the simple version. And then it gives me... [34:39] that and then it totally totally messes it up and makes it like even worse um [34:44] *cough* [34:45] And then I say, I do want magnitude back. And we try that again. So just kind of going back and forth through this. And then it like, you know, gives explanations on the graph of what the different things mean. But this is a really long one. So somewhere... [34:57] So within here, and this is before I was using projects, this is just when there was artifacts, but not projects yet. And so I tried to keep everything in the same chat so that it had context on our whole conversation. Now I would probably have this as one separate thing. I have a separate kind of editor project that will go in and edit things because I got so sick of writing like, hi, I'm packing McCormick. I write a newsletter called Not Boring. You're the top editor from world class tech publication. I've hired you to like, that is now just like a, you know, it's ordnance memory. [35:27] and start a new chat, which saves me seconds every time. [35:31] Um, and, [35:33] And then so like actually one other thing here is,

[35:36] So, [35:37] I kind of know what's going on with this graph, but I asked it to describe what was happening in the graph in the style of not boring. [35:45] And that it [35:46] "Picture this: you're at the helm of a rocket ship called Progress." And so I have to go back and say, "This is a caricature of my style. Try it again. Actually, not boring." [35:56] I apologize for the exaggeration. You're right. And I appreciate the feedback. Let me try again with a more authentic, not boring style. Our vector graph tells a fascinating story about the interplay between government and entrepreneurship and driving progress. At first glance, it might seem counterintuitive. So, you know, it does like, even if I tell it ahead of time not to caricature me, it like really does. And then we'll get it right on the second try. At least it knows who you are. Like, I feel like that's a compliment in and of itself. The first time I asked it to do that and it knew me, I was like, [36:26] whole internet, but it was still pretty cool. That's great. I love this. I think this is super cool. Have you thought about any other ways that you would integrate something like this in future articles? Where does this fit in your creative toolbox? [36:42] My creative toolbox is like very small. It's the internet, which I could say access through the ARC browser. I use Notion to just like kind of dump everything. I use Readwise to save articles and highlight things that then feeds into my Notion. I use Notion. [37:01] Google Docs to... [37:03] to write and Substack to send and Figma to make graphics. And that's pretty much the whole stack. And so these are the first new addition to that stack in a while.

[37:15] Thank you. [37:17] And it's just like kind of a, like I said, like it's when I get stuck or when I have an idea or when I have like something, I'll go in here and jump into it. [37:24] I guess I'm asking like more about this specific, like this interactive graphic. Oh, the interactive graphic. Um, [37:31] Bye. [37:32] It really depends on the piece. In this one, I did this, and then I also asked it for... [37:37] uh, [37:38] Kind of the same thing that I did on the risk piece. I asked it for, give me like... [37:43] The the values for this for all of like the great empires in history. And I wanted to like I didn't influence it. And I wanted it to say that America had was stronger in entrepreneurship and all the other empires were stronger on government. And that's what it said. Got me in a little bit of trouble because then I confidently said that America was the only capitalist empire. And I had a couple of people be like, what about the Dutch? What about the British? And like fair, but I still think the government was like more powerful than than the capitalist system there. And so like maybe it does if it does kind of like. [38:12] do what I expect it to or hope that it does, it might give me a little more confidence than I probably deserve in that thought. [38:21] But so there were a couple of examples here. Another interesting one I'm writing about... [38:28] a nuclear company and [38:30] I had it... [38:31] The founder told me that one of the ways he started designing the reactor was that he played around with this particular nuclear code. [38:39] downloaded, I pulled a bunch of that code and I was like, [38:42] All right, walk me through this. Explain what it's doing. What happens if I change X, Y, or Z value? Blah, blah, blah, blah, blah. Not a nuclear engineer by any stretch of the imagination, but even being able to understand that code and understand what the different types of things you might be looking at when you're designing a nuclear reactor. That's very cool. I haven't written that piece yet, but...

[39:05] there's a good chance that I'll have, if I can figure out how to do it, we'll have something interactive that's like, play with a couple of the parameters of this nuclear reactor and see what happens. And so things like that, that I just never would have done otherwise, I could see doing here. [39:17] That's really cool. I love that. I want to move into, you've said that you're using it for editing and different parts of the writing process. I'd love to see the project you're using for that. So it's like literally just, here's my instructions. [39:34] I'm back in the corner. I can't believe every is not on there. I'm, I'm extremely. I'm sorry. I'm sorry. There's some tech writing on every, that is a different, uh, a different approach to, or a different, um, uh, [39:55] Less optimism, I would say. I do love the writing, though. [40:00] I recently hired you to help me take my writing to the next level. I want to maintain the voice and style of Not Boring, but sharpen the writing and analysis to make it the best tech blog in the world. Something that smart, successful people read to learn what's happening at the frontier of tech. You can push me on everything from writing style to structure to the logical strength of my arguments and the data support I provide for those arguments. I want the writing to be as fresh as it was in the early days of Not Boring. I want you to push me on the novelty of my ideas, letting me know if someone has already written something similar but better.

[40:30] approachable to a wide audience not boring his mission is to make the world more optimistic and that means reaching more people with a fact-based look at the most promising companies and trends in tech i'm very excited to work with you and grateful to have you on the not boring team as our first managing editor i don't know claude claude more than the other ones like really makes me want to talk to it uh like a like a person and that's like pretty much you know pretty much all i have there and then uh [40:53] You know, like, so today... [40:56] I will... [40:59] It does an annoying thing, actually, if I dump something that's too short in, or when I ask for feedback, it just starts... [41:05] typing the whole thing in. Oh, interesting. [41:08] Which... [41:09] I did hear it today. I decided to piss off kind of like both sides of, uh, of the aisle with like a fairly moderate take on tech and, and Democrats. And so, uh, my Democrat readers were offended that I like, didn't just like say that Donald Trump was the worst person in the history of the world. And my Republican readers were pissed that I thought that the Democrats were at all redeemable. Um, so it was, it was a really good idea to write the piece that I wrote today. Um, but just kind of, you know, like, [41:36] Well, you... [41:37] Here's the latest draft from my SAFE back grade. [41:39] And then it'll go back and say like structure and flow. It's well structured, tone and voice, the arguments, the evidence, balanced perspective, timeliness, length, always says it's quite long because they are quite long. I wish I could have written this in a thousand words. I'm sure in like Claude, [41:56] you know, 5.5 sonnet, I'll be able to just say like, can you just rewrite exactly this, my exact style, but like half as long and hitting the key points and it'll be able to do that. And I can't, can't wait for that day, but it's not quite there yet.

[42:10] Um, yeah. [42:12] It always says proofreading. [42:14] And then I'll ask it for specific typos and either it'll make something up or, uh, [42:19] Or be like, actually, I just looked and there aren't any typos, but it always does say proof reading. And then, like I said, it'll give me an A- grade on the first draft pretty much every single time. I do want to try to just put something absolutely garbage in one time and see if it gives me a C, but... [42:38] The other thing that I'll do... [42:40] Just because, like I said, it's what I do when I feel stuck or when I just hit a point where I want to stop writing for a minute. I'll jump in just the intro or the intro in the first half of the next section and be like, by the way, this is only up to the intro. And this is only up to the intro in the next section. There's still a lot more to come. Just give me feedback and a grade on what I have so far. And then it'll invariably tell me like, you know, it does end kind of abruptly and like blah, blah, blah, blah. And so some of those are not quite there. [43:10] pick up what I'm writing, [43:12] understand kind of my normal style, you know, say that it'll fit well with the not boring audience, all of that, and then give me, [43:21] Yeah, so like... [43:22] concrete policy proposals. I didn't take any of these, like, [43:26] Joint task forces on cybersecurity. So some of that is kind of like basic and bland. So it's like total new idea generation. I don't find it to be like as as good at here. [43:38] I asked for it to cut and I didn't love its recommendations on...

[43:43] what to cut. So I didn't, I took one of those. [43:48] And then I'll ask better or worse. There's not been... So I'll send a new draft. I'll say better or worse feedback and grade. And there's not been a single time that it has told me that the new version was worse. And maybe I just am that good at improving every time, but I can't imagine that's actually true. I want to see if we can fix this for you. We've got about 30 or 40 minutes left. And one of the things I love doing in these episodes is just doing something together. [44:11] And so here's my idea. I'll pitch it to you. I feel like we could make a Cod project for you. [44:18] That has actually like a really rigorous set of rubrics for what grades mean what and what essays are best. And then we can basically just you can feed it an essay and you can be more confident that the grades it's giving are actually good. [44:37] Hell yeah, let's do it. Okay, cool. So we're in Claude. So I'm going to make a new project. [44:44] And we'll call it... [44:47] Oops, great project. [44:49] Not boring grader. A grader for Packy... [44:55] to use, oops. [44:58] Use on his not boring projects. Okay, cool. So basically, the way that I would approach this, and I would love any ideas or thoughts or revisions that you have, but I think a good way to start would be to define what good and bad looks like. And the best way to do that is usually like

[45:20] to, uh, to find examples. So do you have like a, an example of like an essay that you've written that you think is like the best of what you can do? [45:30] Ooh. [45:32] . . . [45:36] Hmm. [45:39] Thank you. [45:40] The most popular is probably the [45:41] Yeah, Excel never dies. [45:43] I guess we can just do the first one. I thought that was good. Okay. I wrote it with somebody, so it'll be slightly different, but. [45:51] Maybe that's good. That grades me on what I can do with someone else. [45:54] Okay, cool. So I'll, I'll, we can take a couple. So like Excel never dies. Anything else on your mind? Um, yeah. [46:01] We can do the great online game. Okay. I mean, that's one of the ones I think of. I think it's so good. [46:07] Okay. Anything else? Because one of the things I also want to do is like a lot of what you're doing, we want to find examples of like the kind of pieces that you're writing right now where it's like really deep, technical, in-depth explanations. So let's do... We can do... [46:23] fuse energy. [46:24] It should be a month or so. Oh. Tell me when you see it. Keep going. [46:33] Right there. Perfect. Okay, cool. So... [46:37] I'm going to say like, okay, sorry, I'm going to impersonate you. I apologize. I'm the writer, Facky McCormick. Is this how you spell your name? [46:47] Yep. Okay, cool. From not boring. I want to create a detailed rubric.

[46:56] um, that, uh, explains what goes into my best writing. Um, [47:05] Here are three examples of what I think are my best ideas. [47:12] pieces. Can you write a very detailed explanation of what [47:19] Um, [47:20] what they look like and how they're composed. [47:27] How they compose. How does something like that sound? Yeah, that sounds great. All right, cool. [47:33] Thank you. [47:35] I think I spelled rubric wrong, but we're just going to roll. I was wondering if that was like an alternate spelling. I think it's the Swedish spelling. It's very fancy. [47:46] All right. So we're, we're getting certainly Paki. I'd be happy to analyze this for you. And then it's like creating an artifact. So, and I want you to just tell me like, do you think it's like getting, do you think it's getting it? [47:56] well or not. So structure and flow, clear, compelling introduction that hooks the reader, logical progression of ideas with smooth transitions, effective use of subheadings to break up long-form content. [48:09] engaging conclusion, consistent pacing. Do these feel good or are they too high level? These feel good. Yeah, they feel good. Okay, cool. Deep dive analysis, comprehensive exploration of the main topic, integration of multiple perspectives, original insights and connections, balanced treatment of complex issues, clear explanations. How does that look? That looks perfect. Okay. We've got extensive use of primary and secondary sources. Yeah, just scan through these and see generally, does this look at least reasonably good?

[48:38] Thank you. [48:39] Yes. Okay, cool. So one thing that I like to do, I've actually picked this up from Matt Schirmer recently, is I just say make it better. [48:50] Just to see what it does. And it will often honestly be better, which is kind of funny. Crazy. [48:58] Thank you. [49:01] And it is interesting how it defined for itself what better was. Yeah, totally. [49:09] Um... [49:11] Thank you. [49:12] It seems a little bit less generic, which is nice. Connection of topic to broader cultural societal trends. Injection of humor and wit. Okay. Here's another thing that I think it's missing. Um... [49:23] is, can you provide really specific examples from these pieces? [49:31] Thank you. [49:33] Thank you. [49:33] To judge. [49:35] Thank you. [49:36] Thank you. [49:37] Amen. [49:38] Thank you. [49:39] Thank you. [49:41] Thank you. [49:42] See, it already is confusing. [49:45] it's confused it it's confused excel never dies opens with there may not be a company oh no worth rooting harder for than fuse energy it'd be a weird way to start the excel piece oh man um well uh you know every once in a while claude shits the bed um but well but we'll roll the next one the next one it nailed [50:07] Cool. Next one, it nailed. That one, it nails?

[50:11] comprehensive action. Yeah. [50:14] Okay. So this seems, this seems reasonable. Um, can you, I, I honestly don't think it will correct itself. So we'll just, we'll correct this in the final rubric. Um, yeah, that works very well though. Yeah. We'll make this, we'll make this better as we go along. So, um, last thing I want to try doing is do you have any pieces that you've written that you think are bad? Huh? No. Um, or let's just say like, don't represent the best of what you can do. [50:44] stressed you like hired a ghostwriter one week just to see if it would uh work you know all that kind of stuff i would never um [50:53] Let me see. Praise Elon was not great. Praise Elon. Okay, cool. Praise... [51:02] Elon. [51:04] Whoa. - Pretty cool. [51:06] What? I don't know you could search within Substack. Yeah, I don't know. That's interesting. [51:11] Okay, so here's a piece I don't. [51:15] like very much. [51:18] Based on this rubric, can... [51:24] Oops. Can you tell me why? And before I do this, what is your feeling about why this isn't a good article? [51:34] It was like too, uh, too kind of like in the moment I'm still on Twitter. And like part of the angle was like, you know, Elon, it was like kind of like almost too glib and tongue in cheek where it was like, Elon's going to kill Twitter, but that's a great thing because then I can actually be productive again. And like, you know, like it was, it was, it was just too, uh, yeah.

[51:57] Shit. [51:58] It wasn't representative of kind of what I want to, what I want to be writing. [52:02] Okay, cool. And so based on this rubric, I feel like rubric... [52:07] It still looks weird with a C, but is there some other spelling that I'm missing? With the C is what I thought, but... Okay. All right. Let's just go with a C. The only other possibility is that, but I just... I don't... That's definitely not right. [52:20] Okay. Here's a piece I don't like very much. Yes, and this is a rewrite. Can you tell me why? Okay. [52:24] Just [52:26] All right, here's what we got. So based on this analysis, praise Elon doesn't quite meet the high standards set by your best work. Here are some key reasons. The narrative feels forced and less organic than usual. The premise of Elon intentionally killing Twitter seems like a stretch. The depth of analysis is shallower than your typical pieces. The piece lacks technical depth that often characterizes your work. While the topic is timely, the approach feels more reactive and less insightful than your usual work. I feel like that's actually pretty good. Yeah. [52:50] That's what I said pretty much. That's basically what you said. That's pretty cool. Yeah. So basically, based on this, I agree with this. [53:03] Is there, are there any modifications to the original rubric that, um, might, um, [53:13] uh, make it better that you'd recommend. Um, uh, [53:18] Just see if there's any ways that it might... [53:22] sort of like change the rubric or like want to look out for new things yeah

[53:28] Thank you. [53:32] Thank you. [53:36] Okay, so it's immediately not doing the rubric with the examples. So we'll have to add that back in. [53:49] I think it added this rigoring credibility [53:52] section or at least looks new. [53:55] Thank you. [53:56] I was wondering if it was going to actually just keep adding to the point total. And it does. So it's not redistributing with 100. We just now have 200 points to go on. [54:07] I mean, I think out of 200 points, it's good. It's the not boring style. Let's see. Cool. By the way, last time it... [54:17] when we asked for examples, it took away the points. [54:21] Oh, did it? I think so. Make sure to keep the point structure intact and make sure the examples are accurate. Be careful. [54:33] Let's see if that works. [54:39] Thank you. [54:40] Thank you. [54:42] Thank you. [54:44] Do you think, is this the right hook for Fuse Energy? Yep. [54:50] Okay, great. So it's being less hallucinatory. Yes. Yes. [54:56] So we're getting compelling hook. Good. There may not be a company with earth rooting for a bad white pill. Elon Musk is intentionally killing Twitter for the good of humanity. Um,

[55:05] I think this is... [55:07] potentially good i think so too i need to feed it more about examples so it doesn't just take from that one but yeah [55:14] Yeah. Yeah, well, we can we can make this a little bit more more in depth later. But now the thing I want to do is. [55:28] I want to add this in to [55:31] our knowledge base. So I'm going to say add content, [55:37] grading rubric [55:40] I'm going to paste it and then I'm going to set custom instructions and be like, [55:45] I'm Paki McCormick, the writers. [55:49] Thank you. [55:50] McCormick like rubric, um, the writer of not boring. You're, and, uh, the best editor in the world in the style of the New Yorker, um, and every, just kidding. Uh, New Yorker, the New Yorker and Stratechery. Um, when I ask you for grade, uh, uh, to grade an essay, please use the grading [56:20] rubric. [56:22] in your knowledge base. [56:23] to give my essay [56:26] A grade. Um... [56:28] Okay, cool. So I feel like we should take something that you got an A- on previously and then just throw it in here and see if it gives you a better score.

[56:38] A more accurate score. [56:39] Let me [56:41] Okay, so you just shared with me this doc, and this is a doc that you put into Claude and you got an A- on, right? [56:47] Yep. [56:48] Okay, so basically what I'm going to do is I'm going to just copy it. [56:52] I'm going to go back to my project and I'm going to say, here's an essay draft. Please grade it. And let's see what it says. [57:03] Thank you. [57:04] 27 out of 30 on narrative craftsmanship and premise. What do you think? [57:09] Pretty good. [57:11] Pretty good. Where do I lose points? I don't know. [57:17] Thank you. [57:17] It gives me a perfect score, but it doesn't. Yeah, that's... Okay, we're going to have to... [57:25] Interesting original connections balance treatment only five out of six seven out of eight. So it's definitely still like pretty [57:32] positive like overall like probably too positive um but it is a really good piece [57:38] What did you say? Or it was just like a great piece. Or it was a great piece. [57:44] That's what we're learning here. [57:47] It gave you 182 out of 200. So that I think is basically an A-. It's a 91. [57:57] But we do know that it had some issues adding, so it could be higher than that. [58:08] score and make sure that,

[58:11] It's really accurate. I want you to be as objective as possible and follow the rubric. [58:19] exactly. I keep spelling it with a K instead of a C. So I think we're kind of, this is kind of interesting because it's like, even with the kind of like reasoning, you know, [58:32] It's... [58:33] Um, [58:35] it's still giving you an A minus. So there's, there's something in it that just like wants to go for the A minus no matter what, which I would not have minus. Maybe I'm just an A minus writer. [58:45] Well, the true test is whether I put one of my articles in there and it gives me an A minus. Yeah. Um, [58:52] Wait, we just asked it to be more accurate and... [58:55] dinged me 17 points it gave you 155 out of 200 yeah the hell um [59:01] Interesting. Yeah, I kind of wonder, like, okay, let's see if I take this, if I take the kind of like be accurate thing and I put it in my custom instructions and I'm just like, okay. [59:15] Make sure... [59:17] Um... [59:20] The score you give... [59:22] is really accurate. I want you to be as objective as possible and follow the rubric exactly. All right, let's try that one more time. [59:31] Thank you. [59:32] um another thing that we could try and this might be getting a little bit too in the weeds is like um it's one thing it's doing is it's outputting the score before it outputs the reasoning so it's like oh yeah and i don't want that i want it to see look it's it's we're back to kind of like it's still like a whatever i want it to like output the score only after the reasoning so i'm

[1:00:02] Perfect score here. One sec. Oh, no. [1:00:07] We lost a couple points off. Wow. [1:00:10] Wow. 198 out of 200. Not bad. So we had the exact opposite effect of the effect that we were going for. At least I was going for it. I think you like the A-plus score. I love this, yeah. All right, let's try this one more time. It's not about actually making my writing better any of this that we talked about today. It's just about making me feel good enough to hit send. Only after you output the reasoning, this goes for subsections as well as the full... [1:00:39] the full grade output reasoning first. Okay. All right. One more time. Please grade this essay. [1:00:51] Thank you. [1:00:52] Thank you. [1:00:54] Thank you. [1:00:56] Thank you. [1:00:58] All right, 20 out of 30, 32 out of 35, 22 out of 25. So we're starting to get maybe a little bit more... [1:01:06] Um, [1:01:08] a little bit more accurate. [1:01:10] 183 out of 200. And I think the reason is... [1:01:14] ChatGPT thinks by writing and... [1:01:18] If it's... [1:01:19] if it's outputting the score first, it's going to score it and be pretty positive and then use its, um, use that score as a way to figure out what it should write in terms of its evaluation. We want to reverse that. That's it's, it's so funny too. I, I,

[1:01:37] I was talking to my son about something the other day, and I forget what it was, but it was something that complete... [1:01:42] like made up nonsense, whatever. And he like got the end idea in his head. And then I saw him like his little four year old brain, not yet four year old brain, like rationalize his way into that thing. [1:01:53] by like making up a whole story, I was like, wow, like we, we really all, including, I guess, Clyde all do this. We do. And that's what I think is so funny about people who are like saying like, oh man, like, um, [1:02:07] I, uh, [1:02:10] uh like i don't think these things are smart um and they're not going to get smarter and it's like have you ever interacted with a child like this is how children reason a lot and they're just they just get better and better and better and get better at like you know having more nuance and all that kind of stuff but there still is sort of that basic thing that you're talking about where yeah you kind of like you use the reason to justify the intuition or whatever and um and yeah i i really think it's what makes me sort of confident these things are going to be very very [1:02:40] is you just watch children and they do the same thing. Where do you think that, on this current path, where do you think the smartness ends? I wrote people thinking they're going to get really, really smart, and I don't... [1:02:52] there's agency or like something missing that just doesn't make it feel like they're going to take over. What's your, what's the house view? I don't, um, I'm definitely not a doomer. Um, and I tend to think that right now, like, yeah, we have the whole scaling law thing. Um, um,

[1:03:09] But I generally think that these systems are incredibly, incredibly inefficient currently. [1:03:14] And, [1:03:15] they're going to get orders of magnitude more efficient in terms of training and in terms of like inference, like the amount of compute you need for inference. And as soon as that happens, a couple of things I think will start to become possible. One is right now, in order to get new information into them, you have to prompt tune. And I think that we will figure out architectures that allow them to learn after their initial training is done, which I think is really important. [1:03:45] What you've probably seen me do so far is I'm pressing replay on this chat a few times because I want to see all the different ways that it could possibly grade this essay. So I can map the entire space of what it might think. And then what I can do is I can just sort of flip through them and be like, okay, generally the score is around 180. [1:04:15] different this is 184 the last one is 181 the last one is 182 but that's like basically where it is and i think a lot of creativity is basically doing this but like um a thousand times in half a second and the minute we can do that uh and and they update as you talk to them i think we will i think it will feel very very very similar to like talking to an actual human

[1:04:41] Interesting. Yeah. [1:04:43] I wonder if I... I'm trying to find another bad assay because I want to see if I can get an A- if it gives me an A- or not. [1:04:50] Yeah, give me one that you don't think is very good. [1:04:57] I don't know if this is bad or good, but a good of the unbearable heaviness of being positioned. [1:05:02] okay what are your general thoughts on it before we before we grade it um [1:05:10] They don't remember it super well enough. It was like 15 months ago. [1:05:16] But I think maybe one of the things might be like... [1:05:19] It's a little too cute on like, [1:05:23] words, like in terms of, uh, [1:05:26] sustaining an innovative role. [1:05:28] disruption and counter positioning and being positionally it might be like trying to do too much on like small distinctions but let's say okay cool um so i'll grade this essay narrative craftsmanship 30 points it got 26 out of 30 which is lower much lower than the last one um depth and originality 33 out of 35 which is pretty high technical mastery 21 over 20 out of 25 cultural resonance 14 out of 15 distinctive voice and style 23 out of 25 intellectual rigor 26 out of 30 [1:05:58] actually like significant I was gonna I was about to give quad a lot of credit and Claude just could just prove me wrong if it's math is right I have no idea if that if the math here is right but eyeballing the scores it looked like the the subsections were a bit um we're a bit lower but yeah it just looks like Claude wants to give you an a-minus it's so interesting

[1:06:20] I've actually found this because we have this app, Spiral, that we built. And we have the same grading system for it where... [1:06:33] where it'll allow you to generate a couple different versions of the same thing, and then it has a grader that says, like, this is good and this is bad, and it just gives itself good scores. I really, like, wonder... [1:06:48] I'm sure that there's a way to like prompt tune this and kind of get around the like A minus thing. But it seems like... [1:06:55] Yeah, it's just a... It's like it's an attractor. This is where it wants to go. I wonder if there's like a bell curve trick you can give it or a grading on a curve kind of trick. Yeah, I think we could also like... What I would probably do if we had a little bit more time is I would probably... [1:07:13] have it like explicitly lay out, this is an A, this is a B, this is a C, this is a D. And rather than having it do math, I'd be like, here's a C essay, an example, and here's why. And then like put it in a bucket, depending on what you think it's most similar to. I think that would probably be better than like having it like do math and like reason, reason off over different like parts of the rubric, but not enough, not enough time for now, but I actually think that might work. You know what else I'm going to do? And I can grade this [1:07:43] But then there's also the like – [1:07:45] tweeted, it must have been two years ago now, asking for people's just favorite essays that are more than two or three years old. And it created a list of essays that I go back to

[1:07:55] like fairly often. And so maybe just dumping a bunch of those in and be like, this is the bar grade me against these. [1:08:01] That's really cool. Slightly different because there are different types of stuff than I write, but it'd be interesting to see how I do there. [1:08:06] That would be interesting because like, because your writing style, you're grading it against your own writing style. Right. And I wonder if that has anything to do with the kind of like a minus grade thing. Like if you put a totally different writer in there, how would it do? We should, we should try that. I want, I just want to see before, before we hop off this, I think we got to, I think we have to like go over the full distance and I'm going to take one of my essays and we'll see if I get an A minus on the Packy scale. [1:08:36] essay not against me also and just see if you get an a- okay um we're doing it for science folks um we have to know uh okay so i'm going to take uh let me see what my what are some recent things i've written so [1:08:58] I just wrote, this is every semester plan. I think this is good. I want to find something that's like not that good. [1:09:08] This wasn't that good. I don't think. [1:09:15] Yeah, I don't, I'm not super proud of this one. Okay, cool. [1:09:21] All right, please. I like that you're a more advanced AI user and you just copy the whole page and don't worry about it. I go to like the, you know, I need to stop doing that.

[1:09:33] I think we might be headed for another A- based on what I can see here. It looks A-. Oh, no, no, no. 14 out of 25 on Cultural Resonance. [1:09:41] Oh, it's 173 out of 200. This is the lowest score yet. [1:09:47] All right. But now now grade it without the rubric. Just go ask Claude to grade the essay. OK. All right. Give me your here. Let me just go back to. Yeah. Give me your what's your prompt for this? [1:10:02] Well, you can just do it. I'm Dan Shipper. I write a newsletter. I write every... Please grade this essay. That's it? [1:10:13] I'm [1:10:14] I sometimes do like you're the top editor. [1:10:18] you know yeah let's just let's just try this for now um we'll just we'll just go easy for now [1:10:25] Thank you. [1:10:27] A minus, B plus, A, A minus. [1:10:33] Oh, there it is. We're incredibly mad writers, both of us. I love it. So I think what we're learning is obviously there's this attractor state for A minuses in Claude. [1:10:57] to some degree. But if the rubric is based only on your own writing, it's going to be, I think, pretty likely to stay at that state. So like giving it a wide diversity of examples is probably going to be what, what gets you over the top. Yeah. I'm going to give that, I'm going to give that a try. That's my, my homework from this conversation, but it's also, I mean, like I will say, I do find it very valuable. Like within that rubric, if you, if you look there, I'm sure there's like a lot of just like little nuggets that will at least like change the way that I think about

[1:11:26] the writing because the way that I do a lot of my editing is I'll like go back and forth and make changes and whatever and then I'll hit a certain point where I'm like all right like I kind of know what I'm saying now let me just like go start fresh and I'll like copy things in but I'll write a lot of it fresh like near the end and do it again and I do think a lot of those like little tips that it gives kind of stick in my mind and so yeah. [1:11:48] I do think it's useful, although the grade is always an A-. The grade is almost a safety blanket for me, where if it were a C, I'd be like, fuck, I can't send this. No, totally. I mean, I think just having some of those rules printed out, it's stuff that you already know intuitively, but you might not have said to yourself or totally known in a rational way. And I think that in itself is super awesome and valuable. [1:12:12] For sure. Yeah. So, Paki, this was awesome. Thank you so much for coming. If people, after watching or listening to this episode, want to find you, where should they go to find you? At Paki M on Twitter and notboring.co. I have not sprung for the dot com yet. It's notboring.co. Awesome. Well, really, really, really fun to get to chat. Thank you so much for joining us. [1:12:36] Thanks for having me.

[1:13:06] that will leave you on the edge of your seat. [1:13:09] craving for more. It's not just a show. It's a journey into the future with Dan Shipper as the captain of the spaceship. So do yourself a favor. Hit like, smash subscribe and strap in for the ride of your life. [1:13:22] And now, without any further ado, let me just say, Dan, I'm absolutely hopelessly in love with you.