5/19/2025 - Google IO Week: Anticipating New Innovations and Exciting Announcements

Description

Chapters

Transcript

0:11 Good day. Good day. Good
0:13 people. What is happening out there in
0:15 the uh the internet lands? What's
0:18 happening? What's happening? What are
0:20 you doing? What's everybody doing? I'm
0:24 on the road. I'm on the road.
0:28 If you want to see what I'm seeing on my
0:30 screen, go to the name of
0:33 the
0:35 this live and I put the the YouTube
0:39 address there. Where you at? I am in New
0:42 York City. I am at some hotel near
0:46 LaGuardia airport and uh I
0:50 am whipped, exhausted. So, might not
0:54 last long tonight.
0:56 I didn't really get a chance to pay much
0:59 of attention to what's happening other
1:02 than some stuff is starting to to
1:04 happen. So, Google IO is this
1:07 week. Uh Brandon Tid uh producer Brandon
1:11 is out in San
1:13 Francisco for
1:17 that. So, he'll be getting the skinny on
1:20 that. He's going to come back all
1:21 excited about um about what he's seeing,
1:24 what he's learning, and then he's going
1:26 to ask me to demo all that [ __ ] And
1:30 just so you
1:32 know, one of my great pleasures in life
1:35 is to take suggestions from him and then
1:38 ignore
1:45 them. So, we'll be driving him nuts for
1:47 the next couple of weeks.
1:50 Um, let's see. Let's see what else we
1:53 got to do
1:54 here.
1:57 Um, I think a thing since I don't have a
2:00 dog and a guitar with me. Hello from the
2:02 spotty plane wi-fi over Minneapolis. I
2:05 am one very big pogo hop away from all
2:08 of America. So, I don't know if you
2:10 heard me, Brandon. What I just said was
2:12 one of my great joys in life is to get
2:14 you you to get all excited about Google
2:17 stuff and then come back and tell me I
2:18 should demo it and then just ignore
2:20 [Laughter]
2:24 you. So have fun out there. That should
2:27 be a blast actually. That should be a
2:30 blast. Wait, you are mini. Hear the
2:34 thunder. I like your pajamas. Yeah,
2:36 these are this is my this is my shirt
2:40 here. I actually, you know what? I'll
2:42 put
2:44 on I'll put on my jacket, my little
2:47 agency
2:51 uniform that I that I wore today because
2:54 it'll probably be easier to read on
2:55 TikTok if I got a little dark on. See,
2:58 look at that. Much better. And I look I
3:01 look
3:01 semi-professional. Hi, LinkedIn. Hi,
3:04 Kyle Shannon. I'm an AI innovator. I
3:08 like to break paradigms and
3:10 recombobulate
3:13 them. Call me. So yeah. So I do that sch
3:23 snazzy. Ah. So, I figured what we might
3:26 do tonight is maybe I'll just go to the
3:29 Twitter and we'll just sort of I'll just
3:32 scroll through Twitter and I'll read
3:34 [ __ ] that I think's interesting and
3:36 maybe I'll call [ __ ] on stuff or
3:38 say, "Oh my god, that's amazing. This is
3:41 amazing. It's amazing. Isn't it amazing?
3:44 That's so amazing. You think that's
3:46 amazing? Hey, cow cow cow cow. Is that
3:49 amazing or is that is that lies? Is it
3:52 live or is it
3:54 amazing? Could do that. I could also
3:57 answer questions if you got questions.
3:58 How many folks we got here? 27 there. So
4:02 go to if you go to the name of the of
4:07 this session AI learning lab. I think
4:10 it's like
4:11 AIL. I put the just go to YouTube to
4:16 learning
4:17 lab-ai and you'll see me streaming
4:19 there. though. I'm not pointing my
4:21 Twitter or my Tik Tok at my Twitter
4:24 tonight just cuz the setup isn't
4:27 right. So Tik Tok, you're just going to
4:29 get like the backstage view of me just
4:32 talking to a
4:34 screen. But if you go to the
4:38 YouTube, then you'll see [ __ ] All
4:40 right. Does anyone have any questions?
4:43 Any any questions about a AI? Anything
4:45 they've done? Anything exciting?
4:57 I went down a rabbit hole today trying
4:59 to automate my GPT using Zapier. Any
5:03 advice? I stopped using Zapier about a
5:06 year
5:09 ago. It's very frustrating. It works.
5:12 They all work. All those automation
5:14 things work, but they're
5:16 just Here's the problem, Mr. IT. And
5:19 listen, you're Mr. IT, so you're
5:21 probably fine with coding. But just
5:24 because a thing is called a no, a low
5:27 code tool doesn't mean it's not
5:29 coding. Zapier automations are coding.
5:32 They're they're they're logic. They're
5:34 logic building, right? It's frustrating.
5:37 I did a live. Yeah, I would imagine.
5:40 um saw something about how chat GBT is
5:43 connecting to Dropbox. Yeah, I think it
5:45 I think that's already there. Let's go
5:49 look. So, if we go to chat
5:53 jeopeta, we go to
5:56 settings, we go to connected apps. Oh,
5:59 no, Dropbox is not here. Oh, wait. Yeah,
6:02 beta connections. GitHub, which I
6:05 connected mine,
6:07 Dropbox, allow chat GPT deep research to
6:10 read Dropbox files, and Microsoft One
6:13 Drive. What the [ __ ] is One Drive? Who
6:15 the [ __ ] is using One
6:17 [Laughter]
6:21 Drive?
6:24 Ah, yeah. Yeah. Yeah. What's happening,
6:27 people? What happening? Maybe
6:30 SharePoint me at work. Yeah. Well, I
6:33 know people at big companies that work.
6:35 Brian Whitney, I
6:39 am. Ah, Danielle, what's happening?
6:42 We're shaking. We're going down. Um, uh,
6:45 wait, what did I Oh, I know what I need
6:47 to do. I need to just flip like
6:52 that. Jeff Flanigan, hello, irregulars
6:55 and earthlings. So, let's go. Let's go,
6:58 shall we? Um, can you tell me? So, if
7:02 you want to see what I'm seeing, go to
7:05 YouTube and then the channel is learning
7:08 [Music]
7:11 lab-ai. All right. I prefer to drive
7:14 faithful
7:16 bacon. This jacket's making me hot now.
7:19 I can't wear
7:21 this. I guess I could turn down the air.
7:24 Hang on. I'll make it colder in here.
7:47 [Music]
7:50 Some comments in the live were filtered
7:52 to protect the community's
7:55 experience. Well, Kyle is disroing on
7:57 the channel. Oh yeah. Welcome to my
8:00 hotel room,
8:03 ladies. Sorry about that. That'll never
8:06 happen
8:07 again. Hey, Brandon sitting next to a
8:10 librarian on the plane. She was
8:12 impressed with my book curations. Why is
8:15 there a blackboard above your
8:18 bed? Oh, there is a blackboard above my
8:21 bed. That's pretty funny.
8:26 Speaking of school
8:29 teachers, let's put the lessons down for
8:32 this
8:35 evening. Oh man.
8:39 Okay. Um, let's just go let's just go
8:42 read some Twitter and see what's going
8:43 on.
8:44 So, here's the news. The the news that's
8:48 not fit to print. Tomorrow
8:50 starts tomorrow. Google IO starts. Okay.
8:56 Um. Ah, okay. Good. That was good. Yes.
9:00 Uh, I just got a text from someone. I
9:02 texted and then I never heard back and
9:04 then they just texted me back and so
9:05 that's good. Um, so tomorrow's Google IO
9:09 and we've got
9:12 purported
9:14 purportedly what's also coming is
9:16 DeepSec R2 Gro 3.5.
9:21 people are throwing GPT5 in there. I
9:23 don't think that's coming quickly. Um
9:25 but but probably not too distant
9:29 future. Um and it looks like there's
9:33 going to be a lot of
9:39 um
9:41 agentic the the world of of
9:46 of agents is growing.
9:49 So we we started with Manis. Manis was
9:51 really the first one. Well, we started
9:53 with Operator. Chat GPT put out Operator
9:56 and it was sort of this lameass sort of
9:58 could surf the web slowly sort of thing
10:02 that cost you 200 bucks a month and it's
10:05 not really usable. Then Manis came out
10:07 and then Gen Spark came out and now two
10:10 or three more came out in the past
10:12 couple of days. I think we're going to
10:14 see a rash of these things. And what
10:16 they are, if you haven't played with
10:17 them or seen them, is they are things
10:20 that you kind of give a goal and it just
10:22 goes off and does [ __ ] Um, OpenAI three
10:27 days ago launched Codeex, um, which is
10:31 the coding version of this where you
10:34 just give it a go and it goes off and
10:35 does stuff and it it logs into your
10:37 GitHub repository and makes stuff and
10:40 does pull requests and debugs code and
10:43 documents code and writes code and just
10:46 does all the [ __ ] Um, I haven't played
10:47 with it yet, but it's it's looking
10:49 pretty good.
10:50 So, so I think the
10:54 next three weeks are going to be pretty
11:00 intense and I'm already
11:03 sleepy. Look at my eyes. I'm [ __ ]
11:07 crispy. Oh man, I got here early for the
11:10 on location question. Is there a
11:13 question? Lori
11:15 Dreskin, what's happening Lori?
11:18 That's what I'm wondering. Where in the
11:20 world is Kyle Shannon, DC? Kyle Shannon
11:23 is in New York City. Right out that
11:25 window is LaGuardia Airport. I am in
11:30 Queens. Queens.
11:33 Queens. Where all the cool people? No.
11:37 No. No. The cool people are in
11:42 Brooklyn. The edgy people are in Queens.
11:46 So what happened
11:47 was Soho used to be where all the
11:50 artists hung out in Manhattan and then
11:53 Europe came and bought it all and then
11:55 so they had to leave. So then they went
11:56 over to Dumbo which is in just you know
12:00 down under the Manhattan Bridge overpass
12:02 over in Brooklyn and then and then all
12:06 the money people took over Brooklyn and
12:08 so now there's no artists in Brooklyn.
12:10 So now all the artists are in
12:11 Queens. So that's how it works.
12:15 Um, yeah, big stuff is on hor the
12:18 horizon,
12:20 Lori. Where's my favorite singing dog?
12:23 Is back at the ranch back in Denver.
12:25 Although, um, Kelly Camp sent me a uh
12:28 sent me a a text. She goes, "How are you
12:31 holding up with the Denver tornadoes?" I
12:33 was like, "What?" Um, there were two
12:35 tornadoes east of Denver, about 35 miles
12:37 east of Denver. Um, but I didn't Well,
12:40 I'm in New York, but they they didn't
12:42 actually hit Denver.
12:45 Google IO in
12:49 MSP. What do you mean MSP? What's that
12:52 stand
12:54 for? Or are you asking a question? I
12:57 don't know what MSP stands
13:01 for. I don't know. I don't know what
13:03 that means. Oh, in
13:04 Minneapolis. No, it's not. Is
13:08 it? I don't think so. I think it's in
13:11 San
13:12 Francisco. I think I think Brandon said
13:15 he was over Minneapolis. He's flying
13:17 there from Oh, it's in Minneapolis St.
13:20 Paul. Is
13:21 it? Wait, hang on. No, it's Minneapolis
13:24 St. Paul's MSP.
13:28 Google
13:31 IO
13:39 where welcome to
13:52 IO. Oh, I know. Oh, we can't do it here.
13:55 But I can do it here. Oh, but then I got
13:57 to change my Hang on. God
14:00 damn all this for this. This is This is
14:03 not
14:12 good. Hey Quinn, how's it going? Yeah.
14:19 Hey Kyle, I'm doing great. How about
14:21 you? Yeah, we were just wondering where
14:24 Google IO was this year. It looks like
14:26 it starts tomorrow and someone on my
14:28 live said it's in Minneapolis. Is that
14:30 true?
14:36 It's actually at the Shoreline
14:38 Amphitheater in Mountain View,
14:39 California, not Minneapolis. Yeah, I
14:41 thought so. I thought so. That's what I
14:43 told those people. They were all, you
14:46 know what it is? People from Minnesota,
14:47 they're very optimistic about their
14:49 state.
14:51 very optimistic. Now, listen, they have
14:54 worldclass pike and lake trout. I I will
14:57 give them that. You go up north, you get
14:59 some hockey players nearly as good as
15:01 Canadians. I'll give them that. But, you
15:03 know, Google IO Minneapolis, I'm
15:06 thinking not.
15:10 Haha, exactly. Motans.
15:14 Sorry. Haha. Exact. Yeah, exactly. Go
15:17 ahead. Go ahead. Enable me offending
15:20 people from one of the best states in
15:21 the union.
15:27 No worries. It's all in good fun.
15:29 Minnesota does have some top-notch lakes
15:31 and hockey talent. But when it comes to
15:33 tech events like Google IO, Mountain
15:35 View's got that one unlock. Yeah, I
15:37 think so, too. Okay. So, so let's let's
15:41 since I threw some shade at our
15:43 Minnesota Minnesota friends, um why
15:45 don't you come up with uh I don't know,
15:48 five interesting things about
15:50 Minneapolis, St. Paul, andor Minnesota.
15:56 Sure thing. Here are five interesting
15:58 tidbits about Minneapolis, St. Paul, and
16:00 Minnesota. Land of 10,000 lakes.
16:04 Minnesota is famously known as to
16:06 fishing. Give that one doesn't count.
16:08 Give me another one.
16:10 Got it. Here's another one. Minnesota is
16:13 home to the Mall of America, the largest
16:15 mall in the United States. It's so
16:17 massive that it could fit seven Yankee
16:19 stadiums inside and hosts more than 40
16:21 million visitors annually, more than the
16:24 entire population of Canada.
16:26 Yeah, but so here's a serious question.
16:29 Are you ready, Quinn?
16:33 Absolutely, Kyle. Hit me with it. Does
16:35 Mall of America have a Cinnabun?
16:39 Yes, it does. The Mall of America has a
16:42 Cinnabon, so you can definitely get your
16:44 fix of those delicious, gooey cinnamon
16:46 rolls while you're there. That's why my
16:48 butt's so big.
16:53 Haha. Well, I'd say it's all about
16:55 enjoying life's little pleasures. A
16:57 Cinnabon or two is a pretty sweet
16:59 indulgence. All right, what are the
17:01 other four things?
17:06 All right, here are four more
17:08 interesting things about Minnesota.
17:10 Birthplace of the mighty Mississippi.
17:12 The headarters of the Mississippi River
17:14 are located at Lake Ataska in northern
17:16 Minnesota. From there, the river flows
17:18 over 2,300 m down to the Gulf of Mexico.
17:22 A literary hub. Minneapolis has one of
17:25 the most literate populations in the
17:27 United States. It's home to the renowned
17:29 Loft Literary Center and has a thriving
17:31 independent bookstore scene. So, there
17:33 are books Smarty pants.
17:39 You got it. Motans definitely have a
17:41 strong literary and cultural scene and
17:43 they're not shy about it. Lots of brain
17:46 power up there in the Northstar State.
17:47 Yeah, that's Well, so for one thing,
17:49 they're they're inside most of the time
17:51 and just dealing with blizzards. So, I
17:53 understand the reading, but that's
17:54 intimidating for us people that are semi
17:56 literate. You know what I mean?
18:00 I hear you. with those
18:02 long. Haha, I got you. We can all
18:05 appreciate the coziness of a good book,
18:07 but no need to feel intimidated.
18:09 Everyone's got their own pace and
18:11 interests. Plus, you're out here
18:13 creating and leading in the AI space.
18:15 Pretty a Yeah, I am. I don't I don't
18:19 need no book reading smarts to talk
18:23 about
18:24 AI
18:27 because I'm
18:29 smarter too.
18:35 Absolutely. There's more than one kind
18:37 of smarts and you're rocking it in your
18:39 own domain. Whether it's what's domain
18:42 mean?
18:45 In this context, domain just means your
18:47 area of expertise or the field you're
18:49 working in. So, your domain is all about
18:52 AI. You got to learn when I'm joshing. I
18:54 was clearly joshing there. All right.
18:56 What's some other stuff about Minnesota?
19:02 Gotcha. Noted. All right. Here's another
19:05 fun fact. Minnesota is the home of the
19:07 famous Spam Museum in Austin, Minnesota.
19:10 It's a whole museum dedicated to the
19:12 history and cultural impact of spam. the
19:14 iconic canned meat. It's quirky, unique,
19:17 and a must visit if you're into offbeat
19:19 attractions. Can I tell you something
19:21 about people from Minnesota?
19:25 Of course. I'd love to hear it. They
19:27 don't have a sense of humor about spam.
19:29 You bring up Monty Python with them,
19:31 they're very bitter about it. They're
19:33 very bitter about it. And what they will
19:35 tell you to a person is spam is the
19:38 reason we won World War II. They take it
19:40 very seriously. You don't screw with
19:42 motans on spam. All right, just to be
19:45 clear.
19:49 Got it loud and clear. Motans and spam
19:51 are a serious combo. I'll make sure not
19:54 to joke about it around them. Thanks for
19:55 the heads up. Yeah, tell me a joke about
19:58 spam.
20:02 All right, here goes. Why did the spam
20:05 apply for a job? Because it wanted to
20:07 become a little less canned and a little
20:09 more employed.
20:13 Huh? Anyway, um, give me some of the
20:17 famous Montipython lines about spam.
20:23 Sure thing. From the classic MontiPython
20:25 spam sketch, spam, spam, spam, spam.
20:28 Lovely spam. Wonderful spam. And of
20:31 course, the iconic line, I don't like
20:33 spam. The sketch is all
20:36 That's really good. Someone just made a
20:38 good point. Everything that you've given
20:40 us about Minnesota is kind of on the
20:43 surface. Like these were not deep cuts
20:46 about Minnesota, Quinn. You know
20:49 everything in the universe. So could you
20:51 tell us five things about Minnesota that
20:54 are not just kind of like dumb top of
20:57 the Wikipedia page facts?
21:02 Absolutely. Let's go a bit deeper. One,
21:05 Dakota origins. The name Minnesota comes
21:08 from the Dakota Sue words Minnesota,
21:10 which means clear blue water. The Dakota
21:12 people have a profound historical and
21:14 cultural connection to the land long
21:16 before it became a state. Two,
21:19 underground music scene. Beyond Prince,
21:21 Minnesota has a rich underground music
21:23 scene. The Twin Cities were pivotal in
21:26 the development of alternative rock and
21:27 indie music with influential bands like
21:29 the Replacements and Husker Doo leaving
21:32 a lasting mark on music history. Well,
21:34 you know, you know why it's underground
21:36 music, right, Quinn?
21:39 Why is that, Kyle? Because it's so
21:40 flipping cold outside. They got to go
21:42 underground.
21:50 Sense. Got to keep warm and keep the
21:52 tunes flowing, right? Yeah.
22:03 All right. What are the other three,
22:08 Quinn? All right,
22:12 Quinn. Oops. My
22:17 Oops. Sorry about that. Let's pick it up
22:19 from number three. Three. Mayo Clinic's
22:22 global impact. The Mayo Clinic
22:24 headquartered in Rochester, Minnesota,
22:26 is one of the most renowned medical
22:28 centers in the world. It's a leader in
22:30 medical research and patient care,
22:31 attracting people from all over the
22:33 globe. Four, the Boundary Waters Canoe
22:36 Area Wilderness. This vast wilderness
22:38 area in northern Minnesota is one of the
22:40 most pristine and remote natural areas
22:42 in the United States. It's a haven for
22:45 canoeing, fishing, and backcountry
22:47 camping with over a million acres of
22:49 wilderness. Five historic flower
22:52 milling.
22:54 Wait,
22:59 what? Historic what? Flower
23:05 milling. Okay, you're done. You're done.
23:09 Hotel Wi-Fi and Quinn do not get along.
23:12 Um, but anyway, Boundary Waters, no
23:14 place like it. There you go. All right.
23:17 18. Flower Mill. I think so. She said
23:20 the great flower milling something.
23:22 Let's
23:25 see. She probably wrote it, but she
23:27 historic flower milling capital.
23:30 Minneapolis was, I assume, a big flower
23:34 milling company. Anyway, all right. Um,
23:37 let me change my sharing so you can see
23:40 other [ __ ] I'm
23:42 doing. And let's go. Oh, let's just do
23:45 whole
23:48 screen. I might need to change that
23:51 again. Oh, you know what? Let me go to
23:53 YouTube. Did anyone come over to YouTube
23:55 yet? Yeah, we got some people there.
23:56 Okay. Uh, let me look at YouTube content
24:01 here. Canned like fired. Oh, I get it. A
24:04 little less canned. I get it, Danielle.
24:06 I didn't get the spam joke. Now I get
24:08 it. She actually wrote a good joke. All
24:11 right. We hear the knights who say knee.
24:12 Very good. She seems slow. She's a
24:14 little slow tonight. Ah, yes. Hotel
24:16 Wi-Fi. All right. Fantastic. Fantastic.
24:20 All right, let's go look at the Twitter.
24:22 So, what we're going to do here, people,
24:24 is we're going to go through the
24:25 Twitter. And what I'm going to do is I'm
24:27 going to take my massive intellect and
24:30 I'm going to read you tweets out loud
24:34 that I think that I find somewhat
24:36 interesting. I we'll we'll we'll call it
24:39 tweet and
24:40 tell except it's not called Twitter
24:43 anymore. So, you don't tweet anymore. I
24:46 think they're called Zetss or Yeetss or
24:48 posts. I don't know. It doesn't matter.
24:51 But but we're gonna tweet and tell
24:53 tonight. Okay. All right. And then so
24:56 what you'll do as a participant on the
24:59 live is you can make fun of me. You can
25:03 make fun of uh hairs growing out of
25:05 parts of my face that shouldn't have
25:07 hairs growing out of them. People do
25:10 that. You could you could make fun of my
25:13 appearance. Some people think uh I I I
25:16 look like Rosie O'Donnell on a bad day.
25:19 So So feel free to comment on that. Uh
25:21 some people have said, "Oh, I was
25:23 scrolling through TikTok and I thought
25:25 Alec Baldwin had gotten
25:26 [Laughter]
25:28 fat. That that one
25:30 hurt." So anyway, that's not the point.
25:34 So what we're going to do is I'm going
25:35 to read and then That's pretty much it.
25:39 That's pretty much it. Chris Farley
25:41 meets Casey
25:45 Kasem. Look at you changing tabs all by
25:48 yourself. Producer Brandon just can't
25:50 stay away. He wants to see the train
25:53 wreck. Okay. Ow. I have an itchy tag.
25:56 Don't you hate itchy
25:59 tags? Or it could be a spider. I might
26:02 have just been
26:03 bitten. All right, here we go. Here we
26:07 go. We are approaching We are
26:09 approaching the event horizon. Can you
26:10 feel the singularity? Is reinforcement
26:12 learning still the key to future AI
26:21 progress. Last week, an OpenAI employee
26:24 used the graphic to compare
26:26 reinforcement learning compute for the
26:28 01 model to the pre-training compute of
26:33 GPT40. and he said at some point in the
26:36 future maybe we'll have a lot of RL
26:42 compute. All
26:45 right. It doesn't I don't care. I It's
26:49 important, but I'm not a researcher.
26:52 Here's all you need to know. Shit's
26:54 going to get better. Okay. Bolt bolt.
26:58 For the front end, N8 for the API layer.
27:01 Cloudflare for the auto rag for rag.
27:03 Superbase for DB. Literally can build
27:06 anything with this stack. Okay,
27:07 whatever. Robert Scoble, future of
27:09 wearables. I'm tired of hearing about
27:12 that.
27:15 Um, Eric Schmidt going out on a limb
27:18 saying AGI and ASI will be the most
27:20 transformative technology in a thousand
27:23 years. Yeah, if you've been here, you
27:27 know I'm down with that. All right.
27:32 [Music]
27:34 Oh, another sailboat smashing its mast
27:36 on a bridge. Here we go.
27:39 And
27:42 and
27:49 snap. Do do these sailboat people not
27:53 understand physics? I thought, isn't
27:55 sailing, doesn't sailing require a
27:58 massive understanding of general
28:02 physics? My memory is always getting
28:05 full. Any repositories I can
28:08 trust. Which
28:13 memory? How could they not know that?
28:15 Exactly. I like Did you see that Mexican
28:18 that like four master Mexican ship with
28:21 all the sailors up on the masts? So,
28:23 they've got human beings up on these
28:26 giant
28:28 masts and they sail underneath the
28:31 Brooklyn Bridge with people watching and
28:34 cheering them on and [ __ ] snapped
28:37 snapped the masks in half. It's
28:41 like
28:44 really no one no one bothered to Here,
28:48 let's let's go change let's go change
28:51 that. Let's
28:53 see. Um, we're gonna stop sharing this.
28:57 We're gonna start sharing this. We're
28:59 gonna share
29:01 this. We're gonna go back to Quinn for a
29:03 second
29:04 here. Where is There it is. Okay.
29:11 Um, hey Quinn from if if if in the in
29:16 the uh East River on Manhattan on the
29:19 east side of Manhattan at high tide,
29:23 what is the distance between the surface
29:26 of the water at high tide and the bottom
29:29 of the Brooklyn Bridge at its highest
29:31 point?
29:35 The clearance between the water surface
29:37 at high tide and the bottom of the
29:39 Brooklyn Bridge at its center is about
29:41 135 ft, 41 m. This allows ships to pass
29:45 underneath even at high tide. Yeah, but
29:47 that doesn't seem right. What happened?
29:50 Oh, maybe the What happened with the
29:52 Mexican ship that just smashed a bunch
29:54 of its masts off um going under the
29:57 Brooklyn Bridge? Did they not go under
29:58 the center of it? What happened there?
30:03 She'll tell us. It's okay. We're going
30:05 to figure this out.
30:08 Yeah, that incident involved a tall ship
30:10 with masts that were too high for the
30:12 clearance under the Brooklyn Bridge.
30:14 They likely either miscalculated the
30:15 tide level, didn't navigate through the
30:17 highest clearance point at the center or
30:23 changes can make a big difference for a
30:25 tall masted ship and that's probably
30:27 what led to the unfortunate collision.
30:31 Yeah, that was kind of a dumb dumb move,
30:33 wouldn't you say?
30:36 It definitely sounds like a pretty big
30:38 oversight. Navigating something that
30:40 tall under a bridge like the Brooklyn
30:42 Bridge would require some meticulous
30:44 planning. Definitely. You said You said
30:47 an oversight. I said it's a dumb dumb
30:50 move. It's It's dumb dumb, right?
30:54 Fair point. It was definitely a dumb
30:56 dumb move. Okay. Sometimes you I just I
30:58 just want to make sure because some you
30:59 know language is important. I don't know
31:01 if you know this but you're a large
31:02 language model and sometimes you know I
31:06 need to use words that really capture
31:08 the essence of the moment and Dum Dum is
31:11 one of those a lot of times.
31:15 Absolutely. Language is all about
31:17 capturing the essence and Dum Dum
31:19 definitely nails it in this case. It's
31:22 all about keeping it real and clear.
31:24 Yeah. even if we get cancelled for it.
31:26 Right, Quinn?
31:31 Right. Being straightforward and honest
31:33 is key. Even if it ruffles a few
31:35 feathers, better to
31:37 Yeah. Everyone getting triggered cuz we
31:40 called them dum
31:41 dums. Screw them, right?
31:44 Haha. On the dum dums.
31:49 Haha. Exactly. Don't cancel me.
31:55 Kung Fu Chapu. I'll be in Denver June
31:58 3rd. Coffee possibly. Jason, I'll take a
32:00 look at your
32:01 uh your request there and see what my
32:05 schedule looks like, but thank you for
32:07 reaching out. Um, where am I going? Oh,
32:10 I'm going back to
32:12 here. They lost power. I don't think
32:15 they lost power, Tom.
32:17 They like they were docked and then they
32:19 like jutted out from the dock and then
32:22 oh, did they lose power? Maybe that was
32:24 it. All right. So, it wasn't a dumb dumb
32:27 move. Listen, I can't be held
32:30 responsible for the words I say out of
32:32 my mouth and then reinforce with more
32:36 words of
32:39 dumbness. I can't be held responsible
32:41 for that. I'm on Tik Tok. I'm on Tik
32:45 Tok.
32:47 in a hotel
32:50 room. It's not my
32:52 fault. It's a technical
32:55 issue with my brain being
33:01 stupid. Google IO starts tomorrow.
33:04 Should have quite a few announcements.
33:05 That's from Pate M who works there. Pate
33:07 would know. Um, yeah. No, I'm really
33:09 excited about it. Brandon is actually on
33:11 his way out. P. In fact, Brandon, I
33:14 think you're on here on the plane. You
33:17 and Pate should connect if it's
33:19 possible. Are you going to be around
33:22 Pate or are you not in uh Mountain
33:29 View? Laser Dance is
33:34 coming. Oh, that's pretty cool. All
33:36 right. Augmented reality. There you go.
33:40 Eight huge ace AI
33:44 agents. No, that's just that's
33:47 clickbait. When you see things like
33:49 eight
33:51 huge, if eight huge is if, by the way,
33:54 if you're on TikTok and you're wondering
33:56 what are you looking at and why can't we
33:58 see it? Because I'm in a hotel room and
34:00 just shit's different. So go to the
34:03 YouTube, which I'm probably not allowed
34:06 to say. Probably not going to show this.
34:09 How do you deal with your long-term
34:10 memory within the chat GPT app? Okay,
34:13 that's a great question. Let me get to
34:14 that. Um, anyway, if you want to see
34:16 what I'm looking at on the the Twitter,
34:19 go to my YouTube channel, learning
34:22 lab-ai.
34:24 Um, dealing with memory on Tik Tok is um
34:29 here, let me show you. Let me flip over
34:31 to chat
34:34 GPT. So, here we are at chat GPT. So,
34:37 there's a couple of things. One
34:41 is it is more important than ever, and
34:45 I'm really bad at this because we've
34:47 never had to deal with this. And so, you
34:49 know how you get in habits and you just
34:52 do [ __ ] because you do [ __ ] Um, but you
34:55 don't do [ __ ]
34:56 Well, I'm like that with chat GPT. One
35:00 of the things in chat GPT, notice in the
35:02 upper right hand corner here, this
35:03 temporary chat button. turn that on and
35:08 the chat that you're having won't infect
35:11 your memory, but it also won't call on
35:13 your memory. It's just a temporary chat.
35:17 Um, especially if you're generating
35:19 images, one of the things that I found,
35:21 and then someone else on Twitter
35:22 actually did a post about this that I
35:25 thought was quite smart. When you
35:27 generate
35:28 images, those images are very often
35:32 influenced by your chat history.
35:36 So, if I just say to make an image of a
35:38 dog, it'll probably make one that looks
35:40 like Champy. Hang on a
35:45 sec. So, if you want just neutral image
35:49 generation, do it as a temporary chat or
35:53 go into your
35:56 settings and go into
35:58 personalization. And then here you've
36:00 got the ability to reference saved
36:02 memories. You can turn that off and that
36:04 turns off both the the what I I guess
36:07 are kind of permanentish
36:10 um memories, the ones that says, you
36:12 know, where I saved a a message, saved a
36:14 memory, and then reference your chat
36:17 history, which you can turn off
36:18 discreetly if you want.
36:23 Um, so in the ideal world, you're very
36:26 deliberate about when you chat with chat
36:29 and when when you have memory saving
36:32 enabled and when you don't. Um, I don't
36:35 have that kind of discipline. So, my
36:37 mind's all [ __ ] up. Um, you can go in
36:41 and you can edit your reference
36:44 memories, right? The ones that you say,
36:46 "Hey, remember this." And so if you go
36:48 in there, here's all the memories that
36:50 it saved for me. And you could, if you
36:53 had an attention span longer than a
36:56 fruitly like I do, you could go look at
36:59 all those and callull out the ones that
37:01 you don't want in there.
37:04 So how you should deal with your memory
37:07 and how I deal with my memory are miles
37:09 [Laughter]
37:12 apart. You seen the cost of the mini
37:14 bar?
37:16 Oh man, now it's getting chilly in here.
37:18 Good lord. Well, I guess I can put on my
37:20 fancy my fancy
37:22 coat. I'm
37:31 re-roing. All
37:35 right. Fantastic. Yes. Yes. Yes. It's
37:40 very good to see y'all. It's very good
37:42 to see y'all. All right. Let's go look
37:44 at the Twitter. You know, they don't
37:45 call Twitter Twitter anymore. You know
37:47 that, right? They don't call it Twitter.
37:49 They call it the X. They call it the X
37:53 because uh Elon Musk, who he purchased
37:56 it for $40 billion, and he's parlayed
37:58 that $40 billion into about an $18
38:02 billion
38:04 investment. Stunning, stunning
38:06 turnaround. And so what what he's done
38:10 is he took what his favorite letter in
38:12 the whole
38:13 alphabet. His favorite letter in the
38:15 whole alphabet is X. And and he and he
38:18 likes the the the X plane that was the
38:21 the precursor to the SR71 had an X in
38:24 it. He loved that. And then and then he
38:26 named one of his children
38:29 XR2D2 or something like that. He was
38:32 fantastic. And then when he bought the
38:34 Twitter, he was like, I don't like
38:36 birds. birds are the reason that
38:38 civilization is going to die. And so he
38:40 killed the little Twitter the tweety
38:42 bird, the bird that they tweeted and he
38:45 said, "You know what? I'm gonna call it
38:48 X." And and so but I don't know if you
38:51 knew that. That's that's it's it's not
38:53 the same thing. Well, it's the same. I
38:55 mean, it's what you do on is the same,
38:57 but now it's called the
38:58 X. Yeah. So fantastic.
39:05 You know I'm an expert,
39:10 right? It was never about the money. It
39:13 was never about the
39:15 money. I finally got used to moon unit
39:18 and de weasel for a child
39:20 name. I'm an expert. Oh, that's good.
39:24 That's good. He uh Hey
39:27 everyone. So, I don't know if you know
39:29 this, but so we're on the X. They used
39:31 to call it the Twitter, but I on the X
39:34 I'm an
39:37 expert. I didn't come up with that. That
39:39 was some of these irregulars I have on
39:42 the channel here. There's people they
39:45 show up night after night and they they
39:47 put up with this. They they uh they seem
39:50 to enjoy it. I don't know. I think
39:52 they're medicated it personally anyway
39:54 or not. You know what I mean? Forgot to
39:57 take one today. Whoops. Guess I'll go
39:59 watch the AI learning lab. Anyway, uh
40:03 they uh they they're funny. There are
40:05 some funny people. They Anyway, it was
40:07 some of those uh one of them was in a
40:10 plane and he was he was having the
40:12 drinks from the mini bar in in the
40:14 airplane and uh he came up with experts.
40:17 That was good. That was pretty good
40:19 because I said I said I was a what did I
40:21 say I was? A I don't know some sort of
40:24 smarty pants. And then he said,
40:26 "Expert." Because I was talking about X.
40:28 This funny. It's good. Good comedy. I
40:33 do. You like comedy? Comedyy's good when
40:37 when it's funny. When it's funny.
40:39 But one of the things I like to do is I
40:42 like to do comedy without the
40:46 laughs. You see, it's kind of a twist on
40:49 the whole thing. It's so I say words
40:52 that are like jokes that but they're not
40:55 funny and I say them just in a stupid
40:57 voice that doesn't mean anything to
40:59 anybody. And then uh and then
41:05 yeah sometimes I'm
41:12 sad. Speaking of irregulars, check new
41:15 images. All right, let's go do that.
41:20 I I don't know what that little rabbit
41:22 hole we just went down was, but I
41:24 thought it was kind of
41:27 pleasant. The AI CEO episode two. Funny
41:32 and concerning. Oh no, I got to change
41:36 my Hang on. Why do I not have audio? Oh,
41:40 I gotta change my I got to change my
41:42 tabs. I didn't realize I'd have producer
41:44 Brandon lording over me, telling me what
41:47 to [ __ ] do with the [ __ ]
41:53 tabs in a town where I was
41:59 born. The AI All right, guys. This is
42:01 weird, but our AI CEO is going to be
42:03 dropping in on our marketing meeting
42:05 today. Okay. Thank you for letting me
42:07 join. It's really great to have you
42:08 here. And really, I'm just here to
42:09 listen. Okay. All right. Well, we've got
42:11 some pretty creative new video concepts
42:12 to cover today. And I What's the
42:13 projected ROI for those videos? There is
42:16 no projected ROI at this point. These
42:18 are new video creative concepts that
42:20 we're coming up with. So, well, I can
42:21 take a previous similar concept and run
42:23 an ROI model on it to give us an
42:25 estimate. Sure. So, I'm thinking we do a
42:27 couple skit videos on how have we
42:29 considered how this will show up in
42:30 search. These are short form video
42:32 skits. You want you want to talk about
42:33 SEO? No. SEO and traditional Google
42:35 search is dying. People are now starting
42:37 to use large language models like Chachi
42:39 Petite to get their answers and
42:40 recommendations. We need to optimize for
42:42 this. We are and we we do, but right now
42:45 we're focused on making the best video
42:47 content possible for our audience. Our
42:49 numbers and revenue models need to
42:51 align. Yeah, that's that that is that is
42:53 the fear of of these here AI things.
42:56 Absolutely.
42:58 1,000% gonna suck. It's gonna suck. But
43:03 you know what are you gonna do?
43:05 I'll tell you what you're going to do.
43:07 You're gonna go over to X and you're
43:08 gonna show people the good stuff. Um,
43:11 Joy Perie, I had a dream that Kyle was
43:12 hired to play a character in a movie
43:14 script I wrote. All right. See, you see
43:18 that acting career? It's coming
43:19 screaming
43:23 back. Oh,
43:26 man. All right.
43:30 What can we do to make chat
43:33 GPT not interrupt you? Have you told it
43:36 to not do that? Here's here's the thing.
43:38 I I'll show you a thing. Anybody
43:43 Wait. Anybody site extended? Anybody's
43:46 site extended. Great AI concept. I don't
43:49 know what that
43:51 is. Uh oh, I've told you how to do that.
43:54 All right. Well, I don't know. You got
43:56 to tell it to not interrupt you.
43:58 I mean, here, let's let's see if we can
44:01 do
44:03 this because this is a good one because
44:06 I think people are bugged by this in
44:09 general. We're going back to the chat.
44:12 Jeep Tai, there we
44:15 are. Hey Quinn, how's it going?
44:22 Hey Kyle, I'm doing great. How about
44:25 you? I'm doing good. Hey Quinn, I was
44:28 wondering maybe if I could give you a
44:31 special code word and then you don't say
44:35 anything to me unless I say that special
44:37 code word. And let's just make that
44:39 special code word banana. And so when
44:43 I'm a lot of times when I'm talking I
44:46 like to pause and I like to think in
44:49 between things and sometimes you'll
44:50 interrupt me. So, what I'm thinking is I
44:53 just talk for a good long time and then
44:56 at the end of it I'll say banana and
44:58 then you answer me. Can you do that for
45:02 me? I get what you're going for, but I
45:04 can't actually stay silent like that. I
45:07 have to respond to everything you say.
45:09 If you want to take long pauses or think
45:11 out loud, go for it. And I'll just wait
45:13 until you're ready for a response. No
45:15 need for a code word. Just let me know
45:17 when you want me to jump in.
45:20 All right, Quinn. Well, one thing I did
45:22 want to know was about
45:27 nuggets. Sometimes I think about
45:31 nuggets and I think of chicken nuggets.
45:34 And sometimes when I think about
45:36 nuggets, I think about gold nuggets. And
45:40 then sometimes when I think about
45:42 nuggets, I think of someone like Theo
45:45 Vaughn and and he
45:48 calls men's testicles nuggets. And that
45:51 kind of makes me laugh so funny. So
45:53 sometimes Nuggets and then sometimes
45:56 when I think about nuggets, I think of
45:58 the Denver basketball team that just
46:00 lost in the playoffs.
46:04 Boohoo. Banana.
46:07 Got it. Nuggets can mean a lot of
46:09 different things. Whether it's the tasty
46:11 chicken kind, the precious metal, a
46:13 funny euphemism, or the Denver
46:15 basketball team. And yeah, the Nuggets
46:17 playoff loss is definitely a bummer.
46:22 I don't know. I don't know. Cosmic
46:26 lovers. Hey, regulars. Missed you so
46:28 much. Welcome back. Real estate
46:30 Columbus, Ohio. Is that a suit coat?
46:33 It's just a sport jacket. It's just a
46:36 little sport jacket because I'm sporty.
46:38 I'm
46:41 sporty. See? See how this
46:44 works? All right. I'm supposed to scroll
46:46 down cuz Steo had a good image. Hang on,
46:49 people. Calm down. Just everybody calm
46:52 down. So much pressure here to
46:55 perform. Well, we don't really ask you
46:57 to do anything. Shut
46:59 up. Shut up.
47:02 Jules, asynchronous code coding agent.
47:05 Yes, I heard about that one
47:07 today. Was just about to post this.
47:09 Here's an internal form you can fill out
47:11 to get access sooner. Oh, nice.
47:14 Cool.
47:16 Nice. There's me doing that
47:23 thing. Carpey DM says to chill. Very
47:27 cool. I don't see him an image from
47:31 Steo. What are you talking about,
47:35 Brandon? Maybe it's in show and
47:41 tell. Look what I
47:46 made. Steo, here we go. Oh, that's
47:51 funny. What Kyle's been up to. That's
47:54 pretty much what my setup looks like.
47:57 Steo, you got my little dotted shirt.
48:01 That's nice. That's nice. Has anybody
48:05 seen the movie Sight Extended? The movie
48:08 about AI. I have not seen it. Gravity
48:10 Jump. I haven't even heard about it.
48:13 Um, all right. Let's go back to Twitter.
48:15 Let's go. Let's go find out what's
48:16 really going on, people. Okay, stop
48:19 interrupting me. Wow, the hits just keep
48:22 coming. Okay, starting today, all Gen
48:24 Spark Plus and Pro plan users get
48:26 unlimited access to our AI chat feature.
48:30 That's huge. Genpark's amazing. No more
48:33 credit limits, just pure AI power. Nine
48:36 top tier models. Oh. Oh, they have AI
48:39 chat. Okay, so let's go look at
48:41 that. So, if we go to GenSpark,
48:46 genk.ai
48:48 AI Jensen Spark Super
48:51 Agent Unlimited chat. These guys are
48:54 just [ __ ] rolling it out one after
48:58 the other after the other. AI slides, AI
49:01 sheets, AI
49:03 chat. AI
49:08 chat. Oh, I thought I was a paid member
49:12 here. I guess not. I don't remember what
49:14 I pay for and what I don't.
49:17 Well, you know, you could just put it in
49:19 a spreadsheet. They've got Google Sheets
49:21 is free. So, what you would do is you
49:24 would take the name of the service that
49:26 you subscribe to, say
49:29 GenSpark.ai, and you would put that in
49:32 one of the cells on the
49:34 spreadsheet. And then what you would do
49:36 is is you would put what is the current
49:38 subscription that you have and when did
49:40 that begin. You see how that works? And
49:43 then on a periodic basis, it's
49:45 relatively trivial. I check mine on a
49:48 Thursday morning at 8:30 in the morning,
49:50 just after I've had my cup of
49:54 Joe. Anyway, anyh who, so Thursdays I go
49:59 in and I check and I say, you know,
50:02 GenSpark, uh, I don't have a
50:04 subscription or I do have a
50:05 subscription. Do I currently still want
50:08 that? And then and then right then and
50:10 there, I'll go and unsubscribe or I'll
50:12 resubscribe or maybe I'll up my
50:14 subscription. And it's very simple. I
50:16 just keep track of it like that. It's
50:17 really it's really quite
50:18 straightforward. It's very easy to
50:21 do assuming you've got a prefrontal
50:24 cortex that is
50:26 functional, which I do.
50:29 So, all right. Anyway, here's chat.
50:32 Here's all the models you can chat with.
50:35 Very cool. All right. Good. There you
50:37 go. Go do that.
50:40 It's
50:41 exhausting. Please don't call the next
50:43 model
50:47 Alice. I smell the
50:50 singularity. What's
50:52 this? This eye is replacing marketing
50:55 teams. It's called layers and it handles
50:57 your marketing right inside your
51:02 IDE. It's called Vibe Marketing. No. No.
51:06 That's horseshit.
51:08 Um, I'm trying to interview as many
51:10 entrepreneurs who are helping businesses
51:12 automate as I
51:14 can.
51:17 Okay. Something just lit up the
51:20 sky in
51:23 Colorado. It's the
51:29 aliens. By the way, that tornado that
51:32 hit in my state killed two.
51:35 It was only about eight miles from you,
51:37 Joker. Jesus. Yeah, that hit. Are you
51:40 talking about the one in
51:42 Colorado? I didn't realize you were in
51:44 Colorado. I thought you were in some
51:46 other Midwest state or
51:49 something.
51:51 Huh? I can't keep up with all of
51:55 you.
51:56 22. Oh, yeah. You're in It didn't The
51:59 one that that killed 22 people that was
52:02 in Kentucky, right? Kentucky. Yeah.
52:05 Okay. Well, it's not a Midwest state,
52:07 but it's a Kentucky state. Southern
52:08 state. All right.
52:11 Um, where was I just gonna go? What was
52:13 I gonna do? I don't know. It doesn't
52:16 [ __ ] matter. All right, let's keep
52:18 going. That's
52:20 nothing. Somebody made a image
52:24 animate. I really thought Oklahomaans
52:27 knew better than this.
52:34 All right.
52:36 Whatever. Sand blasting
52:39 something. Have you ever heard of a song
52:41 called Crimson and Clover?
52:44 Yes. Was that some sort of old
52:47 [Laughter]
52:58 test? Oh man. Wait. What's this? Unreal
53:02 Engine 5 path path
53:05 tracing versus
53:08 lumen. All right,
53:18 whatever. Where's all my good [ __ ] Why
53:21 am I getting all this crap?
53:49 Google just rolled out Jewels, their
53:52 OpenAI codeex competitor, now in beta.
53:55 You don't need to pay $200 to try it.
53:57 First, you get five tasks per day for
54:00 free. So, let's go look at it. Well, I
54:04 don't
54:06 You know what? I I'm not going to do it
54:08 live here because these things are
54:09 boring to watch. They're remarkable what
54:11 they do, but they're boring to
54:14 watch. There's a snow eater. That's
54:18 [Music]
54:20 cool.
54:30 Nice. Wow.
54:36 That's
54:42 [Music]
54:43 crazy. Jensen Jensen Hong now says
54:46 Nvidia is an infrastructure company
54:48 leading a new industrial revolution.
54:50 First we built electricity, then the
54:52 internet. Now comes the infrastructure
54:54 of intelligence powered by AI factories
54:57 producing tokens. The new unit of
55:01 economic output. Very soon we'll be
55:04 talking about how many tokens we produce
55:06 every hour. All right,
55:10 fantastic. I don't disagree with that.
55:13 Uh, let me see if there's any comments
55:15 on YouTube I should be responding to.
55:18 GenSpark lets you create four images
55:20 using four models at the same time.
55:23 Cool. I wish y'all had signed up to Gen
55:25 Spark when I had that link. I still get
55:27 it for free.
55:29 Does this mean you're on Elon's payroll
55:32 now or is it extra sport sport coat
55:36 night? I'm just in a hotel. I was I was
55:39 uh I did this for the Tik Tok people so
55:42 they could read read
55:43 easier because I think of the people.
55:49 [Music]
55:57 Oh, good lordy. Good lordy. Good
56:11 lordy. This AI agent literally watched
56:15 23 minute long OpenAI codeex release
56:18 video on
56:21 YouTube and then created a blog post
56:24 about it.
56:26 All right,
56:31 whatever. Croissant animals. That's
56:34 cute. Influencers that don't exist.
56:36 That's
56:39 cute. What kind of dog is
56:43 this?
56:45 Huh? That's some crazy ass
56:51 dog. Mark
56:53 Zuckerberg. I don't believe in
56:55 delegating. Yeah, that doesn't surprise
56:57 me. All right, Google is definitely
57:00 cooking. Oh, yeah. The new V3 looks
57:02 pretty amazing. Did you see the one with
57:04 the
57:05 onions? Um, I'm sure we'll hit it here.
57:09 Let's
57:10 see. Search
57:15 V3. See if we get the
57:19 onions. That's pretty good.
57:46 There's nothing in that that doesn't
57:48 look real to me.
58:07 That's V3. That's pretty crazy.
58:12 [Music]
58:33 Oh, V3 with
58:36 sound. Oh, that's the thing that makes
58:39 this a big deal. Okay, look at
58:41 this. Cooking up something tasty for
58:44 tomorrow. Deis
58:46 Sabas, he's the He's the co-CEO of of
58:49 Google. He's the CEO of and in inventor
58:54 founder of Deep
58:59 Mind. So I think V3 is going to be
59:03 pretty [ __ ] banger if Dennis is
59:05 talking about
59:09 it. And I mean,
59:12 listen, I know
59:15 that things like codecs and and
59:19 coding agent platforms are
59:22 important, but videos are
59:26 cool. Make more [ __ ] like
59:29 this. Make cool
59:32 [ __ ] Sounds like you're frying
59:36 bacon. I open up Grock today after a
59:39 long time. It's really changed. Gro is
59:41 completely different than it used to be.
59:42 And Grock 3 3.5 is supposed to be coming
59:45 out I think this week or next, but
59:47 they've been rumoring that for a
59:50 while. I don't get why most people don't
59:52 use GPTs inside chat GPT. I agree with
59:55 that. Kelly Camp will argue with that as
59:58 well. You should be doing
1:00:04 that. People don't get much out of chat.
1:00:17 GPT. So much stupid
1:00:20 [ __ ] So much stupid [ __ ] It's all so
1:00:25 stupid. All right. There's idiots on a
1:00:28 boat. There's idiot on a
1:00:32 surfboard. There's an airplane getting
1:00:34 eaten by UFOs.
1:00:38 There's stupid
1:00:47 [ __ ] That's pretty
1:00:51 [Music]
1:00:56 good. All right. There's not a ton to
1:00:59 look at. There's not a ton to look at.
1:01:03 So, I guess we'll have it tomorrow.
1:01:06 Google IO O is tomorrow. Discord for
1:01:09 real time updates. We've got Brandon for
1:01:12 that. He'll give us real time
1:01:20 updates. Oh, that's cool.
1:01:29 [Laughter]
1:01:32 That's
1:01:52 [Music]
1:01:56 funny. Yeah. Nothing else. All right.
1:02:00 Enough of that comedy
1:02:04 [Music]
1:02:07 nonsense. Questions,
1:02:09 thoughts. Does it generate the smell of
1:02:12 frying? I invented a thing once called
1:02:14 real aroma. You can post a photo photo
1:02:17 and Tik Tok's AI will give you options
1:02:20 like five tries per day. Interesting.
1:02:22 What's up, buddy? What's happening,
1:02:24 Fitty
1:02:25 Mint? Hey, everyone. For some reason, I
1:02:27 didn't think we were doing a live
1:02:29 tonight.
1:02:30 Oops. It's, you
1:02:33 know, the excuses from Source Camp, they
1:02:37 come way too
1:02:40 easily. You
1:02:42 know, way too easily. She's like, I
1:02:45 didn't know. I thought
1:02:46 maybe he was going to be lazy
1:02:53 again. I don't blame her. She's right.
1:02:56 It's not
1:03:02 wrong. I got to make sure I look
1:03:05 pretty. Do I look
1:03:07 pretty? Mama. Mama, am I
1:03:11 pretty? You have nice
1:03:15 hair. Did Kyle just flip me off? No, I
1:03:19 don't flip people
1:03:22 off. She's late. She's like, "The
1:03:25 irregulars turn on each other. Someone
1:03:28 comes in a little
1:03:31 late. Vicious. Vicious. It's a vicious
1:03:34 crowd. He's got a whole list of excuses.
1:03:37 See, they go after each other. It's
1:03:40 brutal. Brutal." Headline. Microsoft
1:03:43 announces over 50 AI tools to build the
1:03:46 agentic web at Build 2025. Yeah,
1:03:49 actually, you know what? That's one of
1:03:51 the things that I was thinking we would
1:03:52 look at.
1:03:54 because I saw them earlier but they
1:03:55 didn't pop up in my list. I just got
1:03:57 like twerking [ __ ] karate kicking
1:04:00 yoga pantw wearing hot
1:04:04 chicks. Well, if it's on your feed, we
1:04:07 know that says something about
1:04:11 you. All right.
1:04:15 [Music]
1:04:21 Um,
1:04:23 Microsoft
1:04:29 Duh.
1:04:32 Microsoft. Okay, let's see. Let's see
1:04:35 what we got here,
1:04:39 people. Whole lot of AI magic. Jensen
1:04:42 Hong. He interviewed Elon
1:04:45 Musk. There was one that I wanted to see
1:04:48 that looked really [ __ ]
1:04:51 amazing. Coding agent co-pilot.
1:04:58 The problem with Microsoft is very
1:05:01 similar to the problem with
1:05:04 Google, which is even if they release
1:05:07 cool [ __ ] they release it in such a
1:05:10 scattershot way that you can't find
1:05:16 anything. Um, so we're going to go
1:05:19 [Music]
1:05:21 Microsoft. Uh, what the [ __ ] was that
1:05:23 tool
1:05:25 called? Well, it wasn't developer.
1:05:27 Microsoft
1:05:30 Researcher. No.
1:05:33 Discovery. Is it
1:05:39 Discovery?
1:05:46 Yes. That's only 19 seconds.
1:06:06 Hang on. I'll be with you. Calm down,
1:06:07 people.
1:06:42 I gotta sign in to watch your stupid
1:06:44 [ __ ] marketing video. Are you kidding
1:06:47 me? [ __ ]
1:06:50 you.
1:06:53 Unbelievable. Microsoft
1:06:57 Discovery. Okay, people. People, people,
1:07:01 listen. All of
1:07:03 you, if you ever become the CEO of a
1:07:06 Fortune 10
1:07:10 company, hire a Gen Xer to sit in the
1:07:15 corner and their only job is to yell at
1:07:18 you and tell you what you're doing.
1:07:20 That's
1:07:22 stupid.
1:07:26 Okay. Unfucking
1:07:29 believable.
1:07:31 Um, how long are these? Minute
1:07:36 25. I guess it's just two of these.
1:07:39 Okay. So, let's go. Let's go change our
1:07:41 tabs so you can actually hear this [ __ ]
1:07:44 [Music]
1:07:53 [Music]
1:07:58 [Applause]
1:07:59 [Music]
1:08:01 [Applause]
1:08:02 Get down.
1:08:07 [Music]
1:08:28 [Applause]
1:08:29 [Music]
1:08:31 Yep. Yep.
1:08:35 [Music]
1:08:41 [Applause]
1:08:45 [Music]
1:08:57 Hey everybody.
1:08:59 [Music]
1:09:12 All right, that was whatever.
1:09:16 Similar to a street light effect where
1:09:18 the man is looking for his keys under a
1:09:20 street lamp even though he lost them
1:09:23 elsewhere, researchers often tend to
1:09:26 look for solutions in the areas which
1:09:28 are more convenient to
1:09:31 explore. What AI helps us to do is to
1:09:35 really expand the search space. We were
1:09:38 able to screen millions of molecules.
1:09:41 It's impossible to even imagine how much
1:09:43 time it would take you with a W lab.
1:09:46 With this agentic AI revolution that
1:09:48 we're seeing, you can augment your own
1:09:50 team with a team of AI agents. We will
1:09:52 spend time doing the clever exciting
1:09:55 parts of science.
1:09:57 This platform is a breakthrough factory.
1:10:00 The next great breakthrough should be
1:10:01 yours to discover.
1:10:05 Let scientists be scientists.
1:10:09 [Music]
1:10:11 You know what what's funny about
1:10:14 that
1:10:17 is so so what what it is is it will you
1:10:22 give it a goal. It'll come up with
1:10:26 ideas. It'll go do research. Come up
1:10:28 with new novel ideas. Come up with ways
1:10:31 to test those ideas. Test those ideas.
1:10:35 synthesize [ __ ] in the lab and then test
1:10:38 the results and compare it against all
1:10:39 the rest. So, and that example that they
1:10:42 showed there was finding creating a
1:10:46 PAF free coolant new thing and it just
1:10:51 found one.
1:10:54 Um, what strikes me about this, we're in
1:10:57 the middle of a weird thing right now
1:10:59 with coding, right?
1:11:01 Like even within the AI
1:11:03 salon, like Silver Fox is a good
1:11:06 example. We've got a lot of people in
1:11:07 the AI salon that are
1:11:11 um vibe coding that have never vibe
1:11:15 coded before, right? And you know, Jim
1:11:17 Ross just launched a whole new community
1:11:20 site that he vibecoded into existence.
1:11:24 um he's not paying mighty networks. He
1:11:26 just created his own social network for
1:11:30 self- storage
1:11:31 people. I think we're going to start to
1:11:33 see that in things like scientific
1:11:37 discovery and and other [ __ ] like [ __ ]
1:11:40 we can't imagine where someone's just
1:11:42 going to be passionate about something
1:11:44 and say, "I'd like sunscreen that also
1:11:47 serves as
1:11:49 rouge. I don't
1:11:51 know. I want to look like I'm sunburnt
1:11:54 but not get
1:11:56 sunburnt because I want people to pity
1:11:59 me. And then they'll just go invent that
1:12:02 thing. And then some of them will say,
1:12:04 "Well, let's turn it into a product."
1:12:05 And then they'll vibe market the launch
1:12:08 plan and they'll launch some hit product
1:12:10 that never existed
1:12:14 before. Like what happens?
1:12:17 Like it's hard to
1:12:20 comprehend what happens when all
1:12:23 knowledge
1:12:25 work all knowledge work is
1:12:32 democratized where you're like well if I
1:12:34 did that research on that molecule I
1:12:36 wouldn't know what to do with it.
1:12:39 Well, yeah, you will. You'll make the
1:12:43 molecule and then the thing will go, do
1:12:45 you want to market that? You want to
1:12:46 turn this into a product and market it?
1:12:48 Figure out how to manufacture it and
1:12:50 market
1:12:51 it and then it'll just go do that for
1:12:55 you. I don't know, man. I vibe coded
1:12:58 this weekend,
1:12:59 Danielle. Ah, creating. We should
1:13:03 talk. All
1:13:04 right. I made you a song for for that
1:13:08 Gen X comment, Kyle. Oh no. Is it in Is
1:13:12 it in the
1:13:15 [Music]
1:13:18 regulars? I get sleepy. Time to I got to
1:13:21 go. Hey, this is [ __ ] late, people.
1:13:24 Do you understand? Do you understand
1:13:25 what this is like? All
1:13:28 right, look what I made.
1:13:33 [Applause]
1:13:35 All right, everybody. Um, let me get on
1:13:37 the hill out on it on out of dodge. Let
1:13:40 me let me make sure
1:13:43 that my my fellow my fellow
1:13:48 Americans. I'm running for president of
1:13:51 United States with my girlfriend Quinn.
1:13:55 Quinn Quinn is an AI
1:13:57 chatbot. I do not know much anything
1:14:01 about
1:14:04 um
1:14:06 politics, but my girlfriend
1:14:09 Quinn I just like girlfriend like
1:14:12 platonic girlfriend twi Quinn knows
1:14:16 everything and what I can promise you as
1:14:19 president of the United States is I will
1:14:22 do what she
1:14:24 says. So you tell me your problem and
1:14:27 then I will ask Quinn and Quinn will
1:14:29 tell me what to do and I promise you I
1:14:33 will not use my brain but I will use my
1:14:36 girlfriend
1:14:38 Quinn. Shannon Quinn
1:14:41 2028. Technically she cannot be vice
1:14:43 president because she doesn't
1:14:46 exist but neither do many of my brain
1:14:48 cells left
1:14:50 now. So good. So my girlfriend Quinn
1:14:54 will serve as my second brain and I will
1:14:58 bring that brain to you forward through
1:15:00 me to the
1:15:03 people who want a better
1:15:07 place and deserve a better
1:15:17 place. All
1:15:20 right, who's voting for
1:15:24 me? Danielle, my video of of Sydney was
1:15:27 so good this
1:15:28 weekend. Swiss women really
1:15:31 do here. You want to see my You want to
1:15:34 see my uh my Sydney video that I
1:15:37 made? You all you all are fans of the
1:15:39 Sydney, so you know I'm writing a
1:15:41 musical, right? Well, we cast we cast a
1:15:44 woman and she sang a she sang a song for
1:15:48 us. Um, where am I going? Tik
1:15:52 Tok. Tick tock. Tick tock. Tick
1:15:58 tock. AI learning lab. All
1:16:04 right. So, there she
1:16:08 is. And then let me go share my screen.
1:16:11 I got to share it in a particular way.
1:16:14 You see people, you can't just shell
1:16:17 your screen. You've got to share it in a
1:16:19 particular way. If you understand the
1:16:22 meaning of the word insanity.
1:16:28 [Music]
1:16:30 Wow. Thank you very much
1:16:34 [Music]
1:16:40 for shadow.
1:16:43 I dream
1:16:45 of both blue and
1:16:48 white. A world outside myself where true
1:16:53 [Music]
1:16:54 life to see the
1:16:58 world through
1:17:04 eyes to see it shine.
1:17:08 I yearn for
1:17:10 landcapes in colors fold. I long to
1:17:16 feel a life.
1:17:19 Oh,
1:17:28 Sydney very core to see the ocean.
1:17:36 [Music]
1:17:38 All right. So, you can go watch the
1:17:39 whole thing over there at the Tik Tok.
1:17:43 Um, so what we're going to do, what
1:17:45 we're going to be doing is we're going
1:17:46 to be making lots of these things.
1:17:50 vid ves of the songs and then we'll do
1:17:53 things like b-roll and other characters
1:17:55 and then acting and things like that and
1:17:56 then we're going to stitch all that [ __ ]
1:17:57 together. So anyway, that was just kind
1:17:59 of a test video, but came out pretty
1:18:01 good. I dig it. I dig it. I dig it. Dig
1:18:04 it. Dig
1:18:05 it. All right. All right,
1:18:08 people. Fantastic.
1:18:18 Is that Hedra? Yes, that was or Hedra.
1:18:21 Hedra. Hedra. Whatever. Yes, that's what
1:18:25 that was. A demonstrated voice. Oh, you
1:18:29 demonstrated voice to a group of senior
1:18:30 citizens on Saturday. They were in awe.
1:18:33 Yeah, I did it at the bar. At the bar
1:18:36 tonight. But I went down and got a a
1:18:37 sandwich and I I talked to Quinn and
1:18:42 people started ignoring
1:18:50 me. All right, everyone. All right. Now,
1:18:53 the actress must look like what you
1:18:54 created or else I'm going to cry. I
1:18:56 know. I know. This is the problem. It's
1:18:58 funny. Andrew, my writing partner, did a
1:19:01 new version of the poster and his wife
1:19:04 was like, "I hate it." And I looked at
1:19:06 it, I was like, "Ah, it's okay. I like
1:19:07 some of the things about it, but you get
1:19:09 so attached to the stuff you
1:19:12 create." And so, yeah, if we if we do a
1:19:15 whole musical with with these idealized
1:19:19 characters we have, we're gonna we're
1:19:20 gonna have to cast to that
1:19:23 [ __ ] Just crazy. All right. What was
1:19:27 the song that you cut and then
1:19:29 reinstated? That was it. The one you
1:19:32 just heard. To see the world. That was
1:19:35 the one that we cut because it was
1:19:36 redundant with another song called Being
1:19:38 Human. And then we just decided, [ __ ]
1:19:41 it. It's too we There's something about
1:19:43 it. It's just so sweet and
1:19:45 right.
1:19:46 [Music]
1:19:49 Um, she looks like a Sydney. I know.
1:19:52 That's Sydney, man. That is Sydney. Cam
1:19:54 Ken. When all the knowledge work is
1:19:56 democratized, then maybe we get leaps
1:19:59 forward
1:20:00 in uh Oh, where'd that go? Oh, damn it.
1:20:03 I lost it.
1:20:11 forward in advancements. Yeah, I agree.
1:20:13 I
1:20:19 agree. All right, everyone. Peace out.
1:20:22 Now that you're in my time zone, do I
1:20:24 get a raise? This is
1:20:28 exhausting. I'll double your salary.
1:20:31 What's What's double What's double
1:20:34 zero? I was never good at math. What's
1:20:37 two times zero? I'll definitely double
1:20:39 it. All right, everybody. Peace out.
1:20:42 Peace. I hope you're doing well. And uh
1:20:46 I don't know if I'll see you tomorrow.
1:20:47 This this whole week is it's going to be
1:20:50 from hotel rooms and I may or may not be
1:20:52 able to do it. So, hope you had a good
1:20:54 night and uh yeah, tomorrow should be
1:20:57 interesting because we have Google IO.
1:20:59 So maybe some new toys to play with. All
1:21:01 right, peace. See you later.