-
サマリー
activate_samplebutton_t1
あらすじ・解説
Note: This episode was supposed to come out a couple of weeks ago. But Andrew forgot to attach the audio file when he went to publish, so it's just been sitting in our podcast hosting software. Sorry about that. In this episode, Andrew and Sean dig into Paul Graham's founder mode essay that has gone viral in the past couple of weeks. Spoiler alert: Andrew is not a fan. Sean talks about the challenges of evolving client demands with Miscreants, while Andrew recounts his experiences in Mexico City and his decision to pivot ChartJuice towards a freemium model. Then they talk through some of Andrew's new AI-related business ideas, including automating marketing graphics and financial modeling for agencies, and his new co-founder and year-end goals.Links:Andrew’s Twitter: @AndrewAskinsAndrew's website: https://www.andrewaskins.com/ChartJuice: https://www.chartjuice.com/Sean’s Twitter: @seanqsunMiscreants: http://miscreants.com/StackWise: Coming soon...FigTree: Coming soon...For more information about the podcast, check out https://www.smalleffortspod.com/.Transcript:00:00.63SeanCan I start this podcast off by reading you a quote?00:06.34AndrewSure, I'm so curious.00:10.28Seanum Have you seen the Ted Chiang essay about Gen AI?00:15.59AndrewNo.00:16.55Seanokay um It was on The New Yorker. he okay so00:20.85AndrewFor a second, when you said JIN AI, I thought J-E-N, and I thought this was like some new AI company I hadn't heard of that named their AI JIN, as in short for Jennifer, and then I realized I'm an idiot, and you're talking about generative AI.00:34.68Andrewso00:35.09SeanWell, there's there's a startup called, I think, Jim, it's like a 25 year old like Korean person who Korean guy um and he like did like the rounds on YouTube for a little bit about how he like went from like broke to 100 million in net worth overnight because.00:35.41Andrewah00:41.20AndrewOh, that's cool.00:51.05AndrewJesus Christ. What?00:52.11Seanum I mean, it was like the dawn of like LLM stuff and jennna I had just started so.00:57.32AndrewYeah.00:57.44Seanadam00:57.71AndrewOkay. All right. Hit me with the quote.00:59.67SeanOkay, so let me just preface this. Ted Chiang, excellent sci-fi writer, um um ah wrote a really great book called Exhalation, one of my favorites.01:10.54Seanum01:10.83AndrewOh, I've been wanting to read that.01:12.94SeanYeah, excellent, excellent book.01:15.57AndrewCool.01:16.59SeanWould recommend, there's like a two page story in there that I really, really like. But, um You know, this this would have been great if i if I had it pull up, and I didn't.01:28.60AndrewHad it pulled up if you were prepared.01:31.18SeanYeah, yeah, yeah. But that's not how we do things. we you You know what? I can't find in this article. So I'm going to pull it up on a Slack thing.01:42.69SeanOK. um It's just the task that generative a generative AI has been most successful at is lowering our expectations. um And I was like, damn.01:54.53AndrewOh, interesting.01:55.78Seanhot take yeah uh01:57.03AndrewHow do, okay. So how do you interpret that?02:00.61Seanwell um i sent it as like a kind of honestly like uh02:05.49AndrewDo you agree?02:10.72SeanYeah, yeah, I do. um I don't agree with his essay. I think his essay about how his essay is about how Gen AI won't ever make art, and I think that's arguably true, um but I think that he misses the fact that Gen AI would be an art form, and people will use it the way that they make art with Photoshop and then collage things together.02:25.42AndrewSure, but at that point, it's not Gen AI making the art, I don't think. I think it's human artists making art using Gen AI as a medium or a tool.02:29.77Seanyou02:35.06SeanAgreed, agreed.02:35.66Andrewah like I think what he's saying is like the thing that Gen AI, at least in its current form, can't think.02:36.21Seanhence in02:43.90Andrewright it doesn't It doesn't produce creative thought.02:44.72SeanRight.02:46.78Andrewit's not It's not actually that intelligent, even though it's really good at looking intelligent.02:47.70SeanRight.02:53.73SeanRight.02:53.80Andrewum it just it repeats it guesses it and you know there's still this possibility that if we like scale it up enough we find out that that's all we do anyway and original thought is all just like guessing and repeating and suddenly it's doing what we can do but like we think that most likely we will need a new technological revolution like and a different form of AI technology paired with02:58.14SeanYeah. yeah03:13.96SeanSure.03:25.17Andrewpaired with or that replaces our current. Anyway, sorry, you know all this shit I'm preaching to the choir.03:28.49Seanknow um Well, to go back to the quote, I absolutely think it's true.03:33.93AndrewYeah.03:34.93SeanI actually think Gen. AI makes incredibly mid stuff. um Yeah.03:39.30AndrewYeah, for the most part. Yeah.03:41.52SeanBut I think we are always astounded by it because of the speed at which it can make incredibly mid stuff is impressive...