Skip to content
Dalton + MichaelDalton + Michael

Slop vs Craft

As AI tools become increasingly powerful, its easy to output a lot more output... but that doesn't mean the output is good. In this episode of Dalton + Michael, the two discuss the temptation of using the tools to create low quality software/content/startup ideas. They mention some case studies from the past, including SEO content farms as well as the risk of building a "turkey startup." There is also a higher risk for founders that being able to output reasonably sloppy prototypes has the side effect pushing more startups into pivot hell. Having high standards for what you put out in the world is a great signal for what good "taste" looks like. The way to win in a slop war is to not play, and to instead use these new tools to help solve problems for your customers better than anyone else. If you would like to see the "turkey graph" for yourself, here it is: https://x.com/paultoo/status/2028601344534454598 Dalton + Michael is brought to you by @Standard_Cap Dalton Caldwell on X: https://x.com/daltonc Michael Seibel on X: https://x.com/mwseibel

Dalton CaldwellhostMichael Seibelhost
Mar 2, 202618mWatch on YouTube ↗

EVERY SPOKEN WORD

  1. 0:00 – 0:33

    Avoid the symmetric “slop harder” arms race

    1. DC

      I think you wanna stay out of a symmetric battle for who can slop harder. If there's a competitor and they're spamming a lot, and you're like, "Well, our only solution is to spam harder"-

    2. MS

      [laughs]

    3. DC

      ... and you're just in an arms race of who can create the most garbage to make the graph go up-

    4. MS

      Yeah

    5. DC

      ... I, I think don't do it.

    6. MS

      Yeah.

    7. DC

      I think you have to find a way to stay out of trying to produce the most low-quality crap. [upbeat music]

    8. MS

      This is Dalton + Michael. Today, we're gonna talk about slop versus craft.

    9. DC

      [laughs]

  2. 0:33 – 2:06

    Why Claude Code feels powerful—and why that’s dangerous

    1. MS

      Maybe another way to describe this is like what does it mean to have good taste? Why don't we kick this off, you and I were talking about Claude Code. You've maybe gone down the smallest of rabbit holes. How is Claude Code changing how you think about slop versus craft?

    2. DC

      Sure. So I think for context, I've been doing a fair amount of programming over the past, uh, nine months.

    3. MS

      Mm-hmm.

    4. DC

      Um, like I was messing with Cursor and Windsurf a while ago-

    5. MS

      Mm-hmm

    6. DC

      ... and it was cool, and, you know, I've been having fun with it. And then of course, Claude Code came out probably a while ago, but it, it really took off memetically-

    7. MS

      Mm-hmm

    8. DC

      ... Q4 of last year, and definitely over, like, Christmas break is when everyone-

    9. MS

      Yeah

    10. DC

      ... went nuts with it. And what I would say is it's really fun.

    11. MS

      [laughs]

    12. DC

      [laughs] It's like a video game.

    13. MS

      [laughs]

    14. DC

      And you, like, type stuff in, and it just feels powerful.

    15. MS

      Yeah.

    16. DC

      Like, you watch the screen fly by. It's like programming with the best programmer you've ever met-

    17. MS

      Mm-hmm

    18. DC

      ... who's, like, really nice and really fast.

    19. MS

      [laughs]

    20. DC

      And it's so energizing to watch the code fly by on the s- on the screen and, um, see all the tokens it's burning and just, like, you know, it's, it's super cool, so I would recommend it.

    21. MS

      Yes.

    22. DC

      And this is not yet another, you know, VC podcast talking about how great Claude Code is-

    23. MS

      [laughs]

    24. DC

      ... 'cause here's where I'm going with this.

    25. MS

      Uh-oh.

    26. DC

      It is great.

    27. MS

      Is there a dark secret? [laughs]

    28. DC

      No, it is great, but I easily could imagine a younger version of myself-

    29. MS

      Hmm

    30. DC

      ... not really sleeping or taking care of myself-

  3. 2:06 – 2:50

    The Gentoo Linux analogy: “cool” work that creates no value

    1. DC

      It reminds me of, like, I remember setting up Linux when, uh, when I was super young.

    2. MS

      Mm-hmm.

    3. DC

      And I used this version of Linux called Gentoo-

    4. MS

      Okay

    5. DC

      ... where you would compile the whole thing from scratch.

    6. MS

      Mm-hmm.

    7. DC

      And so when you were setting up Linux, it would just be like, stuff would be flying by on the terminal.

    8. MS

      [laughs]

    9. DC

      And I felt like a really elite guy-

    10. MS

      [laughs]

    11. DC

      ... that I was, like, hand-compiling my own Linux kernel.

    12. MS

      Yes.

    13. DC

      And in retrospect, that was, like, not-

    14. MS

      [laughs]

    15. DC

      Like, I'm embarrassed-

    16. MS

      There's a lot of value being created there

    17. DC

      ... maybe I shouldn't be telling this story.

    18. MS

      [laughs]

    19. DC

      But, like, it felt really cool-

    20. MS

      Yeah

    21. DC

      ... to compile your own Linux kernel, when in practice, I- literally nothing beneficial happened with that.

    22. MS

      [laughs]

    23. DC

      Okay? And so where I'm going with this is there's such a temptation that even I felt-

    24. MS

      Yeah

    25. DC

      ... using these super powerful tools to basically make slop.

    26. MS

      Yeah.

  4. 2:50 – 3:41

    Defining slop: demo-friendly, fundraise-friendly, user-useless

    1. DC

      So let's define slop.

    2. MS

      Yes.

    3. DC

      Like, I'd like to hear your definition. Like, what, what... Yeah, what do we mean by slop here?

    4. MS

      Very simply in my mind, slop is products that don't actually help the user. They can seem great in your mind. They can seem great when you're demoing. They can seem great when you're raising money. But when we put it in front of the intended user, we do not see the user getting their problem solved and them being happy. It's like, "Oh, it's just missing one thing." [laughs]

    5. DC

      [laughs] That's a, that's a... It's the, it's the feature treadmill where you're like, "Oh, it's bad, but just let me add 10 more features."

    6. MS

      [laughs]

    7. DC

      You're like... And again, it's so tempting. The tools are so good.

    8. MS

      Yeah.

    9. DC

      You can just keep adding features to something that's-

    10. MS

      Yes

    11. DC

      ... kinda crappy and have a very large surface area very fast.

    12. MS

      Have a lot of crap.

    13. DC

      Yeah.

    14. MS

      Yeah. So that's kinda what I think about with, with slop. What do you think when, when you see slop?

  5. 3:41 – 4:37

    Dalton’s definition: slop is self-deception you recognize later

    1. DC

      By definition, this is subjective, and so one person... Like, you could say anything's slop, and people have debated, and that sounds very tedious.

    2. MS

      Yeah.

    3. DC

      I think the best definition is that in your own objective mind, you know it's not that good. [laughs]

    4. MS

      Hmm.

    5. DC

      Like-

    6. MS

      Hmm

    7. DC

      ... like, especially after you've, like, you know, taken a break and taken a walk or-

    8. MS

      Mm

    9. DC

      ... gotten some sleep-

    10. MS

      Mm

    11. DC

      ... and then you really look at what you're doing with fresh eyes or people that you trust, and you're like, "Actually, this isn't that good."

    12. MS

      Hmm.

    13. DC

      C- 'cause in the, in the, in the Linux example I gave, come on.

    14. MS

      Yeah.

    15. DC

      Like, the hand-compiled Linux kernel-

    16. MS

      Yeah

    17. DC

      ... was not actually better than the one I downloaded from Red Hat.

    18. MS

      [laughs]

    19. DC

      It's the same thing. And so I think you have to have this, like, objective perspective about if something is actually good. Basically, you can't practice self-deception. That's what I'm trying to say. Slop is where you're sort of actively self-deceiving-

    20. MS

      Yes

    21. DC

      ... that your thing is good-

    22. MS

      Yes

    23. DC

      ... when you kinda know it's not that good.

  6. 4:37 – 5:01

    Use the competitor’s product to get real signal

    1. MS

      You know, as an aside, whenever I work with a company and they say, like, "Oh, our competitor gets funded, or Newcoming gets funded," and they're super afraid, the first thing I do is just like, "Have you used the product?"

    2. DC

      Yep.

    3. MS

      "Have you used it?" Like, you can't even really determine whether something is good or bad by not using it. [laughs] Like, if you're taking all of the clues, but you don't try it, you're not getting any signal.

    4. DC

      Yep.

  7. 5:01 – 6:11

    Web 2.0 SEO farms: slop that prints money—until it doesn’t

    1. MS

      So we have some examples of slop, um, which are fun. Um, first one, Web 2.0 with the-

    2. DC

      Yeah. SEO spam. Okay, so the real insight when Google was taking off is that if you could rank number one in Google for something-

    3. MS

      Game on

    4. DC

      ... you, like, would make so much money.

    5. MS

      Yeah.

    6. DC

      Right? If you, if you could, like, spam your way for your e-commerce site to be the number one Google search result, you would, yeah, you'd make tons and tons of money. So guess what happened?

    7. MS

      Or content, and you put a little Google ad on it.

    8. DC

      Right?

    9. MS

      Yeah.

    10. DC

      So guess what happened is a lot of people built slop farms-

    11. MS

      Yes

    12. DC

      ... to game Google-

    13. MS

      Yes

    14. DC

      ... so that they can end up at the top of search results and make money in various ways.

    15. MS

      And they had very nice early graphs.

    16. DC

      Yeah. They made a lot of money.

    17. MS

      [laughs]

    18. DC

      I'm not gonna, I'm not gonna name any of the names of these companies. You can ask ChatGPT about this.

    19. MS

      A lot of money. Not a lot of money relative to successful companies. A lot of money relative to-

    20. DC

      It looked like it was working.

    21. MS

      Yes, it looked like it was working.

    22. DC

      So they were spamming the hell out of Google.

    23. MS

      Yes.

    24. DC

      And some raised, like, VC money for this.

    25. MS

      Yeah. Yeah, yeah, yeah.

    26. DC

      Um-

    27. MS

      Many, I might argue. [laughs]

    28. DC

      Ultimately, none of those are going concerns anymore. Like, it was, like, a, a brief moment in history where you could spam Google and make lots of money, and then you were done. Uh, this, this is something I've been calling it with PB.

    29. MS

      Mm-hmm.

  8. 6:11 – 6:34

    “Turkey startups”: the growth graph that ends suddenly

    1. DC

      He, um, he calls them turkey startups. Have you ever seen the graph of the life of a turkey?Where it's, like, getting better and better and better, and then one day it's over?

    2. MS

      [laughs]

    3. DC

      That's a turkey startup. It's a turkey graph.

    4. MS

      Oh, man.

    5. DC

      And you know what I'm ta- you know what I'm talking about, right?

    6. MS

      Uh, yes.

    7. DC

      It's so-

    8. MS

      It's a little brutal. [laughs]

    9. DC

      It's like where you're like, "Oh, you know, my startup's going great," and then it's just, like, over.

    10. MS

      Boom. Google changed the algorithm.

    11. DC

      Yes.

    12. MS

      [laughs]

    13. DC

      So an SEO slop company is the definition of a, of a turkey-

    14. MS

      Yes

    15. DC

      ... startup.

  9. 6:34 – 7:58

    Mobile consumer slop: subscriptions, forgetfulness, and early App Store junk

    1. MS

      For me when I think back to mobile consumer, there were so many products where when you would dig deep, someone would subscribe to them for a year and their activity would end after 30 days. And, like, the business model was, like, okay, like, I love this business model. It's like, all right, we're gonna acquire the user for this much. If we get them to buy the annual subscription, we al- we have one-year LTV, that's pretty good. And if we get them to, like, somehow forget and subscribe for two years, well, I mean, that's, uh... And then we'd run that play across 100 million users-

    2. DC

      Yep

    3. MS

      ... and all the graphs look great.

    4. DC

      Well, and remember, like, the early App Store, it was a lot of slop.

    5. MS

      Yes.

    6. DC

      It'd be like iBeer.

    7. MS

      Yeah.

    8. DC

      There was the-

    9. MS

      Lightsaber.

    10. DC

      It was just like-

    11. MS

      Yeah

    12. DC

      ... crap.

    13. MS

      Yes.

    14. DC

      And people made money in the short term.

    15. MS

      Yes.

    16. DC

      But again, none of those became durable businesses because it was, it was slop, and Apple didn't want slop in the App Store either.

    17. MS

      No.

    18. DC

      Like, no one was that excited about this stuff.

    19. MS

      No.

    20. DC

      It was effectively a scam.

    21. MS

      It, it was a scam, and, and they were... Users understand when things are scams, so you get, like, a negative word of mouth-

    22. DC

      Yep

    23. MS

      ... and then, you know, you- But to your point, like the SEO thing, you look, in the first 10% of the race, you look like you're winning.

    24. DC

      Yep.

    25. MS

      And it's like, that's why it's, like, so interesting, 'cause it's like, man, it's so desirable to look like you're winning in the first 10% of the race, you can sometimes just forget that you have [laughs] 90% of the race to run. [laughs]

  10. 7:58 – 8:35

    The slop ecosystem: get-rich-quick shills and “courses about slop”

    1. DC

      I just, thinking about this, um, out loud, it's almost like the same shills that are selling courses were into all this stuff.

    2. MS

      Yep, yep.

    3. DC

      Like p- [laughs] Like, basically, on the, the dark corners of YouTube-

    4. MS

      Yep

    5. DC

      ... of, like, get rich quick.

    6. MS

      Yeah.

    7. DC

      They love this stuff.

    8. MS

      Yes.

    9. DC

      Like, everything we just mentioned, there were totally people that were like, "This is the secret to wealth."

    10. MS

      Yes, yes.

    11. DC

      And it was some version of, like, whatever the latest scam was-

    12. MS

      Yes

    13. DC

      ... to, like, fool people with slop.

    14. MS

      Get in early and get out with money before, you know-

    15. DC

      Well, they made money selling courses.

    16. MS

      Exactly, yeah.

    17. DC

      So again, that's the irony, is selling, selling content to explain to people how to make slop-

    18. MS

      Makes more money [laughs]

    19. DC

      ... is probably a better business than making slop, so.

  11. 8:35 – 11:06

    Crypto’s euphoric phase—and the incentive to build low-quality ICO plays

    1. MS

      So then crypto happened, which was slop-free, right? Right? [laughs]

    2. DC

      I mean, I think the, the, the thing with that is it was so intoxicating in the same way. Again, I-

    3. MS

      Mm-hmm

    4. DC

      ... you know, to speak to myself, when I first learned about, you know, Bitcoin-

    5. MS

      Yeah

    6. DC

      ... I thought it was super cool. I remember telling you about it.

    7. MS

      Mm-hmm.

    8. DC

      It was hard to think about anything else for a little while, this was, like, in 2013 and 2014, because it seemed so revolutionary.

    9. MS

      Yes.

    10. DC

      Again, it reminds me of the feeling that I've had messing with a lot of the vibe coding tools, like CloudCode.

    11. MS

      Yeah.

    12. DC

      It's just feeling like I'm seeing the future and feeling almost, like, a euphoria about it and an excitement.

    13. MS

      Yes.

    14. DC

      And because I've been through multiple cycles before, current Dalton is like, "Yeah. [laughs] Um, chill, dude." [laughs]

    15. MS

      Yeah. Ex- Yeah. Yes.

    16. DC

      Um, and that, and that's kind of where I ended up with the crypto stuff too, which is to realize there was an incentive to do low-quality stuff with crypto, like ICOs or, or to-

    17. MS

      Mm-hmm

    18. DC

      ... you know. There, there's a lot of things like that. And ultimately, you know, enduring value was not in- created in as many of those as was created by Coinbase and, and-

    19. MS

      Of course

    20. DC

      ... what have you.

    21. MS

      I think the, the moment that crypto really kinda jumped the shark for me, I remember I was working with a company in the batch who made something, you know, like a Discord plugin or something, and it was an interesting seed that something could've grown from.

    22. DC

      Yeah.

    23. MS

      And then they were, the, the plan was to ICO and monetize it and be done.

    24. DC

      Dump on retail. [laughs]

    25. MS

      And I was like, I was like, "You, you guys have something, like, you might be able to make some..." And they're like, "What are you talking about?" Like-

    26. DC

      That's not the game

    27. MS

      ... "No, that's not the game we're playing."

    28. DC

      Yeah.

    29. MS

      And I was just like, "Oh, come, oh, okay."

    30. DC

      Yeah. And, and again, to put a fine point on this-

  12. 11:06 – 13:04

    Taste as an honesty compass: high standards and positive-sum value

    1. DC

      Well, this is where it goes back to the thing I said earlier, which is you have to have your own good judgment.

    2. MS

      Yeah.

    3. DC

      And just bec-

    4. MS

      Mm

    5. DC

      ... like, if you're looking around and you're like, "Everyone else is scamming, maybe I should scam too"-

    6. MS

      [laughs]

    7. DC

      [laughs] That's not a great way to make life decisions. And so I do think this is where taste comes in.

    8. MS

      Yeah.

    9. DC

      Is that if we... You know, I'd love to hear your definition of taste, but in my definition here, taste just means having a strong barometer, having a strong compass of what is good and what is not good, and not producing things that you know in your heart are bad, right? Or, like, not helping people or hurting people.

    10. MS

      You make such a good point, because I feel like there's, like, a artistic definition of taste which, like, I'm not much of an artist. Like, I don't really, can't really engage with. And then there's just, like, a pure, like, value generation, right? Like, almost in, like, kind of the core of the economy, like, is what, is what you're doing positive sum? It's weird, because, like, I think in a classic definition of taste, people are like, "Well, some people have it and some people don't." I think everyone has the ability to look at something honestly-

    11. DC

      Yeah

    12. MS

      ... and ask whether it's, uh, positive sum.

    13. DC

      Well, and again, the irony, I'm, you know, you've heard me say this before, but anyone that insists what good ta- taste they have-

    14. MS

      Yeah. [laughs]

    15. DC

      [laughs] It's like definitionally-

    16. MS

      Yeah

    17. DC

      ... that's how you spot the people that don't have good taste-

    18. MS

      Yeah

    19. DC

      ... is they, they talk about how great they are.All the time.

    20. MS

      Well, and I think that's so funny 'cause like, like when people are like, "Oh, taste is like where the buttons are on the screen, or like how cool my dark mode is," it's like just-

    21. DC

      Yeah

    22. MS

      ... like, you know, like...

    23. DC

      But to recognize great taste in other people-

    24. MS

      Yes

    25. DC

      ... it's people that have high standards.

    26. MS

      Yes.

    27. DC

      And that they really care-

    28. MS

      Yes

    29. DC

      ... about anything that they release, they're proud of.

    30. MS

      Yes.

  13. 13:04 – 15:56

    Winning the current “slop war”: focus on retention, not extractive growth

    1. MS

      Like at the end of the day, value has to be created to the user. Interestingly enough, one would argue we're in a slop war right now.

    2. DC

      Yeah.

    3. MS

      And maybe one would say like this is what happens in any new platform kind of zone, right? Like, you got a, a lot of easy to slop out things, and then things kinda get sorted out.

    4. DC

      Yeah.

    5. MS

      How do you think about the slop war? Who wins?

    6. DC

      I would almost compare it to the Facebook platform app wars.

    7. MS

      Yeah.

    8. DC

      Where a lot of people-

    9. MS

      Yes

    10. DC

      ... were competing with Zynga, Zynga was buying people.

    11. MS

      Yes.

    12. DC

      They were competing with Farmville.

    13. MS

      Yes.

    14. DC

      And ultimately, no one won that was playing that game.

    15. MS

      I mean, Facebook won. [laughs]

    16. DC

      That's true.

    17. MS

      [laughs] Yeah.

    18. DC

      Um, and so yeah, the people selling tokens-

    19. MS

      [laughs]

    20. DC

      ... were like, "Yeah, this is great," like-

    21. MS

      Slop's great. [laughs]

    22. DC

      Everyone should use cloud code-

    23. MS

      Yeah. [laughs]

    24. DC

      ... to make, like, whatever they want.

    25. MS

      Yes. Yeah. [laughs]

    26. DC

      But if you are the people consuming all these tokens-

    27. MS

      Mm

    28. DC

      ... I think you wanna stay out of a symmetric battle for who can slop harder. If there's a competitor and they're spamming a lot, and you're like, "Well, our only solution is to spam harder"-

    29. MS

      [laughs]

    30. DC

      ... and you're just in an arms race of who can create the most garbage to make the graph go up-

  14. 15:56 – 17:43

    Claude Code can accelerate pivot hell: one-shot prototypes reduce conviction

    1. MS

      Yeah. I think the last thing I like to talk about when it comes to slop, and why I would encourage kinda early stage founders to be very careful, is pre-slop, there was always, there was already kind of a high likelihood of people getting to pivot hell.

    2. DC

      Yep.

    3. MS

      What I worry about is, like, cloud code and slop might accelerate that already pretty aggressive trend.

    4. DC

      Well, 'cause you can one-shot anything.

    5. MS

      Yeah.

    6. DC

      So basically if you're, if you're low conviction and you decide to pivot, you can just-

    7. MS

      Have the new thing

    8. DC

      ... send in a TechCrunch article and say to clone it.

    9. MS

      Yeah.

    10. DC

      You can be like, "Oh, like clone this new startup." And so it's even more tempting-

    11. MS

      Yes

    12. DC

      ... to just randomly change your idea over and over again-

    13. MS

      Yes

    14. DC

      ... and to build plausible prototypes. Again, I'm, I'm saying these are like pretty good.

    15. MS

      Yeah.

    16. DC

      Like, it's good. I like cloud code.

    17. MS

      Yeah.

    18. DC

      And so it's like we're giving-

    19. MS

      [laughs]

    20. DC

      You're getting a really powerful weapon, like, don't hurt yourself with it, right?

    21. MS

      Yeah, yeah, yeah. [laughs]

    22. DC

      I think it's easier-

    23. MS

      Yes

    24. DC

      ... to, because you can crank out a really good prototype so fast-

    25. MS

      Yes

    26. DC

      ... to have lower conviction on what you should be working on-

    27. MS

      Yes

    28. DC

      ... and to constantly change your mind-

    29. MS

      Yes

    30. DC

      ... to seek the next shiny thing.

  15. 17:43 – 18:23

    Closing: don’t play the slop war; turkey graphs come down

    1. MS

      In conclusion, don't play in the slop war.

    2. DC

      [laughs]

    3. MS

      Figure out how to focus on creating value for users, and maybe be a little bit less addicted to graphs that go up but where no value is being created. I love graphs going up.

    4. DC

      Me too.

    5. MS

      Right? Like, graphs should go up, right? But like graphs that go up where no value is created, come down. They've, they have a life of a turkey. [laughs]

    6. DC

      [laughs]

    7. MS

      Great chat.

    8. DC

      All right. Thanks, Michael. [outro music]

Episode duration: 18:25

Install uListen for AI-powered chat & search across the full episode — Get Full Transcript

Transcript of episode YzHyRZMBs04

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.

iOSAndroidClaudeChrome