Skip to content
Dalton + MichaelDalton + Michael

When to Convince AI Skeptics (or When to Move On)

When should you help an AI skeptic see the value, and when should you move on? Dalton and Michael discuss how founders can navigate AI skepticism among employees, customers, friends, and family, and why a useful firsthand experience often does more than an argument. They explain how to make AI adoption part of your company’s culture, what to ask candidates about AI during interviews, and why customers care more about outcomes than the technology behind them. They also discuss helping the people you care about get started, working around skeptical bosses, and recognizing when convincing employees is taking time away from building your business. Discussion includes: showing instead of telling, why trying an old free model can create lasting skepticism, helping friends and family experience AI’s value, sharing useful workflows with your team, avoiding performative AI adoption, screening for AI use in interviews, setting expectations for employees, selling outcomes to skeptical customers, navigating bosses who block AI, and knowing when to stop debating and move on. — Dalton + Michael is brought to you by @Standard_Cap. Standard Capital is the AI-native Series A fund. Learn more at https://standardcap.com. — About Dalton: Dalton Caldwell is Co-Founder and Partner of Standard Capital. He spent 12 years at Y Combinator, where he served as Managing Partner, worked across 25 YC batches, and advised more than 1,000 startups. His investments include Whatnot, Brex, GitLab, PostHog, Stock Space, Rappi, Razorpay, and Oklo. Before becoming an investor, Dalton founded imeem and App.net. About Michael: Michael Seibel is a Partner Emeritus at Y Combinator, where he served as Managing Partner of the early stage accelerator from 2014 - 2024. Michael also serves on the board of three companies: Reddit, Dropbox, and Kalshi. He moved to the Bay Area in 2006, and was a co-founder and CEO of two Y Combinator startups, Justin.tv/Twitch (2007 - 2011) and Socialcam (2011 - 2012). In 2012, Socialcam sold to Autodesk Inc. for $60m, and in 2014, under the leadership of Emmett Shear (CEO) and Kevin Lin (COO), Twitch sold to Amazon for $970m. — Are you an AI builder? Check out StandardDB. Discover offers, credits, tools, and partner programs from the StandardDB ecosystem. StandardDB.com

Dalton CaldwellhostMichael Seibelhost
Oct 19, 202623mWatch on YouTube ↗

EVERY SPOKEN WORD

  1. 0:00 – 0:45

    Show, don’t debate: the Tesla self-driving analogy for AI skeptics

    1. DC

      What's cool is you just ride in a, in a Tesla, and it's a very safe driver. It sees all directions. It slows down. It's cautious.

    2. MS

      Yep.

    3. DC

      It's a great driver, man.

    4. MS

      Yep.

    5. DC

      And it's very hard for someone that's a skeptic to not grant that point.

    6. MS

      Exactly.

    7. DC

      [laughs]

    8. MS

      I, I love that 'cause it's like why, why debate? Just go to the Tesla store.

    9. DC

      Just, yeah, let's take a ride.

    10. MS

      Let's go, let's go to the store. [laughs]

    11. DC

      Let's try it out. [upbeat music]

    12. MS

      This is Dalton + Michael, and today we're gonna talk about how to sell AI to a skeptic. So we're in this interesting world right now where most likely if you're watching this video, you've been AI pilled probably, greater than a 50% chance.

    13. DC

      Or you clicked on the wrong video, I'm sorry.

    14. MS

      Sorry.

    15. DC

      [laughs]

    16. MS

      Sorry. Maybe the right- that may be the right video

    17. DC

      No, you're right, you're right

    18. MS

      ... you may be right

    19. DC

      ... but you're, you're in a weird part of YouTube-

    20. MS

      Exactly

    21. DC

      ... if you don't know anything about AI. Uh-

  2. 0:45 – 2:19

    Why the AI adoption gap feels unusually wide right now

    1. MS

      Yeah. But I think probably 99% of the world is still at AI skeptical.

    2. DC

      Yep.

    3. MS

      And I think that this kinda creates problems across the board, right? You might have employees who aren't adopting AI as they should. Um, you might have friends and family who aren't adopting AI and it's, like, killing you because it would help them in their lives. You might have customers who are like, "Oh, I don't- we don't trust AI. It doesn't," you know, da, da, da, da. And so we wanna talk about if you understand the potential of AI but these people around you don't, what can you do about that?

    4. DC

      I think this is especially pertinent because in my opinion, there's never been a wider gap between the people that are into this stuff and in the know and the mainstream. It's like in San Francisco, there's people like that work at the labs, they're, like, running their life with Claude. They are so far.

    5. MS

      Yes.

    6. DC

      And then there's, like, mainstream people which treat ChatGPT like a search engine at best.

    7. MS

      At best.

    8. DC

      And that's like, "Oh, I'm using it-

    9. MS

      Free ChatGPT. Free ChatGPT.

    10. DC

      Free ChatGPT as a search engine.

    11. MS

      Yeah. Yes.

    12. DC

      Um, and so there's a very wide gap between these-

    13. MS

      Well, by the way, I love how you're saying that's the limit 'cause-

    14. DC

      You're right

    15. MS

      ... there's also the people who don't use AI at all. [laughs]

    16. DC

      Or they're like, "It's unethical."

    17. MS

      Yeah. [laughs]

    18. DC

      Like, "I burn down a rainforest by asking ChatGPT."

    19. MS

      Yes, yes.

    20. DC

      And so it is my moral authority to-

    21. MS

      Exactly

    22. DC

      ... never ask it any questions.

    23. MS

      Yeah.

    24. DC

      I mean, it reminds me a lot when we were growing up, uh, we were teens I would say, when the internet came out, and if you were on the web, it was awesome, and most people had no idea-

    25. MS

      No idea

    26. DC

      ... that the internet was a thing.

    27. MS

      No.

    28. DC

      There was a very wide gap infr- between the people in the know and the people, you know, everybody else.

  3. 2:19 – 2:50

    AI as a force multiplier: bigger than past shifts like mobile or broadband

    1. MS

      But I would say that even that though, 'cause I'm thinking about in my life what were those moments, right? There was definitely internet and high-speed internet. There was definitely mobile phones. But none of them felt like the type of force multiplier that this feels like.

    2. DC

      Yeah.

    3. MS

      I think one of the reasons why I'm qualified to talk about this is that if you would've asked me in November of last year, I would've been like, "I am not an aggressive AI user." And, like, the difference between November of last year and February of last year is, like, massive, at least in my personal life.

    4. DC

      Yeah.

  4. 2:50 – 3:40

    Creating the ‘Uber moment’: the role of one unforgettable experience

    1. MS

      And so I think that there's a lot of people who just haven't had that moment yet. And, um, you know, one of the things I think a lot about is just, like, how do you get people to have that one moment? It's like all it takes is one moment, you know? Um, I remember this with Uber, right? All it takes was, like, calling an Uber, it showing up, your friend getting in the Uber with you, and then the big kicker was when you got out of the Uber and, like, no cash, right?

    2. DC

      They didn't have to tell you the-

    3. MS

      Yeah

    4. DC

      ... the credit card machine didn't work. [laughs]

    5. MS

      Exactly. Exactly. And, and, uh, and it was like... But describing Uber to someone never worked as well. [laughs]

    6. DC

      Correct.

    7. MS

      Never worked as well. So how, how do you think about this?

    8. DC

      I always think it comes down to real personal experience-

    9. MS

      Yeah

    10. DC

      ... where someone personally experiences the value-

    11. MS

      Yes

    12. DC

      ... as opposed to abstract arguments.

    13. MS

      Mm-hmm.

  5. 3:40 – 5:26

    Don’t fight culture wars: avoid moralized AI debates and focus on utility

    1. DC

      And I think what's crazy with the current AI discourse is how much it's become like everything else in the world. There's two poles, there's two opinions, and if you're on one persuasion, uh, of, of your, um, political opinion or what have you, AI is horrible. And it's funny, like I, I still, you know, use TikTok a lot.

    2. MS

      Mm-hmm.

    3. DC

      I don't know, it's entertaining. It's funny how accusing someone of using AI in a video is like-

    4. MS

      Oh, is that a crime?

    5. DC

      ... a slur. Oh, yeah. If you use AI in any way-

    6. MS

      We use no AI in these

    7. DC

      ... there's, there-

    8. MS

      [laughs]

    9. DC

      You'll get comments being like, "How could you sell us out? You sold out your fans by using AI." Like, it's like a-

    10. MS

      Wow

    11. DC

      ... total moral issue to use AI in any way as a creator. I don't know. I, I think it's fascinating for me, um, just to witness this.

    12. MS

      What just scares me when you're in this situation where, like, the set of moral principles that you're thinking about could be so destructive to your life-

    13. DC

      Yes

    14. MS

      ... right? Like, if you don't adopt this stuff and your peers do.

    15. DC

      It's a big cell phone.

    16. MS

      Really big.

    17. DC

      Yeah.

    18. MS

      Really, really big.

    19. DC

      But, but, and they're, again, if I'm trying to steelman, um, their perspective, it's that they have the moral high ground. And unfortunately, like lots of things in the world, when people feel like they have the moral high ground, they'll go pretty far to ju- You know, like, the, it's, it's very hard to convince someone that-

    20. MS

      Yeah

    21. DC

      ... feels like that they have the moral high ground that perhaps there's a different way to look at things.

    22. MS

      Well, and to be clear, I think kind of underlying what you're saying, which I love, is like, don't get into that debate.

    23. DC

      Yes.

    24. MS

      Like, that's not the- [laughs]

    25. DC

      So again, where I'm going with this, this video topic is like, yeah, arguing in the comments of TikTok with someone-

    26. MS

      [laughs] Yeah. Yeah

    27. DC

      ... that it is not, like, a, a crime against humanity for someone to use AI in a video.

    28. MS

      Yeah.

    29. DC

      Yeah, that's probably not the debate to have.

    30. MS

      No.

  6. 5:26 – 6:39

    Hands-on proof: Tesla FSD demo as the pattern for convincing skeptics

    1. DC

      What we would suggest though is to give them that personal experience. And so one thing that I've done for a lot of folks, um, that may be, uh, not huge fans of Tesla for whatever reason-

    2. MS

      Mm-hmm

    3. DC

      ... is to just give them a ride-

    4. MS

      Mm-hmm

    5. DC

      ... and turn on full self-driving.

    6. MS

      Mm-hmm.

    7. DC

      Because they've been fed a steady diet of news about how dangerous full self-driving is, and everyone's dying and, and all these car accidents.

    8. MS

      Well, I heard that Tesla got within 13 feet of hitting someone. [laughs]

    9. DC

      It's pretty bad. So-

    10. MS

      [laughs]

    11. DC

      So there's a whole propaganda machine feeding people stories-

    12. MS

      Yes

    13. DC

      ... that full self-driving is not safe. Um, again, do your own research on this about what the, how safe it is relative to other, other types of driving. I think the data's pretty clear that it's safe. But what's cool is you just ride in a, in a Tesla- And it's a very safe driver.

    14. MS

      Yeah.

    15. DC

      Like, if you sit there in the passenger seat with the latest version of full self-driving, it sees all directions, it slows down, it's cautious.

    16. MS

      Yep.

    17. DC

      It's a great driver, man.

    18. MS

      Yep.

    19. DC

      And it's very hard for someone that's a skeptic to not grant that point.

    20. MS

      Exactly.

    21. DC

      [laughs]

    22. MS

      I, I love that 'cause it's like why, why debate? Just go to the car and take a-

    23. DC

      Just, yeah, let's go take a ride

    24. MS

      ... let's go, let's go to the store [laughs]

    25. DC

      Let's try it out. It's much easier than debating Elon Musk.

    26. MS

      Yeah [laughs]

    27. DC

      [laughs] Right? Like-

    28. MS

      Yeah

    29. DC

      ... which of those is an easier thing to do?

    30. MS

      Hunt on that.

  7. 6:39 – 7:38

    The ‘I tried AI once and it was bad’ problem: recency and free-tool bias

    1. MS

      In my experience, when I'm trying to help people along this path, the biggest thing they tell me is, "I tried something once, and it was so bad." And the two things that almost always happen is they tried something, like, a year and a half ago. Like, the first time they were promised AI was changing the world, they tried something and it hallucinated, and the second thing is so common is they're trying free tools.

    2. DC

      Yeah.

    3. MS

      And I think that a lot of people still feel like they live in the world where most of the best things on the internet are free, and this is one of these weird non-physical things on the internet that is 100 times better when you're paying, and that is, in a scary way sometimes, a function of how much money you put into it. I, I can't think of other things-

    4. DC

      Well, it's because-

    5. MS

      ... like that

    6. DC

      ... it's easy to cherry-pick. Like, if you use a free model, especially an older one-

    7. MS

      Yeah

    8. DC

      ... and you tried doing certain things with it, it's so easy to cherry-pick a bad user experience.

    9. MS

      Yeah.

  8. 7:38 – 9:08

    Cherry-picking failures: TikTok “break ChatGPT” content and misleading conclusions

    1. DC

      Again, here's another TikTok reference. There's a really, I think, amazing TikTok creator.

    2. MS

      Okay.

    3. DC

      But he's a very anti-AI guy. But again-

    4. MS

      Yeah

    5. DC

      ... I love him.

    6. MS

      Yeah.

    7. DC

      He's, he's great, and his entire shtick is he uses free ChatGPT voice mode.

    8. MS

      Oh, God.

    9. DC

      And he asks it-

    10. MS

      Oh, like dumb questions?

    11. DC

      He'll... Well, no, he finds ways to break it. So he'll be like, "Hey, um, I'm running a marathon. Uh, can you, can you record the marathon for me and time how fast I'm going?"

    12. MS

      Yeah.

    13. DC

      And it's like, "Yes, I'd be happy to do that." And he's like, "Okay, I've started." And he'll be like, "Okay, I'm done. Uh, how fast did I go?"

    14. MS

      Yeah.

    15. DC

      And it'll be... It, it basically can't time things.

    16. MS

      Oh.

    17. DC

      So he found a bug in ChatGPT.

    18. MS

      Oh.

    19. DC

      And it's... These videos are hilarious. They have millions of views.

    20. MS

      Yeah.

    21. DC

      He does another one where he's like, "Hey, um, can you bark like a dog? Um, 'cause, like, I need someone to bark like a dog. There's someone that broke into my house."

    22. MS

      [laughs]

    23. DC

      And it's like, "Sure, I'd be happy to do that."

    24. MS

      [laughs]

    25. DC

      And it's like, "Arf, bark."

    26. MS

      [laughs]

    27. DC

      "Bark." Like, it's like just-

    28. MS

      Yeah. Not good

    29. DC

      ... these are really good videos.

    30. MS

      Yes.

  9. 9:08 – 11:16

    A practical tactic: share a paid account and collaborate on a real task

    1. MS

      I, I, I get it, but once again, like, if you, if your beliefs are significantly impacting your future livelihood, be careful with those beliefs. One way that I've tried to break through this problem of, "Hey, I tried it, it didn't work," um, I use an AI product called Manus, and one of the things that makes it really easy to do, and I'm sure other products can do this too, is collaborate on a task. And so it's really easy for me to effectively share my Manus account with other people, where I pay for it. Those people get to use all the skills that I set up and get to try fun stuff. And so I've tried this with someone in med school, and it was incredible. Like, it was a f- literally just started med school and was, like, doing anatomy, and just instantly was like, "Well, if I don't have to worry about cost, it's already set up, give this a try," and just started putting in, like, "I need study guides, I need this, I need that. Make it into a website." Just, like, you know, almost like there's no way it could do it, and then just, like, bing, bing, bing, bing, bing. And I didn't do anything other than share an account. Uh, another person I shared, uh, AI with was a, a screenwriter. Super skeptical.

    2. DC

      Yep.

    3. MS

      Right?

    4. DC

      Of course.

    5. MS

      Super skeptical. And then-

    6. DC

      He's a turncoat to use AI, right, to his industry

    7. MS

      ... and I saw he just started having dialogues about, like, screenwriting, which in my mind is hilarious 'cause, like, hand me, like, a script and ask for critical feedback and I'd be completely lost in every way. And he's having this back and forth, and suddenly he's like, "I lo- I like talking to this thing. It's making my ideas better." It turns out that there's a lot of things in screenwriting where, like, you know when you watch a movie about something you know and they get all the lingo wrong-

    8. DC

      [laughs]

    9. MS

      ... 'cause the person who's, like, writing it doesn't know your industry? AI does.

    10. DC

      Yeah.

    11. MS

      You can actually make the lingo. And so it was interesting kind of sharing this tool. I'm trying it with someone in government right now, where literally, like, I was given these, like, gnarly spreadsheets, and I was like, "I'm just gonna put this in here and we can ask questions." In a meeting I'm, like, asking questions, and they're like, "I wanna do that."

    12. DC

      Yeah.

  10. 11:16 – 12:40

    Inside companies: normalize AI via concrete workflows (not slogans or performative metrics)

    1. MS

      And it's like... So to the point about getting people to get on a side project, sometimes just sharing your AI account, sharing a task, doing a collaboration is a big win. But I kinda wanna get into this from a company perspective. So there are a lot of people in companies that, for whatever reason, are not AI pulled. And if you're a founder, um, I think you kind of have this dilemma of how do I convince someone that they could be 10 times more effective if they learn how to use AI? How do I train them to use AI? Should I have to train them to use AI? W- how do you think about this challenge? Because, like, I think in a tech company today you would want everyone in y- like-

    2. DC

      Yeah

    3. MS

      ... in your company using it

    4. DC

      ... I mean, one of the fixes a couple years ago was to do token maxing, where you would make a leaderboard and track which employees were using AI.

    5. MS

      Yes.

    6. DC

      And in some ways that worked. In other ways that incentivized performative use of AI, which is tricky. But- Effectively, you have to normalize that this is how we do business, and you have to normalize that not using it, you're just sort of hobbling yourself. And I think the most effective thing to do is to be forced to share specific workflows and value adds-

    7. MS

      Mm-hmm

    8. DC

      ... and not in the abstract, "Our organization is now using AI for everything. Go." 'Cause people will build really silly stuff.

  11. 12:40 – 13:38

    Hiring and management: bake AI competence into the ‘front door’ and expectations

    1. MS

      Yes. I think that's important. I would say another interesting place to attack is interviews. I think that, um, at this point, if I were hiring in a startup in any role, I would have a set of questions around how you're using AI, and if those answers are not exciting and interesting, I would strongly question to make- whether to make that hire. And I think that's so important, like if you wanna have a culture of people using AI, but the front door is letting everyone in-

    2. DC

      Yeah

    3. MS

      ... you're making your life harder. Um, I'd also say this, and I think this is the uncomfortable point. At some point, your job is not to convince your employees this is a good tool. It's kind of like being like, "Hey, we're like a, uh, work from the office company," and employee's like, "I'm just gonna work at home." And then you're like, "Well, let me convince you of all of the great [laughs] advantages-

    4. DC

      You're right

    5. MS

      ... of working from the office." Um, I, I think at some point you might have to put your foot down.

  12. 13:38 – 16:05

    The ‘great sorting’: why non-adoption can compound inequality (family, kids, careers)

    1. DC

      Yeah. Well, and I think, let me underline the point you're making, man, which is I think we almost have an imperative to try to teach the people that we know and the people that we care about-

    2. MS

      Yes, yes

    3. DC

      ... and teach our children-

    4. MS

      Yes

    5. DC

      ... because if we do a poor job of this, it's gonna make their lives harder.

    6. MS

      Yes.

    7. DC

      Like, like there's a great sorting happening, and you're not doing folks any favors by not helping them. Right, man? Like, that's what you're saying, and you're right.

    8. MS

      Well, and I remember this with using computers in school. I think there was a set of kids, and this goes back to show how old I am, there was a set of kids who had computers at home, and there was a set of kids who only had access to computers at school, and like computers were expen- I mean, they still freaking aren't-

    9. DC

      Yeah

    10. MS

      ... but they, they were expensive, and that one thing, I mean, if you fast-forward 20 years, every single job that those kids had-

    11. DC

      Yeah

    12. MS

      ... for the most part-

    13. DC

      How, how-

    14. MS

      [laughs]

    15. DC

      ... expensive w- was that to grow up in a computer skeptic household?

    16. MS

      Holy crap.

    17. DC

      Right?

    18. MS

      Oh, I mean, and like-

    19. DC

      And we're just talking computers, but like, can you do any- like, can you farm without computers now? [laughs]

    20. MS

      Yeah.

    21. DC

      Like-

    22. MS

      And so I think that's what really scares me is that like there is this deep harm if you don't keep up with the advances, and like that's not spoken about, right? What scares me is like this advance maybe gives you the power... It's like a weird analogy. Like, okay, this advance, depending how much money you spend, gives you the power of deploying hundreds of smart people on a task.

    23. DC

      Yeah.

    24. MS

      Imagine that you had two people to hire, and one person, for relatively little money, could deploy hundreds of people, and the other person couldn't.

    25. DC

      It's pretty uneven. So-

    26. MS

      Yeah

    27. DC

      ... so again, I guess what we're saying, and this reminds me, I remember I spent a lot of time and effort-

    28. MS

      Yeah

    29. DC

      ... showing my friends and family how to use Google.

    30. MS

      Yeah.

  13. 16:05 – 19:00

    Selling to AI-skeptical customers: sell outcomes, then demonstrate speed and capability

    1. MS

      Thank you, Ted. [laughs] That was very helpful. All right, so let's think about AI skeptics in the context of sales and context of working with a customer. I imagine that there are a lot of situations where you know AI can do this, and your customer is a skeptic. How do you navigate that?

    2. DC

      You sell them the benefit and the outcome and not the product, and there's been a lot of startups being super successful with this, like Harvey and Lagora, where, yeah, it's, it's AI, but what they're trying to do is sell almost like a services component. They're selling what you get if you buy it-

    3. MS

      Mm-hmm

    4. DC

      ... as opposed to a tool that they have to use. And so again, I would just focus on the shiny prize you get from adopting AI and not the technical details of what the thing does.

    5. MS

      Yeah. I, I would extend that to say I think that using AI you can deliver a speed that makes any customer start guessing things.

    6. DC

      Mm.

    7. MS

      So like for example, the chances that you actually have a fully functional prototype between the first meeting and the second meeting in the pre-AI world, nonexistent. Usually it's like a general spreadsheet to a more specific spreadsheet. [laughs] If you can basically be like, "Hey, you know that thing that we talked about and da, da, da? Here it is," that gets people asking different questions. If you can change the delivery timeline-

    8. DC

      Yeah

    9. MS

      ... that gets people like, "Oh, every other vendor says this is a one-year project. Oh, like we, we can get this to you in a month," that gets someone to start asking questions. I would also say like if you can take on work that other people wouldn't, right? Like, "Oh, we're happy to do this, but like we don't wanna build all these old integrations, or we don't wanna do all this kind of shit work." Like-

    10. DC

      Yeah

    11. MS

      ... if you can take on shit work, so big. And then the last one that I've seen work so well is like dashboards and graphs and stuff. Like AI can just f- just-

    12. DC

      Oh, yeah

    13. MS

      ... just-

    14. DC

      You can one shot all that stuff

    15. MS

      ... just, oh my God.

    16. DC

      It's great. Yeah.

    17. MS

      And so just like infinite observability, you can use AI in meetings to answer your customer's questions about the product or da, da, da, and it's just a completely different experience from, "Oh, those are great questions. I'm gonna take it back to the team, and we'll come back to you a week later." So it's so, one, I agree with you, not selling the AI, but two, it's like you can almost use it- ... in front of them, with them, and just changes the tenor of the meeting. So I've seen people go from skeptical to pitching their boss based on that [laughs] very, very quickly.

    18. DC

      And maybe one point about the boss thing, um, I think if you work at a place that is AI skeptical and your bosses and upper management is-

    19. MS

      Uh-oh

    20. DC

      ... probably, like, find a new job.

    21. MS

      Mm-hmm.

    22. DC

      Right? Like that's, again, in terms of the Great Sorting that is happening, um, if you're into this stuff and you face constant roadblocks for management, man, is that a good reason to take your services elsewhere, right?

    23. MS

      Well, and to extend that point, every month that you're prevented from using AI in this company is a month you're not learning.

    24. DC

      Right.

    25. MS

      You're not growing.

    26. DC

      It's bad.

    27. MS

      Really bad.

    28. DC

      [laughs]

  14. 19:00 – 22:20

    When authority blocks AI: rules vs. self-improvement, and the student/professor case

    1. MS

      Really, really bad. I wanna talk about bosses. I wanna talk about security. Um, I even heard this in the context of professors. A student told me, "I can't use my, this to help me write my senior thesis 'cause my professor won't like it." This was at a school that costs hundreds of thousands of dollars. [laughs] What do you do? What advice would you give when people tr- can't just leave, and the authority figure is either explicitly preventing it-

    2. DC

      Yeah

    3. MS

      ... you know, "We have these security rules," or-

    4. DC

      I think-

    5. MS

      ... culturally preventing it?

    6. DC

      I think you should follow the rules, and you should do what they say, but they can't control with what you do with your own time, and they can't control what you do with your mind. And as a student, whether in high school or college, most of the cool stuff I did was kinda on my own time [laughs] versus just for the classwork.

    7. MS

      Mm.

    8. DC

      And so again, yeah, I don't think you should try to hack the system and get in trouble by actually using AI. I don't know, maybe you should do that, but that probably wouldn't be my first choice.

    9. MS

      Mm-hmm.

    10. DC

      But man, in your extracurricular time, I would be going really deep on using this stuff, and there's nothing they can say or do about that.

    11. MS

      So I'm gonna, I'm gonna take the opposite position.

    12. DC

      Okay.

    13. MS

      Yeah, I think you should... Break the rules is, like, a hard thing to say. I'm not saying break the rules. I'm saying I don't think rules are written particularly well. I don't think-

    14. DC

      I, I knew this about you. Yeah.

    15. MS

      [laughs]

    16. DC

      I, I, I saw this going, Michael. I understand.

    17. MS

      I think, you know, for example, I don't think a workplace often conceives of the idea that you would spend money on your own AI account to the benefit of the company, right? Like, I-

    18. DC

      Right. Again, I think maybe we're on the same page

    19. MS

      ... I don't think no one's written that policy. [laughs]

    20. DC

      This is a distinction without a difference. Yes, you are, as long as you understand what the rules are and you comply with them-

    21. MS

      Yes

    22. DC

      ... you'd be surprised at, at how much you can do with your own budget and your own time and things like that. That's a way to say it.

    23. MS

      Exactly, and, like, it was interesting. I was talking to this student, and you know, the professor, I understand. It's like the professor doesn't want you to write a 50-page senior thesis that AI wrote.

    24. DC

      Correct.

    25. MS

      That's like painfully obvious. [laughs]

    26. DC

      You don't wanna upload a ChatGPT 50-page paper. That would be bad.

    27. MS

      Yes. [laughs] And the professor already doesn't wanna read your paper.

    28. DC

      Yeah.

    29. MS

      But the fact that AI wrote it is, like, kind of an insult. But I was having this funny conversation with a student because they were like, "So, so kids aren't using AI?" And I'm like, one, they definitely are. Two, I'm sure papers have changed in the last two years.

    30. DC

      Yeah.

  15. 22:20 – 23:42

    Wrap-up: engineer the ‘self-driving car’ moment—and know when to move on

    1. MS

      ... you can research all your custo- all of your competitors. You can research the customer way better. Like, you can come to that meeting way more prepared on your own dime. So I would basically be like, if you're blocked from using the best models at work and you're not spending that money on your own dime, that's like the parents not putting a computer in the house. Like, you're only hurting yourself. So yeah, so you know, maybe to wrap this up, I think that if you are AI skeptical watching this, right, we don't wanna engage in all of the-

    2. DC

      I don't think they're still watching.

    3. MS

      Oh, shit, okay.

    4. DC

      [laughs]

    5. MS

      If you made it this far, give it a try. If you know somebody, I, I agree, don't engage in the political debate. Don't engage in the theoretical debate, right? Like, figure out some crack where you can get them to use it, some equivalent to putting them in the self-driving car. And then if you're at a company and you're sitting here and you're spending too much time asking yourself, "What do you do about these people that refuse to use AI?" that's, yeah, a problem. Um, and I think the scary thing is that, like, that's preventing you from spending your time thinking about how we can use AI to win. So be careful with that. Like, your responsibility only goes so far. All right? Good chat.

    6. DC

      All right. Thanks, Michael. [upbeat music]

Episode duration: 23:42

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

Transcript of episode xY7GrFh9wRg

Get more out of YouTube videos.

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

iOSAndroidClaudeChrome