7/2/2025 - Chain of Creativity: Crafting Video Games with AI and Passion

Description

Chapters

Transcript

0:01 Oh, champic
0:04 champella.
0:06 [Music]
0:41 Let us be lovers. We'll marry our
0:44 fortunes together.
0:47 [Music]
0:48 I've got some real estate here in my
0:51 bag.
0:52 [Music]
0:54 So, we bought a pack of cigarettes
0:58 in Mrs. Wagner's pile.
1:01 We've all come to America.
1:08 [Music]
1:27 Kathy, I said as we
1:31 [Music]
1:34 see like a dream to
1:38 [Music]
2:12 Yeah.
2:15 Oh,
2:17 [Music]
2:41 thank you.
2:46 [Music]
2:59 Good evening, good people. Welcome to
3:02 the AI learning lab. We're going to
3:04 learn
3:05 something tonight.
3:08 I don't know what, but we're going to
3:11 talk about chains.
3:17 Not tire chains,
3:19 not Allison chains,
3:25 not whatever else you use chains for,
3:29 but chains. We talk about chains
3:32 tonight. Chains are really important. I
3:34 think chains
3:37 are actually the key,
3:39 a key
3:42 [Music]
3:46 to help combat
3:48 some of the AI haters.
3:51 There, I said it.
3:54 [Music]
4:17 Sitting in this lonely town,
4:22 wondering when things are going to
4:23 change.
4:25 [Music]
4:27 Dream my life away.
4:30 Seems these dreams have turned to a
4:32 bunch clouds.
4:36 Get my nerve up, but my past is pulling
4:40 me down.
4:44 Somebody told me once before.
4:48 [Music]
5:10 Somebody told me once before you can
5:13 never go home again.
5:16 Won't you leave?
5:18 Said thanks to steer me away. Yeah. From
5:23 the truth of who I am and what I
5:25 believe. So I thanked him for his two
5:28 cents with a handshake
5:30 and some sympathy. Yeah.
5:33 And I packed up my blue jeans,
5:36 headed for this big prize
5:40 of my freedom.
5:44 Bye-bye,
5:45 black sheep. To the black sheep of the
5:49 family.
5:52 Bye-bye.
5:55 Oh, it means so very much to me. And
5:59 bye-bye
6:00 [Music]
6:02 sheep with my friends and my family.
6:07 Bye-bye.
6:08 [Music]
6:10 Going to strip my soul.
6:14 Set it free.
6:16 [Music]
6:19 Um, chains, chains, chains, chains.
6:26 My god, he's lost his mind.
6:29 He's lost his mind. People, what are we
6:31 going to do?
6:34 He is not in a stable mental state.
6:38 He's lost his mind. He's lost his
6:40 marbles. He's unstable.
6:44 All he's speaking about is chains.
6:46 Nobody knows what he's talking about.
6:48 Chains. Chains. Chains.
6:53 Curious.
6:55 Has anyone discovered
6:59 his qualifications?
7:01 [Music]
7:37 [Music]
7:46 Um, yes, I have lost my mind. and I have
7:49 less lost my ever loving mind.
7:57 But if you have questions about AI, I
7:59 can answer it.
8:03 [Laughter]
8:06 Ah, what's everybody saying here? Have
8:09 you seen Martin Ston live? He is
8:12 amazing.
8:14 He is amazing.
8:18 [Music]
8:32 Um, he is
8:38 I've never heard a voice like his.
8:42 I got to hear Martin Ston sing. I did a
8:44 songwriting workshop about a decade ago
8:48 and he was the guest artist that came in
8:50 and he sang he was standing right in
8:52 front of us. It was a room full of about
8:55 15 of us, 20 of us.
8:57 And they had a microphone but it's like
8:59 a little small coral room, right? Like
9:01 you know three benches,
9:03 you know, the length of the room and
9:05 we're it's a little teeny room and he's
9:06 singing and he stepped in front of the
9:08 microphone.
9:10 Forget what song he was singing. He's
9:14 singing um
9:18 wasn't black sheep. But anyway, he got
9:21 to the chorus and he just hit this note.
9:25 I've never experienced anything like it.
9:27 It was like trumpets were coming out of
9:30 his face.
9:33 Like there were there was
9:37 music coming out of his mouth, but there
9:40 were like harmonies coming out of his
9:42 face.
9:44 It was insane. It was just insane. It
9:48 was just like, you know,
9:52 um he's one of those performers that um
9:56 in a lot of a lot of ways he's got some
9:58 similarities to Jacob Collier where he
10:00 plays the audience, but he also he also
10:03 knows his songs so well
10:06 that on any given night, you know, he
10:10 might turn So Long Susanna into
10:14 um a 12minute riff.
10:20 [Music]
10:23 He'll do that with Purple Rain, too.
10:25 [Music]
10:27 [Applause]
10:28 [Music]
10:39 Yeah. So, you should go see see him. If
10:41 you haven't seen him, you should go see
10:42 him.
10:45 I didn't mean to cause you any sorrow.
10:51 Didn't mean to cause you any pain.
10:54 [Music]
10:57 Only want one time to see you laughing.
11:03 Only want to see you laughing in that
11:05 purple rain. Purple rain. Purple rain.
11:12 Purple rainbow rain.
11:18 Purple rainbow rain.
11:23 Um, yeah, he's amazing. Amazing.
11:25 Amazing. I've seen him, I don't know,
11:26 seven, eight times.
11:30 Your guitar sounds so wonderful. Thank
11:32 you. This is a This is a guild. Wait,
11:35 where's my It's a guild.
11:39 D30, D28. What is it?
11:43 D30 AB. I bought this I bought this
11:47 guitar
11:49 in 1987.
11:52 I graduated Penn State
11:55 and I accepted the job as the jester for
12:00 the Pennsylvania State Renaissance
12:02 Festival.
12:04 And so I just had to walk around and act
12:06 like an idiot. and they paid me $1,100
12:09 in cash.
12:11 And because I'm good with money,
12:14 I spent $400 on a mountain bike and $700
12:19 on a guitar.
12:24 But honest to God, I have never
12:28 [Music]
12:31 Hear
12:34 [Music]
12:38 how deep it is and rich,
12:41 but it's still got
12:43 [Music]
12:53 It's just It's just a beautiful guitar.
12:55 never I can't Everything I play just
12:57 sounds flat and hollow to me because
12:59 this has got such richness to it. Um but
13:01 the high-end is still there. So I feel
13:05 like Taylor's really um great on highend
13:09 but not on bass. I feel like Martins are
13:12 kind of level, not great on bass. This
13:15 one just there's something about how it
13:16 resonates. It's just amazing.
13:20 And man, I've beat the [ __ ] out of this
13:22 thing. Like the frets are the frets are
13:24 all need to be redone. The neck's too
13:27 low. Um
13:30 it's got all sorts of scratches and [ __ ]
13:32 from my from when I walk into screen
13:34 doors and [ __ ] like that.
13:39 But it's really good.
13:42 Yeah. Thank you. Thank you. It's nice to
13:43 have someone. It's magic. Yeah, it
13:46 really is.
13:48 [Music]
14:07 [Music]
14:39 Okay, let's see. What are we going to
14:41 do? What are we going to start? Where
14:42 are we going to start? Where? Oh, a yes
14:44 riff, right? Yeah, exactly.
14:47 [Music]
15:13 Um,
15:15 things we learn in college.
15:18 I'll take I'll take songs we learned 10
15:21 seconds of in college. Bob,
15:28 [Music]
15:31 wait.
15:34 [Music]
15:45 [Music]
15:52 Anyway, I used to know that Blackbird
15:57 agent since since the music sound nice,
16:00 let's try some musical ideas. We can
16:02 play some music tonight.
16:06 Um, got my Oh, Serena stickers today.
16:09 Yeah, if if you knew Serena, so we have
16:13 these really amazing stickers
16:17 in honor of Serena who passed away about
16:19 three weeks ago. Now, I don't know if
16:21 you knew knew her or not, if you knew of
16:24 her, she was absolute she came to every
16:27 single one of these AI learning lab
16:30 lives. She's the one that coined the
16:32 term the irregulars with this phrase,
16:34 we're not weird, we're irregular. So, if
16:38 you want to order these uh stickers, you
16:40 get this one and you get an AI salon
16:42 sticker and you get a nice little
16:45 memorial sticker that Tobias has kindly
16:48 generated. And the the sticker set for
16:51 those three stickers stickers is 20
16:53 bucks, but all 20 bucks goes to her
16:56 family. Um so, if you go to uh the AI
17:00 salon
17:02 um aison.mn.co
17:05 co and scroll down into clubs and hubs
17:08 and you go to the irregulars channel
17:11 here. Let me share my screen. Am I not
17:13 sharing? I am.
17:17 Uh why did the There you go. And go to
17:20 the irregulars channel. And then we've
17:21 got remembering Serena here.
17:25 And
17:27 there's the stickers. And then you can
17:29 order them. So you just click click to
17:31 order them from here from Tobias's shop.
17:34 And any any sets that you order there,
17:36 the proceeds go straight to her family.
17:40 So yeah. So that's cool. They came,
17:42 Kelly. That's awesome.
17:44 Um,
17:46 she was in one of my lives. Nice, Mr.
17:48 It. That's awesome.
17:52 [Music]
17:55 If I add the guitar to the stickers,
17:58 you'll buy them. No, I'm not selling
18:00 this. This guitar. This guitar. Too
18:02 nice. too too. Uh,
18:07 [Music]
18:13 every time I see it now, wait. Every
18:17 time I see you now, get that look in
18:19 mine.
18:22 Every time I see your mouth, I hear that
18:26 smile.
18:29 The early misty morning light. I heard
18:32 the engine burning
18:35 the old outside.
18:39 Um, okay. Here's what we're going to
18:42 talk about. We're not going to do any
18:44 show and tell for a minute. I'm going to
18:45 talk about chains. You're be like,
18:47 "Chains? What the hell are you talking
18:48 about? Chains? What are you talking
18:50 about? Cow. Cow. Hey, cow. Cow. Cow.
18:52 Cow. Cow. Cow. Cow. What are you talking
18:55 about?" Um, anyway, if you're new here,
18:57 how many folks here? Oh, it's it's a
18:59 light crew tonight. It's probably all
19:00 people we know. Amelio's wife. Yeah,
19:02 exactly. I know, right? Um
19:06 really, really sad. Really sad that we
19:08 lost her. Um she was she was a I would
19:11 argue that the
19:15 the culture, the tenor of of this group
19:21 that that hangs out here every night um
19:24 is as generous as it is um because
19:27 because of how she showed up here. She
19:29 was just always positive, always always
19:31 full of love. Um, okay. So, I've been on
19:37 a I've been on a bit of a journey and I
19:39 I took a little bit of this journey the
19:41 other night where and this going to seem
19:44 weird. We're going to meander a bit.
19:47 It's welcome to chat add. That's how we
19:48 do it here. I might even finish a
19:50 sentence tonight.
19:54 Um, I've been trying to figure out
19:59 when someone says what tool should I use
20:01 for X
20:05 or if we want to put together a list of
20:07 here's a list of recommended tools how
20:10 do you organize them
20:14 and the challenge is is that the tools
20:18 are getting very sophisticated a single
20:21 tool probably has
20:24 anywhere between five and 50 features
20:28 Some of which are very specific and very
20:30 useful for one very specific thing. Some
20:33 of which you don't know if you're ever
20:35 going to use that little piece of some
20:36 of which are like okay but you like the
20:38 other tool better. So how do you talk
20:41 about that? And Silver Fox had a good
20:44 idea of putting up u
20:47 you know doing keywords and making it
20:49 you know sort of dynamic.
20:51 So like you could find a video tool also
20:54 while you were looking for image tools
20:55 because some of the video tools now do
20:58 images as good as the image tools do
21:00 images.
21:03 And so three nights ago, four nights
21:05 ago, I don't know, we did um a Gemini
21:08 deep research thing on this dilemma and
21:11 we said go off and research
21:13 how we should deal with talking about
21:16 these tools in this
21:19 in this uh
21:21 complicated environment. The example I
21:24 gave was
21:26 um if you've seen the Bigfoot videos or
21:29 the Yeti videos on Twitter or LinkedIn
21:32 where it's like the Bigfoot walking
21:34 through the woods selling you some, you
21:35 know, mosquito repellent or whatever the
21:37 hell they are and someone says, "Hey,
21:39 that I I want to make one of those Yeti
21:42 videos. What tool do I use?"
21:46 And historically, what I would do is I
21:48 would answer that.
21:52 I've been to a Renf fair. Does that
21:53 count? It does. I have worked at many a
21:57 Renf fair. It does not surprise me that
22:00 we have Renaissance fair geeks in this
22:02 channel.
22:08 Oh, nice. Kelly is Kelly is hawking the
22:11 uh the LinkedIn channel, the LinkedIn
22:12 live over there. Yeah, if you're
22:14 watching this on LinkedIn, um this is
22:16 not your typical LinkedIn fair. uh we
22:20 are talking about thing. If you hang out
22:23 here for many nights in a row, you will
22:25 probably learn more here than than you
22:28 will learn uh taking some fancy ass
22:31 course. However, um it's like osmosis.
22:35 There's not a curriculum. It will just
22:37 come in over time. Um,
22:41 so the the Gemini deep research on this
22:45 tool thing came back with this concept
22:47 called
22:49 jobs. What was it? Jobs.
22:53 [ __ ] I can't remember it.
22:56 See if I've got it here.
23:02 We close that.
23:27 Jobs to be done.
23:31 the jobs to be done framework,
23:36 which that's not a good name,
23:40 but what it basically talked about is if
23:43 if someone says, "I want to make one of
23:45 those yet Yeti videos."
23:47 Historically, what we do is we say,
23:49 "Well, I use this video tool or I use
23:51 that that
23:53 large language model, right?"
23:57 But in almost all cases, that's not
24:00 actually what happened. What what
24:02 happened was there was a sequence of
24:04 events. There was a chain of events
24:08 that happened that that you stack
24:10 together all of these different in this
24:12 case what what they're calling jobs to
24:14 be done. You're like, "Okay, we want to
24:16 make a Yeti video." Okay, well, do we
24:20 know what a Yeti is? Do we know what a
24:22 Yeti looks like? We could go Google
24:23 that. We go to chat GPT.
24:27 Do we know what we're trying to
24:28 accomplish? Do we want these videos to
24:30 be funny? Do we want them to be scary?
24:32 Do we want them to be photorealistic?
24:36 What are the If we're going to have the
24:37 yetis talk, what are we going to have
24:39 them say? Are we saying it for a
24:41 business? Do we want it to have some
24:43 coherent message? Is there a target
24:45 audience? Right? So, all of these
24:48 unknowns have to get answered.
24:53 What is the what does the voice of the
24:54 Yeti sound like? Does that matter? Do we
24:56 want to design that? What is the look
24:58 and feel of the Yeti? Is that we're
25:00 going to leave that up to us or up to
25:02 the model? I don't know.
25:06 And so chains
25:09 are chains of activities
25:12 that when someone says, "Hey, what tool
25:14 did you use to do that thing?"
25:17 I think there's something really
25:19 powerful in saying
25:23 in not answering them
25:30 in not saying which tool you used
25:33 because maybe and I I do this here all
25:36 the time, right? I'll go, "Oh, that's a
25:39 cool tool or that's a cool output. Let
25:42 me go look at how we can make that." and
25:46 we'll go to chat GPT and and I'll have
25:48 sort of an idea and I'll kind of
25:49 brainstorm with chat GPT. Then we'll
25:52 come up with a concept. Then I'll have
25:54 it write a creative brief and then it'll
25:57 write a shitty creative brief and then I
25:58 yell at it
26:01 and then we'll go back and forth and
26:02 then it'll write some scripts. It'll
26:04 sort of get the scripts right but sort
26:05 of not. Then we'll turn the scripts into
26:07 prompts and then we'll take the prompts
26:09 over to VO3 and it'll do a shitty job
26:11 and then we'll be like, "Well, wait a
26:13 minute. What if we took that script and
26:15 we did the voice in 11 Labs and then we
26:16 took the thing out of 11 Labs and we put
26:19 that into Hedra? Maybe that's another
26:20 way we could do this.
26:24 That chain
26:28 is way more important than whatever tool
26:31 you ended up using along the way because
26:34 that chain is where the human
26:36 contribution is happening.
26:40 The trope that we hear from a lot of
26:42 people talking about
26:44 AI slop
26:49 is that
26:51 especially people who don't use AI
26:54 assume
26:57 that how you make something with AI is
26:59 you go make something with AI.
27:03 Push the button here. I'm going to push
27:06 the button. You can make money with
27:08 judging. I did it. Yay.
27:17 And so I think when someone asks you,
27:18 "What tool did you use?" You say, "Well,
27:21 the tool I started with was my brain."
27:27 Huh? Yeah. I saw this cool thing and I
27:30 thought, "How could I do that?" And then
27:33 I didn't want to just copy it. So, I
27:36 actually went to my whiteboard and I I
27:37 drew up a couple of So, the second tool
27:39 that I used was a whiteboard. I think it
27:42 was invented in 196. When was the
27:44 whiteboard invented? Let's take a chat
27:47 moment, shall we?
27:49 When was the white board? I'm going to
27:53 guess 1963.
27:56 No, no, no. 1954.
28:00 When was the white board invented?
28:03 Because they're going to say
28:04 chalkboards. It's not a chalkboard. The
28:06 modern whiteboard was invented in the
28:09 50s. Bang.
28:14 1950s. The concept is credited to Martin
28:17 Height, a photographer and Korean war
28:20 veteran who reportedly used marker on
28:22 film negative to take notes and wipe
28:24 them clean, leading to the idea of a
28:27 reusable surface.
28:31 So, I started out using my brain and
28:33 then I went to the whiteboard and then I
28:34 went to chat GPT and I took these
28:37 concepts and I had and I and I help had
28:39 it help me flesh them out.
28:42 And then I thought, well, rather than
28:44 just having it write ads right away,
28:46 maybe I should have it write a creative
28:48 brief because I know if you give it more
28:50 context, it gets better. So then I
28:53 talked to chat GPT for a while about a
28:55 creative brief and at this point they
28:57 will have w the people that ask you the
28:59 question will have walked away right
29:02 which is good like you don't want to
29:03 really talk to them anyway
29:06 because all they want to do is [ __ ] all
29:08 over you. Oh, what'd you do? Push a
29:10 button and make a video you little baby?
29:13 Yeah, I did.
29:17 And then and then after I did the
29:19 creative brief, I had it come up with
29:21 some some campaign concepts. And then I
29:24 had it write some scripts. And then I
29:25 had it describe the Yeti.
29:28 And then I thought, wouldn't it be cool
29:29 to design a voice? And I went over to 11
29:31 Labs and we were designing a voice. And
29:33 then it hit me, wait a minute, if we
29:34 make the scripts 8 seconds long, we can
29:37 use V3 that just does the acting and the
29:39 voice. It's it's just all built in. Why
29:41 don't we just try that?
29:43 And then we went to VO3 and we put it in
29:45 there. we couldn't get it to look quite
29:47 right, but it sounded great. And so now
29:50 we're going to work on that.
29:53 Or we went to V3 and we realized, oh
29:55 [ __ ] we actually to do this right, we
29:56 need a 15-second video and we can't
29:58 extend V3 videos with voice. So maybe
30:02 we'll use Hedra. So then we've got to go
30:04 to a right.
30:07 That chain of events is craft.
30:11 When people talk about AI is lazy, we're
30:15 losing the craft. When you hear
30:17 filmmakers say there's no craft in AI
30:20 video, horseshit.
30:24 the chain, the sequence of events that
30:27 Joy Pertie goes through when she has an
30:30 idea for a film
30:32 and then she writes a script and then
30:34 she goes into one of four or five
30:37 different image tools and one of four or
30:39 five different video tools and then she
30:42 knows which of those video tools can do
30:44 sound effects. And if they can't, she
30:46 knows she can go to 11 Labs to do sound
30:48 effects. And then she can go get
30:51 synthesized voices and then use Runway
30:55 ML's act one to take a moving video and
30:58 have that moving video now say something
31:01 with synced lips.
31:03 And then she does that and then she
31:04 takes it into her nonlinear video
31:07 editor, which she took a year to [ __ ]
31:09 figure out and learn from the ground up.
31:12 And she can stitch all these elements
31:14 together. That is [ __ ] craft.
31:18 It is no different
31:21 than today's film making craft except
31:24 that it's completely different.
31:29 But it's a human being who has a story
31:31 to tell going through a process having
31:34 made some mastery of some amount of
31:37 tools along the way.
31:40 So when people say to you, "What tool
31:42 did you use to make that video?" What
31:45 they're effectively doing is diminishing
31:48 your contribution into this single push
31:50 of a button
31:54 so that they can stay on the sidelines
31:56 and judge you as less than.
32:02 of chains are really important.
32:07 And I think it also gives us an
32:09 opportunity to say, "Hey, wait a minute.
32:13 this song that was an AI generated song
32:16 I actually wrote the lyrics for and I
32:18 and I hummed the beginning of the song
32:21 into suno and then it turned it into
32:22 something and it took my lyrics and then
32:25 I made
32:27 let me count 157
32:30 songs before I got one that sounded like
32:33 the thing I have in my head.
32:37 Wait, you made 157 songs that didn't
32:40 make it out of the edit booth? Yeah.
32:45 Huh? I thought it was more simple than
32:47 that. Yeah, cuz you don't do it. You're
32:50 on the sidelines
32:52 [ __ ] all over people who are trying
32:54 to figure out this new craft.
32:59 And when you really don't know what
33:01 you're doing, like me, most nights that
33:03 chain of events is some random
33:06 meandering through the forest.
33:09 But then what happens is you figure out,
33:13 oh, this is repeatable. If I do this, I
33:17 can do this. I can do this, and I get
33:18 some predictable result.
33:21 Well, now that's something you could
33:22 actually automate or turn into an app or
33:26 document so that other people can do it.
33:32 So, I don't know thoughts
33:36 they say such silly thing because
33:37 they're scared. I know it's not terribly
33:40 different for contemplating someone's
33:42 talent versus a hard earned skill. Yeah.
33:45 Someone says, "Oh, you're talented." No,
33:47 no, I worked my ass off. I've taken two
33:49 and a half years to figure out how to
33:51 stitch together these [ __ ] janky ass
33:54 tools that some work half the time maybe
33:57 if you're lucky and good.
34:05 This is not a magic button. It is not.
34:09 Wait, who's having Korean barbecue?
34:13 Joy Birdie.
34:16 Okay, Korean barbecue is one of my
34:19 favorite foods when I was doing
34:20 agency.com. So, my um my co-founder of
34:24 agency.com is a Korean dude named Chan
34:26 Sue. And so whenever we would like win
34:30 new business or I don't know wanted to
34:32 celebrate, we would go down to
34:36 32nd Street in Manhattan between Fifth
34:39 and 6th, Little Korea and oh my god, we
34:43 would we would eat at New Wanjo or if we
34:46 wanted to get fancy, we would eat at
34:48 Wule Oak on I think it was on West 47th.
34:53 Was it 47th? Yeah, I think so. It burned
34:56 down.
34:58 It was amazing, man. [ __ ] Korean
35:00 barbecue is the best.
35:03 Anyway, I don't know if you heard me,
35:05 Joy. I was I was I was uh
35:08 invoking your name here
35:11 about realizing that that
35:16 we should stop talking about tools. I
35:18 use this tool to make this thing.
35:23 We should talk about the the chain of
35:25 events, the chain of tasks,
35:28 the combination of
35:30 human ideiation,
35:33 curation, taste,
35:36 persistence, creativity,
35:42 mastery
35:45 of these tools that are changing on a
35:47 weekly basis.
35:51 That's That's it.
35:57 See, being told you're so smarter. I
35:59 could never do that. Yes, you can, but
36:01 it's not easy. Yeah.
36:04 Yeah. It's you you know Pate it's you
36:07 know I think I think about engineering
36:09 and and it's like
36:12 you know engineering and storytelling
36:14 and film making
36:17 are all just series of tasks and and
36:20 there might be there might be things in
36:21 there that that require specialty
36:25 knowledge but especially with today's
36:28 tools you can figure out all the
36:30 individual steps but it's having the
36:32 passion for the process
36:34 is what makes a good engineer, what
36:36 makes a good storyteller, what makes a
36:38 good filmmaker, what makes a good
36:39 artist, what makes a good educator
36:46 is that chain of events.
36:50 And I think because we we've gotten so
36:52 seduced, myself included, we've gotten
36:54 so seduced by the tools that we want to
36:57 say Chad GPT did this. No,
37:03 you did that
37:07 with chat GPT.
37:11 But what was the actual chain of events?
37:19 How many iteration how many how many
37:22 times did chat GPT give you a result
37:24 that was so bad that you're like, I've
37:27 got to try a different tool. Then you
37:30 went and tried a different tool and it
37:31 couldn't do it either. But it gave you
37:33 an idea for how to go back to chat GBT.
37:36 Oh yeah, maybe I should use a different
37:37 model. Yeah, let me try a different
37:39 model. That's not quite right, but it's
37:40 close, right? That's all day every day
37:44 with AI.
37:46 That process is craft
37:49 [ __ ] craft. Anyway,
37:52 um I'm feeling good about this. I've
37:54 used both creative thinking and and this
37:58 is um Chef Kelly. I've used both
38:00 creative thinking and analytical
38:02 thinking more in the last 18 months than
38:04 I had in 15 years. Yep. Me too. Me too.
38:10 I talk about this all the time that my
38:12 brain has changed
38:14 with AI.
38:16 And one of the things that I talk about
38:20 is um the positive reinforcement you get
38:23 from large language models that are
38:25 trained to be nice to you that when you
38:27 put an idea in there, it's like, "Hey,
38:29 great idea. Oh, thanks." You know, don't
38:31 mind if I do.
38:34 They did call me clever in college.
38:39 So tick tock pit.
38:43 It's like saying you didn't build that
38:45 house. The hammer did. Yeah. Exactly.
38:48 Exactly.
38:50 Yeah.
38:52 How'd you build that house? What tool
38:54 did you use to build that house? Oh, you
38:55 used a hammer. Yeah. Yeah. I mean,
38:58 hammers are simple. You just It's you
39:00 just do this. I've I've hammered things
39:02 before.
39:04 Yeah. I made a house with hammers
39:06 before. Yeah. I I mean it was a
39:08 doghouse, but you know, but I used a
39:10 hammer. Do you use a saw, too? Yeah.
39:13 Yeah. Yeah. I mean, saws are good.
39:17 You the Did you do the H or the H? Oh,
39:21 you Oh, nice. I didn't realize you were
39:24 fancy.
39:27 But that's a nice house. No, you
39:29 wouldn't talk about how'd you build the
39:31 house. Well, I've always wanted to build
39:33 a house.
39:35 So, I went out and I learned about wood
39:37 and I learned about how you pour
39:39 concrete and I learned how you dig holes
39:41 and I learned why you did all that. Oh,
39:46 and then what'd you do? I figured I
39:48 didn't Who knew how you got concrete? I
39:51 didn't I saw cement trucks. I didn't
39:52 know where you found them.
39:55 So, I went and I found cement truck guys
39:57 on in the yellow page. One of the yellow
39:59 pages. It's a thing old people use. Oh,
40:03 then what'd you do next? I found a guy
40:05 named Bucky who'll bring you a a truck
40:09 full of cement and he'll pour it in the
40:11 hole you put in the ground. Oh,
40:15 right.
40:22 Joy Pertie,
40:24 30 years as a a sleep analyst,
40:29 has stories to tell.
40:32 doesn't have a degree in cinematography,
40:35 doesn't own a film camera, doesn't know
40:37 editing,
40:40 starts playing around with image
40:42 generating tools. She's like, "Huh, I
40:45 can get these ideas I have in my head
40:48 out in a way I never could." Huh.
40:53 I wonder if you can make them move. Oh,
40:55 here's a tool that makes images move.
40:59 Huh.
41:01 Wonder if I could get him to talk.
41:03 Right. So, so Joy over the course of
41:06 six, eight, 10 months, a year, two years
41:12 is figuring out how to build her house,
41:14 figuring out how to tell her stories
41:16 that have been in her head for most of
41:19 her life.
41:24 And then she puts a film out and someone
41:27 shits all over it and says, "Oh, it's AI
41:29 video. That's worthless, you thief.
41:40 You just pushed a button." No, she
41:43 didn't.
41:45 She had to live her life
41:49 [Music]
41:51 to be able to inform the stories.
41:53 [Music]
41:55 that motivated her enough
41:58 to dig through these shitty, janky,
42:01 [ __ ] up tools
42:06 in a in a series of unlikely events
42:11 that she could just do at home at the
42:13 dining room table
42:16 instead of having to go
42:18 raise half a million dollars.
42:21 and she could tell her stories
42:25 and is telling her stories.
42:29 [Music]
42:35 So chains
42:39 chains are important.
42:44 There ain't nothing easy about
42:45 blacksmithing. Yeah, exactly.
42:49 My whole life I could never paint, draw,
42:52 make a good photograph, sculpt, design.
42:54 And now she can
42:59 anything. Anything. This is this is my
43:02 absolute conviction.
43:07 I think this is the reason I'm most
43:08 excited about AI.
43:12 All right.
43:13 Who How many people we got here? Got a
43:16 handful of people. 36 people.
43:20 I want you to think about in your life.
43:25 Maybe even close your eyes. We'll do one
43:27 of those close your eyes exercises. I
43:29 want you to close your eyes
43:33 and I want you to think about something
43:36 that you know that you know
43:41 you're no good at.
43:46 And I want you to think back and see if
43:48 you can go back in time to the moment
43:53 where it became clear to you that you
43:57 couldn't do that.
44:02 Was it that you shared a drawing to the
44:06 girl you had a crush on and she laughed
44:08 at it?
44:10 Was it your mom telling you you weren't
44:12 talented? Was it a teacher telling you
44:15 that
44:17 you should really go into accounting?
44:20 Think back to that moment
44:26 where
44:29 you put yourself
44:32 into a kind of purgatory
44:35 with that particular thing. Because if
44:38 there's a thing that you know you're not
44:40 good at and it's kind of painful that
44:41 you're not good at it,
44:44 that means that's a thing that you've
44:46 wanted to do for a long long time and
44:50 you know
44:52 deep in your soul you can't do it.
45:00 Guess what?
45:04 AI lets you do it.
45:10 You can draw, you can make music, you
45:12 can do science, you can do math, you can
45:13 do programming,
45:15 you can do business strategy, you can do
45:17 marketing.
45:20 You can write scripts for Tik Tok, a Tik
45:22 Tok channel. You can do animation.
45:25 Anything. Anything. Anything.
45:32 Yeah, but I'm not good at No, no, no.
45:34 Anything.
45:36 And it's going to get better and better
45:39 and better and better and better and
45:41 better. And so a thing we get to do now,
45:46 which is really the first time in
45:48 history that this has been possible, we
45:50 get to
45:53 say, "Okay, life world,
45:58 I know that I'm not good at this one
46:01 thing, but if I were,
46:04 if I could do that thing, if I could
46:06 draw, if I were better at songwriting,
46:10 if I were better at coming up with a
46:12 business plan.
46:15 You know the thing I would do
46:19 that's our job.
46:22 Tik Tok pin
46:25 that would have cost me tens of
46:27 thousands of dollars in probably 18
46:29 months. I did it in 10 weeks. I don't
46:33 know what you did but yes
46:36 that kind of thing is possible.
46:40 Joy Perie when she was asked I I might
46:42 have asked her I because when I saw the
46:44 first the first film I saw from Joy
46:46 Pertie was the one she did for the
46:48 salon. I think I might have seen one
46:50 before that but whatever.
46:52 But it was so good and it had such good
46:58 film craft in it. I'm like clearly
47:02 you've got a film background.
47:05 No,
47:06 but I've always had stories I've wanted
47:08 to tell, just never been able to tell
47:10 them.
47:13 Holy [ __ ]
47:16 So, I would encourage you to think about
47:18 what's that thing for you?
47:21 What's that thing that you've denied
47:23 yourself your whole [ __ ] life?
47:26 Because, you know,
47:30 it pains you.
47:33 you're [ __ ] crushed
47:36 because you're not good at that thing.
47:40 And then like literally just
47:43 ask yourself the hypothetical, what if
47:46 that weren't true?
47:49 What would I do?
47:52 And and what I would do if I were you,
47:56 if you have an answer to that, I would
47:58 start it tonight.
48:01 Well, but how do you start it? I don't
48:02 know. I don't know what your thing is.
48:04 Is it I've always wanted to make a video
48:07 game.
48:09 You want to see something fun?
48:12 [Music]
48:15 Motorcycle.
48:17 Put this down.
48:19 Um,
48:23 my fave.
48:27 Can you see what I'm doing here? Let me
48:29 make this bigger.
48:35 My fave
48:39 video games are
48:42 old school
48:44 trigger twitchers.
48:47 I hate sim games.
48:52 I loved
48:55 asteroids
48:57 and
48:59 centipede.
49:02 and
49:04 Missile Command,
49:08 you know, the good games.
49:12 I want you to come up
49:17 with five new
49:23 video game concepts
49:30 that
49:31 pay homage to the 80s arcade games,
49:39 but will be relevant for a current
49:45 gaming audience
49:52 and have
49:55 nostalgia
49:57 points for the Gen Xers.
50:04 All right.
50:07 You've always wanted to make a video
50:09 game your whole life, but you're not
50:10 good at programming. You weren't good at
50:13 math. You applied for Stanford and they
50:16 told you that your scores were too low.
50:18 You were an idiot.
50:24 And so you do your spreadsheets or you
50:26 do your marketing or you do the thing
50:29 that pays the bills.
50:31 And then moments of weakness, three
50:33 whisies in, you're like, I had an idea
50:36 for a video game once. It could have
50:38 been so big.
50:41 But then daddy told me he wouldn't get
50:43 me a Timex Sinclair and I couldn't learn
50:46 basic programming. Give me another
50:49 drink.
50:54 Debrisfield, a spiritual successor to
50:56 Asteroids, where you're a salvage pilot
50:58 in a cyberpunk future. Instead of just
51:01 blasting blasting rocks, you're
51:03 harvesting valuable tech from floating
51:05 space junk. Neon Swarm. Think Centipede
51:08 meets Tron. You're defending a city.
51:10 That That one sounds cool. Sky Shield
51:13 Protocol. Fractal Fighter. Vector Storm.
51:16 You know what? Neon Swarm it is.
51:20 Um, my daddy told me I'd never be a
51:29 programmer and wouldn't pay for me to go
51:34 to Stamford
51:38 and ruined
51:41 my life
51:44 and now I sell carpets.
51:49 But all I wanted was to make video
51:55 games.
52:00 Make me
52:02 neon swarm.
52:06 It sounds awesome.
52:11 Awesome.
52:14 Awesome. And the first person I am going
52:19 to have play it
52:24 is
52:28 daddy. Play it is daddy. I don't know
52:31 why this took such a dark turn, but this
52:35 is life. Like life is [ __ ] up like
52:37 this, right? We've all got that dark
52:39 place. That dark the dream died sometime
52:43 probably between when you were five and
52:45 seven.
52:46 That's the time in life where you
52:48 realize that life is dangerous.
52:52 It's when you start shutting down that
52:53 part of you, the joyous part of you that
52:56 is only possibility.
52:58 You start shutting it down then. And
53:00 most of us go through our whole life
53:04 living that reality.
53:07 The choice that a five-year-old make
53:09 made is what's driving our life today.
53:13 So no more. So daddy told you wouldn't
53:16 send you to Stamford even though he had
53:17 the money [ __ ]
53:20 Watch this.
53:25 I'm suing my stepdad with the help of
53:27 AI. Perfect. There you go.
53:30 [Laughter]
53:32 I'm going to make you What did it say?
53:35 Um, I'm going to make you an amazing
53:37 version of Neon Storm. Swarm that'll
53:40 show your daddy exactly what you're
53:42 capable of. And look at this. You didn't
53:45 You said you couldn't code, right? You
53:47 weren't smart enough to go to Stanford
53:48 or MIT. Well, you're coding now,
54:00 right?
54:05 Let's create something that captures
54:07 that classic arcade magic with a modern
54:09 cyberpunk twist.
54:12 So, I don't know what it's doing. I
54:14 don't care.
54:18 It's doing more than I would have ever
54:21 done.
54:24 Now,
54:26 if you're someone like Pate, who is an
54:28 actual
54:34 smarty pants with a PhD
54:37 in I think computer science,
54:39 mathematics, something I don't know.
54:45 You do care what this code is and you
54:48 want this code to be good and a lot of
54:50 times it's not.
54:52 But in his chain of events, he can
54:54 figure some stuff out. Okay, how does
54:56 this work? What are we supposed to do?
55:01 Hack firewalls to change them. Arrow
55:04 keys move. Spacebar fires.
55:07 Okay, there you go. Complete fully
55:09 playable version of Neon Swarm. This
55:11 isn't just a demo. It's a real game with
55:12 professional polish.
55:15 All right. Arrow keys to move. Spacebar
55:17 to shoot. Shift for overclock mode. R to
55:21 restart. All right. So, how do I start?
55:27 Oh, I just fired something. Wait.
55:32 Oh, okay. I got it.
55:37 Look,
55:42 that's pretty good. Doesn't have any
55:44 sound effects, but
55:50 it's actually pretty good. It's like
55:52 Space Invaders. This is like Space
55:54 Invaders meets the like the speed of
55:57 asteroids or centipede
56:00 with the sort of mechanics of Space
56:02 Invaders. This is pretty cool. This
56:04 kicks the [ __ ] out of Space Invaders,
56:06 quite frankly.
56:12 The little the little uh centipede
56:14 thingies are uh a little too slow,
56:18 but damn
56:20 damn that traffic jam. So, guess what,
56:22 Daddy? You want to play the video game I
56:25 just [ __ ] built? Son, you're 54 years
56:28 old. Why are you making video games and
56:30 wasting everybody's time? Cuz, dad, it's
56:33 all I ever wanted to do in my life. I
56:35 work at your stupid [ __ ] carpet
56:37 store. But not anymore. I'm a game
56:39 developer now. Look at my game, Dad.
56:44 By the way, I don't code.
56:49 I have a degree in acting. It was only
56:51 acting. There's a shocker. Oh, he's an
56:54 actor. That explains more more than I
56:57 wanted to know about this channel.
57:04 How good is this game? It's I mean, it's
57:06 a little easy, but like I love how
57:08 flipping fast it is.
57:13 All right, let's see. Um,
57:17 okay, this is good.
57:20 But nothing is shooting back. And the
57:25 red
57:29 centipede
57:31 things
57:34 are really slow
57:38 and easy to kill.
57:42 And
57:44 there is no audio.
57:49 Come on, man. The sounds
57:54 are everything in a game like this,
57:59 right?
58:03 Right peoples.
58:08 Space Invaders visits asteroids.
58:16 Dang, look at that long code. Like chat
58:18 GPT has daddy issues. I'm actually on
58:20 Claude. This is Claude.
58:26 Season two of Daddy's Last Game.
58:33 Oh man. Anyway, questions, thoughts.
58:41 So the reason to start thinking about
58:43 chains, chains of activities
58:47 is that
58:50 if you've got those things in your life
58:52 where you're like, I can't do that one
58:54 thing, you'll think about a chain of
58:56 events and then you'll get to that
58:57 blocker. Oh, I've got to do programming.
58:59 Yeah, I'm not a programmer. I can't
59:01 program. I didn't go to Stanford.
59:04 And you won't pursue it, right? And
59:06 historically in life, I think whoever
59:08 said it here, historically in life,
59:11 you've got you got to put in 10,000
59:13 hours to be able to acquire a skill to
59:16 get over that hurdle. And if you've got
59:18 that mental block like I cannot do that
59:21 thing, I will never be able to do that
59:24 thing. 10,000 hours isn't enough because
59:27 you need 10,000 hours of therapy on top
59:30 of 10,000 hours of
59:33 skills training.
59:36 But now
59:38 what if you could just do [ __ ]
59:42 Oh, well then I've got a whole lifetime
59:46 of ideas of things. I said, "Boy, if I
59:49 would have gone to Stanford, here's what
59:51 would have learn what I would have done
59:54 with my life." Right? We all have a
59:56 brain full of those ideas.
1:00:02 This thing is now refactoring our game,
1:00:04 redesigning it. Is this Are we good? Are
1:00:06 we here? Come on. Are we done? Yes.
1:00:10 System breached. Press R to restart.
1:00:12 Okay. R. Wait. Click. R. R.
1:00:17 Wait. Daddy's gonna be Daddy's gonna say
1:00:22 he's right. I was a jerk. Press any key
1:00:25 to enable audio.
1:00:27 Uh, system breached. Uh,
1:00:31 game isn't
1:00:33 restarting after
1:00:37 system breach.
1:00:40 message.
1:00:44 Yeah. Where I am right now is I'm in
1:00:45 claude.ai.
1:00:47 And this thing over here is called an
1:00:49 artifact. So you you do your chat over
1:00:51 here in this window.
1:00:54 And then over here you got your code.
1:00:57 Uh oh. Wait. What's going on? Okay.
1:01:01 And it's it's fixing the code. It's
1:01:03 doing the code [ __ ] Uh uh. Hang on.
1:01:25 Yeah, this is better. It's actually kind
1:01:27 of enjoyable.
1:01:32 Okay, except
1:01:35 first of all, daddy can go [ __ ] himself.
1:01:38 I'm a good programmer. I'm a good game
1:01:40 designer. Daddy, look at this.
1:01:44 Well, that's not bad, son. I always
1:01:46 thought you could do that if you just
1:01:47 put your mind to it. You wouldn't send
1:01:49 me to Stanford.
1:01:53 Oh, sorry. I was a little drunk that
1:01:55 day. What? Yeah. If you'd asked me the
1:01:58 next weekend, you probably could have
1:02:00 gone, "What? Really? Do you love me?"
1:02:06 That's how life is, you know?
1:02:09 I remember I did one of those with my
1:02:11 father with my stepfather where I went
1:02:15 and I did the the forum which you know
1:02:17 landmark forum and one of the things I
1:02:21 realized is that
1:02:24 my stepdad
1:02:26 would make me go to my room before he
1:02:28 got home.
1:02:30 And so every night I would hear like him
1:02:32 and my mom watching TV and having fun
1:02:35 and I would have to kind of sneak out
1:02:36 and get some water. It was al it was
1:02:38 very sad, right?
1:02:43 And I realized I did this at like 35. I
1:02:46 realized
1:02:48 that that little boy that was sent to
1:02:50 his room was who was controlling my life
1:02:54 when I was in my 30s.
1:02:58 I was like, well, that's [ __ ]
1:03:00 ridiculous.
1:03:02 like why is why are the rules that were
1:03:05 set up by this seven-year-old, you know,
1:03:08 put in his room by his stepfather.
1:03:12 And so one of the things you do is you
1:03:13 call and you talk to people about [ __ ]
1:03:15 And so I called my stepfather.
1:03:18 I said, "Hey, I just want you to know I
1:03:20 kind of [ __ ] up my whole life here
1:03:22 sort of living by these rules where you
1:03:25 made me go to my room every night."
1:03:28 And you know what his response was?
1:03:32 He was He was like, "What are you
1:03:34 talking about?" I said, "Yeah, every
1:03:36 night,
1:03:38 you know, I'd get sent to my room
1:03:40 because you were coming home and then
1:03:42 I'd hear your car sort of rumble down
1:03:43 the street and you'd pull in and then I
1:03:45 just live my life in my room." And he
1:03:47 goes, he goes, "Every night when I got
1:03:50 home, you were already you're already in
1:03:52 your room. I didn't He It wasn't him.
1:03:57 It was my mom.
1:04:00 My mom was doing it to protect him from
1:04:04 having to deal with me and to protect me
1:04:05 from having to deal with him.
1:04:10 So I lived like 25 years of my life
1:04:15 blaming this man for his evil motives
1:04:21 that he had no [ __ ] clue about.
1:04:25 That's life.
1:04:29 So now AI
1:04:34 has the possibility of freeing you from
1:04:38 some of those shitty decisions you made
1:04:39 when you were seven.
1:04:42 Oh man, I'm making that de debris field
1:04:45 one with Gemini right now. Oh, that's
1:04:46 awesome. That's awesome. Awesome.
1:04:48 Awesome. Awesome. Um let's I'm gonna
1:04:51 make this a little better and then I'm
1:04:52 going to publish it. Let's do uh this is
1:04:56 really
1:04:59 good now,
1:05:02 but let's build some tension
1:05:07 where the
1:05:12 firing worms
1:05:17 should
1:05:19 not be read to start.
1:05:24 That should be
1:05:28 saved for higher levels
1:05:33 and should fire a little more slowly
1:05:40 slowly. Also,
1:05:44 the static
1:05:48 square blocks
1:05:52 should be more like
1:05:59 uh like flying
1:06:03 or like
1:06:05 floating
1:06:08 space ships.
1:06:13 Maybe
1:06:17 a nod toward space invaders.
1:06:25 They should also
1:06:28 move and slowly
1:06:33 descend
1:06:35 to
1:06:38 increase
1:06:40 tension and risk.
1:06:49 The
1:06:52 boss versions of them should also fire.
1:06:56 All right, so we'll just add a little
1:06:58 more gameplay in there.
1:07:02 I like philos philosophical catharsis
1:07:05 lives. Thanks for sharing that, Kyle.
1:07:07 Perception is reality for many many of
1:07:09 us. Exactly. Listen for all of us. Joy,
1:07:12 this is life.
1:07:14 I'm not I'm not passing any judgment on
1:07:17 anyone who's every every single human
1:07:21 being
1:07:24 at some point in their childhood decides
1:07:26 that the world is dangerous and they
1:07:28 shut down part of themselves.
1:07:32 And some welladjusted people manage
1:07:34 through that quickly. And some people
1:07:36 never are even aware that they made that
1:07:38 choice back then.
1:07:40 And most of us are spending our lives
1:07:42 struggling between what's real and
1:07:44 what's not. What is actually possible
1:07:46 and what we've told oursel is or is not
1:07:49 possible.
1:07:55 So like when I when I
1:07:58 you know
1:08:02 to say that I have you know drunk the
1:08:04 Kool-Aid on AI is an understatement
1:08:06 right?
1:08:08 I think this is this is a a series of
1:08:12 the most profoundly powerful
1:08:15 technologies in the history of humanity
1:08:18 and really has the
1:08:21 the power
1:08:24 to release
1:08:28 anyone who uses it
1:08:32 to release their potential.
1:08:34 And I just I can't overstate how
1:08:38 profound that is.
1:08:45 But it's not about it's not about AI's
1:08:47 over here being the genius,
1:08:52 right?
1:08:54 To do AI right, it's not the genius.
1:08:57 You're the genius.
1:09:00 and and the ideas that you have and the
1:09:03 things that you want to put into the
1:09:04 world are the genius
1:09:08 and you kind of strap AI on like a
1:09:10 jetpack
1:09:13 and you go, "Okay, I want to do this
1:09:15 thing, but I know I don't know how to
1:09:16 code
1:09:18 and I know I'm not smart enough to do
1:09:20 the math to figure out how the bullets
1:09:22 should go,
1:09:27 but I'm going to try it.
1:09:31 And here we are.
1:09:35 What's this thing doing? I don't know.
1:09:39 Do I care? I don't care. Someone else
1:09:42 might go, "This is amazing that this is
1:09:45 doing this. I want to use the fact that
1:09:47 this is doing this to learn how to do
1:09:50 what it's doing. How is it doing? What
1:09:52 do all these things mean?"
1:09:56 Well, go figure that out. That's pretty
1:09:58 cool.
1:09:59 Oops. I gotta move here, people. Oh,
1:10:02 yeah.
1:10:06 This really good now. We have a good
1:10:08 game.
1:10:13 Yeah. Uh oh. Uncut error. Console
1:10:17 message.
1:10:19 Copy. Put that over here. I just got an
1:10:25 error.
1:10:29 And then I'm going to put also there is
1:10:32 a weird
1:10:36 yellow
1:10:38 line on the lower
1:10:44 left of the play field.
1:10:50 That should either do something
1:10:54 or
1:10:56 or go away.
1:11:00 All right,
1:11:01 building right along with you. Jim Ross.
1:11:03 Yes, I use this time to play with AI. So
1:11:06 Jim Ross is another person in here who
1:11:10 I don't know what switch he flipped in
1:11:13 his head, but at some point early with
1:11:16 this AI stuff, he just decided
1:11:20 every new thing that I learn I'm going
1:11:22 to try to apply to my business. like
1:11:24 literally every like I would be
1:11:26 surprised if Jim Ross right now is not
1:11:29 making some sort of video arcade game
1:11:32 about um self- storage management owners
1:11:36 trying to manage boxes falling from a
1:11:39 storage unit or something like that.
1:11:41 He's in the self- storage business. He
1:11:43 is he has like fully given himself over
1:11:48 to huh
1:11:50 maybe these tools can do something cool
1:11:52 for me and my business and it has
1:11:55 absolutely transformed his business his
1:11:58 life his marketing. He's now giving
1:12:00 keynote speeches about AI in in the
1:12:03 industry. He's known as the AI guy.
1:12:14 And I think I mean you've said it before
1:12:16 Jim he's he's you know he feels like
1:12:18 he's operating like a 10person company
1:12:20 at this point
1:12:22 and he's not. He's just him but he's him
1:12:26 augmented with AI.
1:12:34 I love infinite I love infinite bullets.
1:12:39 This is so good.
1:12:45 All right. It's It's a little
1:12:46 non-challenging at this point.
1:12:54 So, what's shift? Shift is overclock
1:12:56 mode. Oh. Oh, wait. Where did I go?
1:13:02 System breached. R to restart. R. Okay.
1:13:05 Uh oh. What just happened?
1:13:13 Okay, we got to do one more thing. Okay,
1:13:18 this is great now.
1:13:21 But hopulate
1:13:27 the screen
1:13:29 with more
1:13:32 descending
1:13:33 objects.
1:13:41 Starting
1:13:45 lower in
1:13:52 the frame
1:13:54 so that not
1:13:57 everything is
1:14:00 falling from the sky. It makes it too
1:14:07 easy to kill
1:14:10 everything
1:14:12 before it drops lower.
1:14:18 Let's see if it understands what I said
1:14:20 right there. All right. What software
1:14:23 are we using tonight? I'll tell you what
1:14:25 we're using there. What's your name
1:14:26 there? Blah blah blah blah blah blah
1:14:29 blah blah. So listen blah blah blah blah
1:14:32 blah blah blah. Tonight we're using
1:14:34 Claude. Now Claude is one of them.
1:14:36 They're what they call a it's a large
1:14:38 language model. It's like a chat GPT
1:14:41 except it's Claude and it's Claude.ai.
1:14:44 And then we went in there and we said,
1:14:46 "Hey, listen. Daddy kind of [ __ ] up
1:14:48 our life and I like video games. I
1:14:50 always wanted to make video games, but
1:14:52 daddy wouldn't let me go to fancy school
1:14:55 because he was drunk one night
1:14:56 apparently and I could have gone, but I
1:14:58 didn't have the guts to just speak up
1:15:00 again. So anyway, but that's not sad.
1:15:02 Anyway, whole point is I just said to
1:15:05 Claude, "Come up with some video
1:15:07 concepts, video game concepts for me,
1:15:10 and I want to make a a good video game.
1:15:11 I'm going to show daddy that I made a
1:15:13 video game." So that's what we're
1:15:14 making. And then it just broke the code
1:15:16 and just made a game that we can play.
1:15:19 So, if you've not played with Claude
1:15:22 before in artifacts, it is pretty
1:15:25 stunning.
1:15:29 Yeah, it's good.
1:15:41 Wait, what happens if I if one of those
1:15:46 Yeah. Yeah, I saw it. I saw it there,
1:15:47 Brandon. I talked about it. Never met a
1:15:50 person didn't didn't have some childhood
1:15:52 baggage. Even perfect ch childhood leave
1:15:54 scars. Yeah, this is this is the the
1:15:57 childhood stuff I'm talking about is no
1:16:00 commentary on anyone's parents or on
1:16:03 them.
1:16:05 This is one of those life things
1:16:08 that
1:16:10 you learn at some point in your life
1:16:12 that
1:16:15 sharing who you really are is dangerous
1:16:20 in some way. It's dangerous socially.
1:16:22 It's dangerous physically. It's
1:16:24 dangerous emotionally.
1:16:27 And in that moment, you make choices.
1:16:30 Ah, I get it.
1:16:33 when I sing show tunes
1:16:35 as a little boy in this house.
1:16:39 That's not good. Let me not do that ever
1:16:41 again. Right. You you make those calls
1:16:45 um very often. Well, I think everyone
1:16:49 makes, you know, at least one of those
1:16:50 in their life, probably, you know,
1:16:52 dozens.
1:16:54 So, and I think I think AI really does
1:16:57 have the the the possibility to free
1:17:00 people from that [ __ ] prison.
1:17:05 [Music]
1:17:08 or to restart.
1:17:15 This is actually a good game. All right,
1:17:16 so let's publish this. You want to go
1:17:18 play my game?
1:17:20 Okay, publish.
1:17:22 Boom.
1:17:26 Anthropics usage policy prohibits use of
1:17:28 cloud for harmful use cases like
1:17:30 violent, abusive, dis deceptive content.
1:17:33 This is kind of violent because the
1:17:35 space things are coming to get us and
1:17:37 we're going to shoot them out of the
1:17:38 sky.
1:17:40 Uh, publishing this artifact will make
1:17:41 it accessible to everyone on the
1:17:43 internet. Publish
1:17:46 and copy link.
1:17:48 Copy link. Okay, published. So now I'm
1:17:51 going to go over to the AI salon. If you
1:17:52 haven't been to the as salon, go over
1:17:55 here
1:17:57 and
1:17:58 I'm going to put this in. Look what I
1:18:00 No, I'll put it near regulars. So,
1:18:03 over at the AI salon, go there and if
1:18:06 you say click on if you click on the
1:18:08 button that says join our community,
1:18:11 that'll take you to this site. This is
1:18:13 our community site. And down the left
1:18:15 hand side here, there's an area called
1:18:16 clubs and hubs. and go to the irregulars
1:18:19 and I'll say here is my what was it
1:18:23 called? Neon.
1:18:25 Neon what? Neon.
1:18:32 Neon.
1:18:37 Okay. Yeah, look at that. No safe
1:18:39 moments.
1:18:41 Neon swarm.
1:18:45 Here's my neon swarm
1:18:51 video game I made
1:18:57 to avenge
1:19:04 that moment.
1:19:08 Some fictitious
1:19:12 daddy
1:19:15 told
1:19:16 their son they could
1:19:22 never write code.
1:19:29 Screw daddy.
1:19:35 It's pretty good. I like it. Okay,
1:19:38 there's that. Do we get a preview? No,
1:19:41 we didn't. No. Okay, fine. Let's go do a
1:19:44 screenshot of our little game, shall we?
1:19:46 Published. Can we go see it? How do I
1:19:50 get to it? Oh, copy link.
1:19:53 Go to
1:20:01 That's not how you spell moment or
1:20:02 fictitious. Five.
1:20:13 All right, there's our thing. Let me go
1:20:14 fix my spelling because producer
1:20:16 Brandon's going to yell at me. So, let's
1:20:20 Oh, do I have to Can I paste this image?
1:20:24 No, of course not.
1:20:28 Date. If you remove the preview, it'll
1:20:30 let you paste it. Oh, it will. Okay.
1:20:34 It's not how you spell video either
1:20:35 while we're doing We're there.
1:20:40 No, it didn't let me paste it. Damn it.
1:20:43 Is this our game? Where's our game?
1:20:46 No. Too big to look. No. Where did I put
1:20:50 it? Is it here? Yes.
1:20:55 Is that clip where we'll do this to
1:20:56 desktop?
1:20:59 Boom.
1:21:00 All right. Go back here. Go here.
1:21:09 Desktop. Oh, wait. So, screenshot.
1:21:16 There's our game. Open. Okay.
1:21:22 Neon Swarm video game. Okay. Here is my
1:21:24 neon swarm video game with the D I made
1:21:28 to avenge that moment.
1:21:32 That's not the right letter.
1:21:35 Some fictitious.
1:21:39 Yes, daddy told their son
1:21:43 they could never
1:21:45 write code.
1:21:48 Screw daddy. All right. Boom. Thank
1:21:53 Thank you, Brandon. I'm so proud of you,
1:21:55 son. Now you go. Go tell daddy to play
1:21:58 your little video game. Bang. Bang.
1:22:02 See what I'm happening? See what's
1:22:03 happening here?
1:22:05 [ __ ] Bishop just become number became
1:22:08 number 18 in the fan club. If you're not
1:22:10 in my fan club on TikTok, what are you
1:22:12 doing with your life?
1:22:15 Be like [ __ ] Bishop.
1:22:18 Why is he swearing so much? It's just
1:22:20 the guy's name. Okay.
1:22:23 Not flipping bishop.
1:22:28 All right. Let's play the game. R
1:22:31 R
1:22:36 Oh, wait. These things hitting the
1:22:39 hitting the ground should do something.
1:22:42 That should be bad.
1:22:56 It's a good game.
1:22:58 It's good game, people.
1:23:05 Now,
1:23:07 let's go back to the where we were at
1:23:09 the beginning. What's the chain of
1:23:11 events?
1:23:13 What's the chain of events?
1:23:16 Well,
1:23:20 we got to talking about things we didn't
1:23:22 think we could do in our life and maybe
1:23:24 programming was that thing for you. Then
1:23:27 we thought, huh,
1:23:29 I heard
1:23:31 that AI can help you code things. If you
1:23:34 don't know how to code, AI can help you
1:23:35 code things. And so, we knew enough
1:23:38 about AI that we thought Claude's pretty
1:23:41 good at that. It's got this thing called
1:23:44 artifacts where you can let it code and
1:23:46 then it can turn it into an application
1:23:48 that you can actually share with the
1:23:50 world.
1:23:52 And so in less than I don't know what
1:23:54 that was, maybe 45 minutes, probably 35
1:23:59 of which were me just whining and
1:24:01 rambling
1:24:04 in about 10 minutes of actual work time.
1:24:10 We came up with five video game
1:24:12 concepts, built and published a video
1:24:15 game. That's fun.
1:24:19 Is it perfect? No.
1:24:22 Could I have written that if I wanted
1:24:25 to?
1:24:29 Not in that amount of time. Pretty
1:24:30 impressive, right?
1:24:35 You have to upload an image, not paste
1:24:37 it.
1:24:39 Um, let me go to irregulars. Lower here
1:24:41 in irregulars. I think Steo put
1:24:42 something in. Just me doing what?
1:24:48 Feel the quality, young lady. Oh, the
1:24:50 carpet salesman.
1:24:56 Y'all are funny.
1:24:59 Oh my god, that's funny. Steo, look at
1:25:01 Oh, that's cute. I love the little
1:25:03 mushroom house. Me and Champy up on the
1:25:04 roof. Nice. Like it.
1:25:11 All right. So, there you go.
1:25:15 Go play my video game
1:25:19 and or be like Jim Ross. You should be
1:25:21 making one. Use this time. So, if you're
1:25:23 new here
1:25:26 and and you're befuddled,
1:25:28 if you're discombobulated, you're like,
1:25:30 I don't I don't get this channel.
1:25:34 It's kind of entertaining. It's kind of
1:25:36 educational, but it's really neither.
1:25:43 This channel isn't about me.
1:25:47 All I'm doing here
1:25:50 is creating space for anyone who shows
1:25:53 up to be in the conversation
1:25:56 about AI as a possibility. What does it
1:25:59 make possible? What does it make
1:26:00 possible? Huh? What if I go try this?
1:26:02 What if I go try that? Oh, that makes
1:26:04 sense what he said there. Yeah, let me
1:26:05 go try this. Let me go try that. Hey,
1:26:07 does anyone have any experience doing
1:26:08 blah blah blah? Maybe I should go join
1:26:11 the AI salon and help. And right, that's
1:26:14 what this is about.
1:26:17 So, the reason you come here is to be in
1:26:19 the conversation.
1:26:22 Jim Ross is one of those guys who's like
1:26:24 he comes in here just fully open to ah
1:26:27 yeah I I kind of enjoy Kyle's stupid
1:26:30 jokes and I'll as soon as he teaches me
1:26:33 shows me how to do something I haven't
1:26:35 tried before I'm going to go do that
1:26:36 right now right while he does it
1:26:45 changed his
1:26:52 So that's what this is about. All right,
1:26:54 I'm going to get out of here.
1:26:57 And a lot of laughs. What better way
1:26:59 than Frogger than that Frogger one with
1:27:03 the logs from a while back. Yeah,
1:27:05 exactly. Oh, way better than that
1:27:07 Frogger one. Well, yeah. You know what?
1:27:08 You know, it's funny, James. This We're
1:27:11 now at Claude 4. the the Frogger game
1:27:14 was was Claude 2.5 or three. Then we
1:27:18 tried it again with Claude 3.5. This is
1:27:20 now Claude 4. So the the quality of the
1:27:24 video game here is just better. Just the
1:27:27 the tool is better, right?
1:27:30 More sophisticated. Like if I wanted to
1:27:33 put real time into this game, I'd
1:27:34 probably spend another two or three
1:27:37 hours on it and really [ __ ] dial it
1:27:40 in.
1:27:42 But can you imagine that
1:27:46 we now live in a time
1:27:52 where someone like me who's got no
1:27:54 coding skills, moderate tech and project
1:27:58 product management skills, moderate
1:28:00 interface skills, good conceptual
1:28:03 skills, someone like me can actually say
1:28:06 with a straight face, if I really wanted
1:28:09 to put in more than half an hour on that
1:28:11 game. I could make it really good in
1:28:13 like three or four hours, two or three
1:28:15 hours.
1:28:18 And I know that to be true.
1:28:24 This is science fiction. We live in
1:28:25 science fiction.
1:28:28 If you come here a lot, I know we kind
1:28:30 of normalize the remarkableness of the
1:28:33 era we live in, but remind yourself how
1:28:36 [ __ ] lucky we are that we get to live
1:28:39 in a time in history where the knowledge
1:28:42 of humanity that's come before us has
1:28:45 been compressed into a softballsized
1:28:49 sphere that we can just tap into.
1:28:55 The brilliance of image makers and music
1:28:58 makers and coders and writers and
1:29:01 business people and psychologists and
1:29:05 academics
1:29:06 and people into animal husbandry and
1:29:10 plant genetics. All of that is now at
1:29:13 our fingertips. All of it. All of
1:29:15 humanity
1:29:18 instantly available. Anything you
1:29:21 thought you were not good at is no
1:29:22 longer true.
1:29:25 What are you going to do with your life?
1:29:28 All right, that's it. I had the
1:29:31 programming schools with Forran 4 typing
1:29:34 it on cards and feeding them in into an
1:29:36 IBM card reader. So, James, that was I I
1:29:40 was going to either go into theater or
1:29:41 computer science. And I said, I got to
1:29:43 go into computer science. It's the
1:29:44 responsible thing to do. I walk into my
1:29:47 first computer science class. I walk,
1:29:49 it's a room full of IBM 3870 mainframes.
1:29:53 And this [ __ ] dweeb dweeb nut walks
1:29:55 out of the back room with like big thick
1:29:58 glasses. It's like like these glasses.
1:30:00 He walks out. He walks out and he's
1:30:04 holding a stack of punch cards this
1:30:06 tall. He goes,
1:30:09 "This is a computer program. Each card
1:30:12 represents a single line of code which
1:30:15 is encoded in 80 characters across the
1:30:20 and he said you're going to stick this
1:30:22 stack of cards in this reader and it was
1:30:24 like
1:30:26 reading the cards. The card's just
1:30:27 slowly going down
1:30:29 [Laughter]
1:30:32 and he goes and and it will run the
1:30:34 computer application and the computer
1:30:37 application generally is analyzing data
1:30:40 in the stored on magnetic reels and and
1:30:43 we're running reports to analyze that
1:30:45 data and we're outputting that on these
1:30:47 large printers.
1:30:53 I was like I was like, "Uh,
1:30:57 yeah, no, that's not going to be my
1:31:00 life." So, I walked I walked out of that
1:31:04 class and down to some guidance
1:31:06 counselor's office. I say I said I'd
1:31:09 like to change my uh major to theater
1:31:11 and uh I'll do a minor in computer
1:31:14 science,
1:31:15 but that ain't that ain't going to be my
1:31:18 life
1:31:21 because all I knew about computers was
1:31:23 that's how you made video games. I
1:31:24 didn't know [ __ ] about Forran and
1:31:27 corporate computing. I had no [ __ ]
1:31:30 interest in it. So anyway, all right,
1:31:33 I'm gonna get on out of here. Um thanks.
1:31:35 Hope you hope you guys had fun tonight.
1:31:38 I really want you to think about
1:31:41 stop limiting your own thinking and
1:31:43 catch it when other people say, "What
1:31:45 tool did you use?"
1:31:49 And talk about the chain of activities
1:31:53 that allowed you to produce a something
1:31:54 that someone asked about or if I want to
1:31:57 make one of those, what tool do I use?
1:31:59 Well, it's not a tool. It's a whole
1:32:01 process, whole chain of things that has
1:32:04 to happen. And some of them are just me
1:32:06 and some of them are a whiteboard and
1:32:08 some of them are chat GBT. What are you
1:32:10 talking about? Right.
1:32:13 Um, so since Friday is the 4th of July,
1:32:16 tomorrow night is going to be Thursday
1:32:19 night date night. So we'll be here
1:32:22 tomorrow. Um, but but I'm not going to
1:32:25 be here Friday. Gabby's got an art show
1:32:27 opening. It's the 4th of July. just
1:32:30 whole bunch of [ __ ] going going on.
1:32:33 So, tomorrow night 8:00 here, be here or
1:32:36 be square.
1:32:40 If I did something that rhymed with
1:32:42 here, I'd probably get lifetime banned
1:32:45 from Tik Tok. Um,
1:32:51 oh, good lord. All right. Um, yeah,
1:32:55 that's it. I'm out of here. Um, trying
1:32:58 to think if there's anything else. No.
1:33:00 Join the salon.
1:33:02 Keep coming back to these. Do me a
1:33:04 favor. How many folks are here? I got 46
1:33:06 people here. I got 31 over there.
1:33:12 Here's here's a request. I don't ask
1:33:13 much of this community, but here's my
1:33:16 request.
1:33:18 If you know someone that used to come to
1:33:20 these that has stopped coming to these
1:33:22 or if you know someone who you think
1:33:25 would
1:33:27 benefit from getting their head around
1:33:29 AI, invite some new people to this.
1:33:32 I'd like to start expanding this and
1:33:34 growing it. And I haven't been paying
1:33:37 any attention to that. And much to the
1:33:40 just chagrin of producer Brandon who's
1:33:42 been really trying to to get it get me
1:33:44 to do stuff. Um I want to start working
1:33:47 on that. I want to I want to start
1:33:48 expanding our footprint. I think what
1:33:50 we're doing here is actually important
1:33:53 and as irreverent as I am with this
1:33:55 stuff, um I'm deeply deeply passionate
1:33:59 about it. So, if you would be so kind,
1:34:02 um bring a friend. Tomorrow's Thursday
1:34:06 night date night. Bring a [ __ ] date.
1:34:09 Um the date won't last that long.
1:34:12 They'll probably leave early, but um but
1:34:15 do that. I'd like to start to get this
1:34:16 out to some more people. All right,
1:34:18 cool. All right, have a good night and I
1:34:20 will see you on tomorrow.