Introduction and Warm-Up

Raise your hand if you can hear my voice. Keep it up. Okay. Okay.

Now high five your partner. Go ahead and high five your partner. Okay.

So for those who are just joining, you've missed half the fun. So sorry. That was the most important thing we're going to talk.

No, I'm kidding.

Training Imagination in the Age of AI

So what we're doing is we're trying to spark our imaginations collectively. 1We recognize that in the age of AI, your imagination is your primary limitation.

And when Thomas Edison was asked how how to be a great inventor, you know what he said? Train your imagination. And he said a few things, but that's like the part that matters here.

I think most, I would say 90 % of people don't think the imagination can be trained. They think, oh, it's either you got it or you don't. No, that's not true. Training, you can actually build the skill of imagining.

Invite the Unexpected to Spark New Ideas

And one way you do it is by opening your mind to unexpected inputs. The imagination is triggered by the unexpected.

And there are lots of different tools and techniques for that one amazing one that's gonna sound crazy is talking to strangers okay and by the way if you're here amongst this crew you're not really strangers but talking to

strangers is a great way and all of us who live our little tidy lives we never talk to anybody and we wonder why we don't aren't imaginative see above start talking to strangers okay so what i

Brag on Your Partner: A Listening Exercise

want to do and we're going to do this a couple times but what i want to do is if you heard heard something that you go, no way, that's so cool. I've never thought of that. I want you to brag.

Raise your hand if you heard something that you thought was really cool from your partner, not said something cool, but heard something cool. Hey, like four people. That's great. Okay.

The Cognitive Bias: Why Sharing First Helps You Listen

By the way, as an aside, we've studied this. What research suggests is that folks who share first tend to be more impressed with their partner than folks who share second.

you know why this is there's a lot it's not because they're men you know or or because they're engineers or whatever it's simple reason I have four daughters by the way so I hear a lot about what's wrong with daddy I what our research finds is somebody who's sharing second lacks the sufficient brain space to attend to their partner the person who shares first gets their mind empty and And then they're available to listen. And the person who's available to listen generally finds their partner quite interesting.

But the person who's going to go second, the whole time their partner's talking, they're going, is that an interesting story? Am I going to look good when I say that? Am I smart? Am I thoughtful? Right? And they're incapable of finding their partner interesting. So that's like an interesting cognitive bias. Right?

So chances are, our research suggests something like 90 % of people who raised their hand actually shared first, which is fascinating. fascinating it's because you were actually available to listen to your partner um you don't now now that you know that you can actually use that as a superpower just choose to listen rather than keep thinking about yourself but you didn't know that before now you know it so i would love

Stories From the Room: Real-World AI Augmentation

yannick you want to pass the mic to one of these folks who raised their hand and i want you to brag like be like a hype man you remember like in the hip -hop days you had the hype man is like you you you check this guy out i want you to do that okay i know it's going to be a stretch but you You raise your hand, will you hype your partner please? Then stand up if you want.

Can you hear me? Yes, sir. I'm not very good at hyping things. I'll do it, yeah, yeah, yeah, yeah, yeah, yeah, yeah, here we go.

The interesting part he told me is when his company got acquired and reacquired and reacquired and then he moved to Parag, Parag, yeah. He tried to set and get things working there I thought it's very good with the changes that he's seen but still going on with, I'm working with them. I thought it's very good.

Yeah, maybe that should be the, yeah, yeah, yeah. Yeah, yeah, yeah, yeah, yeah, okay, okay, great. Thank you, thank you, give it up, give it up. That's great.

Okay, now refine the search parameters. Who has an AI story, something cool they heard from their partner regarding how they're working with AI or how they're getting augmented by AI?

who's got a cool AI story yeah right here oh yeah yeah it's cool come on yeah we're gonna hype you I mean it's like it's surefire way to get applause give this guy a round of applause folks yes yes

yeah so what I learned from my partner today was he's the creator of a unique product design company in Argentina one of a kind and what I learned is that he uses AI for almost everything in his business to the point that he's almost 300, 400 times more efficient than his competitors.

Don't confirm it. Wow. And he's good for scheduling meetings and taking notes, making presentations. He's ahead of the curve in my opinion. And yeah, I think that's really impressive. Wow.

All right. You got a hype man. Nice. Okay.

A Reframe Moment: “I Know Kung Fu” (Delegating to AI Agents)

I'll tell, so we're gonna do this one one more time. So you're about to, just as a warning to everyone, you're about to make awkward eye contact with another stranger. Okay. That's, that's a warning.

I'll tell you a quick story. This is, this is a true story from my life.

As I mentioned, I'm working on a new book. I wrote a book a couple of years ago about creativity and innovation. And one of the things you learn about

writing a book is you got to like read a lot. You have to know a lot of stories. I spent like 12 years learning a bunch of stories to write a book.

You know what they say? I've heard, I heard a musicians say one time, you have a lifetime to write your first album and two years to write your second album.

That's why they call it a sophomore slump. Oftentimes the second album of an artist isn't that good. It's because they don't have any time.

They had like their whole life to write the first record. Right. And, and all the heartbreaks and all, you know, all that

stuff. And then they're on the road for the second record. It's like, I guess I'm going to write about like how hard life on the road is.

Right. Which is really interesting. Anyway, in regards regards to a book, I spent 15 years accumulating knowledge that went into my first book.

And my second book, I go, man, I mean, we're learning at warp speed here, but still, it's not a long time.

I was at home over Thanksgiving, and a friend had recommended a book. It's called Primal Intelligence. It's about how special operators develop imagination and instinct, etc. It's kind of interesting.

In this book, there's this amazing story about Van Gogh, about how he he reinvented color theory and how we used to think it was RGB, but it's actually not. It's CYK and the reason why all this stuff, super cool story.

And I found myself, it's like, this is the first book I've read in a year. I'm like, I have one day where I'm just chilling and I'm able to read a book. And I had this thought, there's no way I have so much time to read this time around. That was my first thought.

And I don't know if you have this feeling ever in your your life where you're like, there's no way I have time to blank. That's actually a really important trigger. There's no way I have time to blank.

And I'm sitting in my living room with this book. And it was like, you know, um, in the matrix where Keanu Reeves is like, I know Kung Fu. I know Kung Fu. I had an, I know Kung Fu moment because I go, I have a team and I got up

and I went into my office and I opened Claude Code and I said I want you to read my book outline I want you to search all of my, I have thousands

of blogs and hundreds of podcast transcripts, I want you to read everything I've already written and tell me where in my outline those things go and then identify gaps

and then commission sub -agents to research on those gaps you have a tireless army of analysts who can go and read every scientific study about any of the things that aren't represented in the book commission sub -agents for every single chapter and then brief me as the senior research lead on why I need to include them in the relevant chapters and I hit enter

and I went away and I went and ate turkey and when I came back it had It worked for over three hours, and I had 90 pages of research per chapter of my book. It was one of those moments. It's like, I know Kung Fu. And now I have 900 pages of research to read, right?

The Overnight-Assignment Mindset

But the point is that what I thought I need more time for, that was the reframe moment. and we've all got to find these reframe moments where you go why would I go to bed without giving a team a research assignment when I started my career at BCG more than 20 years ago we had this

thing called BCG tiger where you could like sketch a slide on graph paper and then send it to you know overseas I don't even know where you'd send it overseas and when you wake up in the morning there would be a beautiful powerpoint slide that was like inspired by your sketch right and I

I realized I never wanted to go home without sending something to BCG Tiger, right? So it's like, I get a new slide in the morning, right? That's the same kind of mindset.

If you have a basic subscription to any language model, you have BCG Tiger available to you, whether you know it or not.

And any night you go to bed without giving an agent an assignment, it's time you're never getting back, right?

But it requires a mental shift. And what I want everybody to do is be inspired to commission an agent to do something on your behalf before you go to bed tonight. That's the goal.

I hope we commission. I hope OpenAI, you know, Sam says like whenever you say thank you and please, it costs us a lot of money. I hope we cost collectively OpenAI hundreds of millions of dollars tonight.

Just this group. Cry me a river, Sam. I hope we cost them a ton of money because we're all like, oh, we're already paying for

this why would i not commission the t the research team that i have employed should i leave them on vacation another night or should i give them work to do right so that's your this is what you're seeking you're like an acai berry like looking for free radicals i don't even know if that's how it works but you get the idea right you're an acai berry you're looking for something you go i'm

going to do that tonight. What's the thing?

Now, turn your head and make awkward eye contact with somebody else in the room. Find someone else. Find someone else.

Don't look at me. Not helpful. I'm not going to talk to you. You got to find somebody else. Okay.

And now again, I want you to spend time finding out why they're the most interesting person in the room and what you can steal to do you tonight. Ready? Go. Go.

All right. Raise your hand if you can hear my voice. Raise your hand. Raise your hand if you can hear my voice. Raise your hand if you can hear my voice. Okay.

Build the Right Thing: Customer Discovery Before Product

Who wants to brag on a partner? Again, you're not bragging on yourself. Brag on your partner. Who heard something cool that they want to share? Yeah, we got one one here. Okay, go ahead, please. And tell us who your partner was.

So, my partner was Sanjay. Sanjay, okay.

He owns a company called Unix. Okay. And he sells a SaaS software which is a decision tree for customer service and stuff.

But the coolest thing I heard from him was not about the company or anything, but I'd simplify what he said. Try to simplify it, because you can correct me. He said, I can be building the coolest of products, but if you don't have a customer, it's no use. So, the hardest thing in any business is getting the customer, not the product itself. So my brain is deep into AI right now, building this cool stuff and stuff, but getting a customer was and is still the most difficult thing I'm still facing, so thank you.

Yeah, well said, well said. There's the can we build it question is different from should we build it. And a lot of, no offense to the engineers in the room, I teach in the School of Engineering, I would say the vast majority of engineers I know love the can we build it question. It's like, oh, this is so hard, right? We've got to do this. And then four months later, they're like, now who's the customer again? And finding ways to quickly, through scrappy experimentation, have conviction about the should we build it. Yes, we should build it. If you can build conviction there, oftentimes that will actually inform specifications of the what can we build.

You know, I'll give you one of my favorite examples is, and by the way, my agent has said, approve the right, click approve on the prompt, and I'll continue, or you can manually save the CSV. I don't even know what that is. Okay, I'll look at it later. That is a great insight.

Case Study: Bridgestone’s $14.60 “Tread Alert” Experiment

One of my favorite examples of this, I have some friends at Bridgestone Tire. They're, you know, a Japanese company, but their headquarters in the U .S. is in South Carolina. Anybody know why? It's where BMW's factory is. So they want to be close, you know. It's vertical integration.

So they've got a lot of operations in the Midwest. They had an idea. This is, you know, probably 10 years ago. They had a really cool idea. They were trying to get out of the commodity tire business. Like, that's like a rubber business, okay? Okay, it's low margins, high competition. They said, we want to be in a differentiated service business.

So they commissioned a bunch of experimentation around new businesses they could start. One thing that became clear, which is actually pretty cool, is Uber drivers, Lyft drivers, really care about maintenance in a different kind of a way.

So the reason is, if you get a flat tire as like a ordinary person, you're kind of indifferent, like it kind of sucks, but whatever, whatever, it's like you gotta repair it. As an Uber driver, a flat tire on a Tuesday is fine because you don't have any rides anyway. On Friday, you're basically killing 90 % of their income. And so they really care about preventative maintenance.

And so Bridgestone had said, hey, what if we created like a sensor that you put it in Uber driver's garage and it would just read the tread wear and let them know when they need to take preventative measures and so they could get in front of needs instead of being blindsided by repairs right so erica wall she's the head of innovation

there said ordinarily what we do we'd like toss that idea over the wall to r &d and r &d we geek out like typical again lovingly engineering style like is it infrared is it lasers is it and they'd be like down to the 14th decimal place of like we're gonna we are gonna measure this tire right and she said i knew that would take six months and approximately six hundred thousand dollars of development cost and she said thankfully we took a low resolution kind of experimentation driven approach we said let's answer the should we build it question before we answer the can we build a question r d is going to deal with can we build it our question is should we build it

and so anybody have any does anybody know this story anybody know what they did i'll tell you one interesting kind of aside about it in bridgestone's headquarters there is a a plaque on the wall. It's a receipt from Target for $14 .60 because it's the cheapest patent that they've ever developed.

So the team went to Target, bought four bath mats. Bath mats. And then they went and brought them to a parking garage

and they said to people as they pulled in, hey, and they put a big sign outside, Bridgestone Tread Alert. You know, and somebody would roll down the window and go, what's that? Oh, well, instantaneously read your treadwear and let you know if you need to take maintenance issues.

If somebody said, how does it work? That's data, right? Somebody said, oh, that's stupid. That's data, right?

So then some people go, how does it work? And they'd say, oh, just give me your mobile number and I'll load you into the system. As soon as you pull into the spot, it'll send you the treadwear, right? Cool. So some people, they give them their number. Some people, they don't.

They pull into the spot and they're going, texting them, warning low tread wear right you know they tried all sorts of things right and they go hey hang on we got a ruler we want to you know the system's a little finicky here we're going to make sure we'll measure it right but what they learned really quickly was what do people do when they learn they've got a dangerous condition do they people started asking well while i shop can you change my tires for me right they didn't think about that right can i get a rental car while you're handling my, right? All sorts of ideas started coming and all it cost them was $14 .60

in bath mats. They didn't need to develop any technology. They needed to simulate a believable environment where somebody pulls onto sensors and gets a message, right? It's the, should we, that's answering the question, should we build this thing? Not, can we build it?

And to your point, customer discovery and validating needs when you can build anything it's really really important that you have high conviction there's a customer there right and so learning and in fact that's actually those are some non -ai skills right i mean they can certainly be augmented by ai of course i know folks in many companies are doing simulated audiences for example right where you can you can survey a panel of 200 users about basically anything right with certain demographics or things like that and it's with you know 80 90 accuracy relative to actual human panels right so a lot of that can even be aided by ai but the point is focusing on finding the right problem before you then kind of geek out on the solution is really important okay do we have any other

More Examples and Inspiration From Partners

bragging anybody else want to brag on a partner something cool that they heard that their partner is using or working with ai for this is a serious problem if no one heard any cool stories. Serious problem. I would recommend you never come back. If you aren't hearing cool stories you should not come back.

Yeah you got one here? I have one from previous.

This gentleman here was talking about an AI application for his. Take this. This gentleman in front of me was talking about an AI application his son created for his girlfriend who wanted to be I guess developing learning learning more about AI tools and things like that, and was finding herself getting really distracted and not really focusing on what needed to be done.

So he wrote an AI tool to essentially track or monitor when she was on the computer, sort of learning these different things.

That's not creepy. No, not at all. But I thought, I was like, hey, whatever works. It seems to be working for her in terms of knowing when her attention is wandering from the task at hand that she really wants to complete.

and I don't know I thought it was really interesting when he was chatting about that earlier so very cool very cool I love it nice anybody else we have time for one more story are we smelling pizza and all of a sudden people just like actually but there's pizza yeah yeah one here please Yannick bring this man a

Marketing in the AI Era: How Do You Earn Attention Now?

microphone I just want to talk about what we discussed in our group here and and we got caught up in Lucy here. She is studying marketing, and we got caught up in what is AI in marketing capable of?

And we talked about the attention that, the way marketing used to work is to get your attention. But in the age of AI, how do you get the attention of people? And I think she has got an interesting exploration in her career as to how a marketing changes in the age of AI so come on yeah

no pressure no pressure I mean I'm still a college student I go to Santa Clara and I have found that he was basically asking like how do you market things for different people and I think it's hard for you to market something for for everybody because we kind of are split up in our different generations.

I feel like for me, Gen Z is like TikTok, Instagram, like all these different social media platforms. That's where people my age normally get their information and news.

And I think for like millennials or people older than me kind of focus more on reading or maybe a little bit of social media. And so I think it's just like, how do you kind of market things and like where do you market things in order to like get your products out there excellent yeah this is a great time to

be in marketing for sure let's give it give a round of applause here nice job nice any others maybe we got time for one more anybody else want to brag on a partner doing something cool yeah right behind you if you want to pass the microphone back yeah right there perfect

I also want to speak about my previous partner, she was a software dev and just six months back she was doing hackathon with a random stranger and then after one day of hackathon they created some solution and they won the hackathon and then both of them got inspired and they started a company. So now she has a running company called Perfect Free

and they create AI agentic solutions to optimize workflows for dermatologists, right? Yeah, clinics and dermatologists and they also got their first customer. So I think that was a really inspiring thing to hear and I think that's what we all are here for.

Yeah, yeah. super cool well done well done well done yeah we had a uh we uh on the podcast there's a great

founder his name is nicholas thorn and he's using he's basically using ai agents to build businesses through a company called autos i'm not i'm not financially involved in this company anyway i am a huge fan of he's one of the smartest people i know but if you slide into their dms on instagram which is how that's basically like the way they connect with with customers uh if you i think it's It's A -U -D -O -S.

But if you go to autos, you can start a conversation with them, and they'll basically build you, like, a pitch deck and a landing page. And they'll start running Facebook ads and Instagram ads, like, everything, basically. It's a pretty interesting model.

For folks who are kind of curious about the whole demand gen side of things and marketing side of things, it's fun to play with just to go. So, like, in an hour, your mind will be blown about how much of a business can be created in, you know, before you go to bed tonight. And so, again, in the spirit of kind of sparking imagination, logging on to stuff like, you know, interacting with and playing with stuff like that, it just, it will blow your mind. Because, again, it ratchets up your imagination and it ratchets up your ambition of what you think is possible.

and so I mean if there are other tools that folks know or other ways that people know to kind of get in and get inspired definitely share them because that's the purpose of this group okay so with that being said we're gonna do is

Wrap-Up: Community, Resources, and Dinner

we're gonna break for dinner we got pizza Yannick you want to say anything about the pizza what kind of what kind do we have just so folks know

use your hand use your hand as a plate if you see anyone with more than two slices just shame them okay just shame tackle them take the piece that looks the best um last thing we want to say here before we break for dinner is

is last gathering, we had Peter Norvig here, you know, OG of AI. Some of you probably were here.

If you didn't see that talk, probably every other meeting or so, we're going to have special guest speakers and different things like that.

We actually had somebody and they got COVID earlier today. So glad they're not here, to be honest with you. Love them. Glad they kept that at home.

But all of the recordings of these sessions, as I said, Milan, Paris, Amsterdam, Boston, Austin, San Francisco, they're all on the MindStone site. So if you want to check out Peter Norvig's talk, again, he's literally an OG AI developer, wrote the defining book on AI. He was here last time, and we have that talk available online.

Call to Action: Commission an Agent Tonight

And if you deploy an agent tonight to do something cool, we want to know it. Okay?

Yannick at MindStone? Yeah.

Check out the video recordings, like Jeremy said. They're on the MindStone event page. If you click on the past tab you can see any of the talks from any meetup, including PeerNorweg.

The other thing I'll say really quickly is we would love to have more speakers, and it doesn't have to be a fancy talk, it doesn't have to be a fancy demo. It's something simple, at least two minutes, five minutes, ten minutes, half an hour if you really want to. But, you know, please come see me if you're interested in that, because it is part of the community to have people from the community also be speakers.

hackers, they don't have to be here in Norway, we love them, but now we want to hear from you guys as well. If, for example, you built a company at your last hackathon and you want to do a demo of it, it's a great chance to do it.

All right, with that, thanks. Thanks, Yannick. Thanks, Hacker Dojo.

Grab pizza. Tackle anybody with more than two pieces and email the stuff that you get agents to do to Yannick at MindStone because we want to be able to hype you next time. Thank you.