11/4/2025 - Putting AI Presentation Makers to the Test: Gemini vs. GenSpark vs. Gamma

Description

Chapters

Transcript

0:25 Uhoh.
0:30 Oh,
0:59 Listen,
1:22 I don't
1:43 She came up to him like slow moving coke
1:48 front.
1:52 Well, his beer was warmer than a look in
1:56 her eye.
2:01 She sat on a stool and he said, "What do
2:05 you want?"
2:10 She said, "Give me a love that don't
2:13 freeze up inside."
2:19 They said I have melted all in my tit
2:29 next to you. Well, I shiver and shake.
2:37 And if I knew love, well, I don't think
2:40 I'd be here
2:46 asking myself if I've got what it
2:54 But your eyes see blue.
3:04 to stop.
3:07 Turn what's been frozen for
3:11 years
3:15 into a river of tears.
3:32 All right, I've got black bar failure
3:34 going on. Ruining everyone's life on Tik
3:37 Tok. It's just a horrible thing.
3:40 Horrible. Horrible. Horrible. All right.
3:44 Fantastic. Bob. Hey, Bob. Hey, Bob. Tell
3:47 everybody he's won. There really only
3:49 three people on Tik Tok. Yeah, there's
3:51 four people on YouTube, three people on
3:53 TikTok. Hey, Corey Sandler. We'll have a
3:56 We'll have a nice We'll have a nice
3:58 intimate evening.
4:00 I'm exhausted from the salon. Um, if you
4:02 weren't at the salon tonight,
4:04 we introduced this idea of the the AI
4:07 salon mastermind practice. This idea of
4:11 designing
4:13 an intentional daily practice using AI.
4:20 Heard a lot of tales of how people do
4:24 that in their lives with their work and
4:26 with what they do.
4:33 Vicki, sorry. I was off commenting on
4:36 another Tik Tok. Now I'm here.
4:45 Thank you. Thank you for being here.
4:48 Thank you for acknowledging that
4:50 commenting on other Tik Toks is very
4:52 threatening to me.
5:10 I did add a summary of all the AI
5:12 practices in the salon. Oh, that's so
5:14 cool, Vicki. I was thinking about that.
5:16 Um, like as they were happening, I'm
5:18 like, damn, we've got to capture these.
5:20 And I mean, I know they're being
5:22 recorded, but that's that's super cool
5:23 that you did that, Vicki. Um, yeah,
5:25 those were just that was just a really
5:27 nice
5:32 was a really powerful uh just hearing
5:35 everyone that's been doing this for a
5:37 long time like how their relationship
5:40 with AI has evolved and like everyone
5:43 that talked tonight has got this
5:47 rather sophisticated relationship with
5:50 AI and it was all different, right?
5:52 everyone had a different sort of thing
5:53 and I think I think that's the whole
5:55 point. The idea of the framework is just
5:58 to give people buckets to think in but
6:02 but in each of those buckets they're
6:04 going to design their own
6:07 what's whatever is important to them. I
6:09 it's it's I think this going to be a
6:11 really powerful thing. Put it in Look
6:13 what I made. Okay, great. That's
6:14 awesome.
6:17 Did anyone get a memory update today on
6:20 Claude? Uh, persistent memory.
6:28 I don't know. I mean, I I haven't been
6:30 to Claude today. I You know, it's funny.
6:33 I don't use Claude
6:36 nearly ever.
6:40 Ah, Claude now has memory.
6:42 Use memory. Claude can make relevant
6:44 connections across your chats.
6:49 Memory includes your entire chat
6:51 history. Here, let me share this tab. Oh
6:55 [ __ ] I wonder if I can get that popup
6:57 back.
7:00 Opus 4.1.
7:07 Let me see if it'll come back. Probably
7:08 won't.
7:11 No.
7:14 Um projects, artifacts, code code
7:23 with Claude anywhere. Run multiple
7:25 coding tasks in the cloud seamlessly.
7:33 Oh, did this pop up a new tab? It did.
7:36 Damn that thing.
7:40 $250 credit.
7:43 Trycloud code on the web on us. We've
7:47 added $250 in free usage credits.
8:02 That's that little chord thing. That's
8:05 what turned into a song two weeks ago.
8:15 Wait.
8:51 Oh man, maybe I'll just go listen to
8:55 some music.
8:57 You know what I'm going to go listen to?
8:59 I'm going to go listen to an AI
9:01 generated song that can't be creative
9:05 because it was made with AI.
9:08 And because it was made with AI, it
9:10 means that I'm a talentless,
9:13 lazy oaf.
9:21 Truth [ __ ] hurts, man. You know, they
9:25 [ __ ] found me out. They found me out.
9:30 Wait, where's my song?
9:33 No, it was after that.
9:37 How did I miss it?
9:45 Oh, maybe it's insuno. We'll go to suno.
9:58 Hello, Mr. It Michael Jackson. Country
10:01 style.
10:04 Um,
10:10 something went wrong. An error has
10:12 occurred. Rut row. Did Universal Music
10:16 by Sunno too?
10:22 All right, let's see here now. Let's
10:25 see. Published.
10:50 The ferris wheels asleep, its arms still
10:53 folded tight. We walk between the
10:56 shadows and the strings of halfway
10:58 light. The popcorn stands are dreaming
11:01 of the laughter they'll be fed and your
11:04 reflection in a puddle's turning pink
11:07 instead of red.
11:08 >> Yes.
11:13 >> You say it's strange to see it empty.
11:18 I say it's strange to feel this calm.
11:23 You smile like you know something I
11:27 don't know. I've already begun.
11:32 Before the lights come on. Before the
11:35 crowd arrives. Before we name what this
11:38 could be. Before the day survives. We
11:42 linger in the middle. Where the maybe
11:45 feels like song. Two hearts not yet
11:48 decided.
11:51 before the lights come on.
12:01 A prize bear in the windows got a rip
12:04 along the seam. You laugh and say he's
12:08 just like me, stitched up from a dream.
12:11 The smell of rain and sawdust makes the
12:14 air taste almost sweet. I count your
12:17 footsteps next to mine. They nearly
12:19 match the beat. You toss a coin, it
12:22 spins between the quiet and the breeze.
12:26 It lands on something fragile. Something
12:29 trembling between.
12:33 Before the lights come on, before the
12:36 noise begins, before the world demands,
12:39 we choose where all this story ends.
12:43 We're written in the margin where the
12:45 ink's still barely strong. Two souls
12:48 half in the promise
12:51 before the lights come on.
13:02 Cotton candy ghost the whisper of a
13:07 secret that your hand could hold but
13:10 hasn't yet tried. If I breathe too loud,
13:13 I'll break the spell we've spun.
13:18 A heartbeat before gravity decides which
13:21 way we run.
13:25 Before the lights come on, before the
13:28 sky turns gold. Before our voices learn
13:32 the lines they're scared to have told.
13:35 Let's walk this midway silence till the
13:38 dawn has found its song. And maybe we'll
13:42 decide it
13:47 before the lights come on.
14:07 That's a good song. Which one's the real
14:09 one? Um, it's an orange logo. Hang on.
14:20 Sunno su nuno.
14:25 It is
14:26 from Sunno Inc.
14:29 It's Sunno AI songs and music maker.
14:36 It's It's from Sunno Inc. The the
14:39 company is Sunno Inc. and it's it's I I
14:43 just went to the app store to just check
14:44 on the the real name of the app. Um it's
14:49 Sunno Song and Music Maker from Sunno
14:52 Inc.
14:53 I didn't get an an alert again today.
14:56 Ah, damn you, Tik Tok. Tik Tok, I'm
14:59 definitely in uh 200 view jail right
15:02 now. None none of my videos are getting
15:05 any any visibility at all, which is
15:07 frustrating, but
15:09 but I haven't been posting a lot, so
15:12 it's that's on me. That's on me for not
15:14 being a robot content creator.
15:20 Um, we had a we had a nice uh a nice
15:23 evening tonight at the AI salon. And uh
15:29 one of the things that struck me,
15:32 we've been we've been talking a lot
15:33 about this idea of a daily practice
15:35 using AI.
15:38 And I said this in the salon tonight,
15:40 but one of the things that struck me is
15:43 there's a difference between a habit and
15:46 a daily practice. So I show up here
15:49 night after night after night
15:51 and
15:54 it struck me that that
15:57 it is more like a habit than a daily
15:59 practice. And that's something I want to
16:02 work on. And I've got to really the
16:04 thing about a daily practice is it's
16:05 very intentional and it's it's rooted in
16:08 your values and it's rooted in your
16:11 goals and what you want to do and how
16:13 you want to impact people and who you
16:15 want to impact.
16:18 as opposed to a habit which is just show
16:20 up and do stuff,
16:23 which is what I do. Um, and I think the
16:27 difference could actually be quite
16:29 subtle, but I think it's powerful. And
16:32 so,
16:33 so where we're headed, what we're
16:35 launching in two weeks inside the AI
16:37 salon is um the AI salon mastermind
16:41 practice lab where if you join the
16:45 mastermind and you come to those
16:47 meetings,
16:49 you're going to get to design. We're
16:50 we're designing a framework right now
16:52 for what are the core components of
16:55 daily practice
16:58 and then together we're going to design
17:00 them. we're going to design our own
17:03 daily practice.
17:05 And and one of the things I got tonight
17:07 from people sharing their daily
17:08 practices on the salon is how
17:11 sophisticated they were. Um how personal
17:15 they were, how different they were.
17:19 Um
17:23 and it feels like important work. It it
17:26 scares me enough to make me want to not
17:29 do it,
17:32 which I think is the reason uh why I
17:35 should. What are you showing off? I
17:36 recently started an AI podcast hosted by
17:38 AI about technology and AI.
17:43 I don't know what I'm featuring tonight.
17:45 I'm I am in a I am in a liinal state
17:49 with what I talk about, what I do.
17:54 Um,
17:58 and so I don't have a good answer for
17:59 that. I don't have a good answer
18:01 tonight. I probably won't have a good
18:02 answer this week or maybe even next
18:04 week.
18:05 Um,
18:10 but it but it's something I'm figuring
18:11 out like
18:14 in in putting together a framework to
18:16 help other people design their daily
18:18 practices.
18:19 I am going to be confronted with
18:22 if I think this is such a good [ __ ]
18:24 idea, maybe I should design my own and
18:28 kind of live into that, right? And so,
18:31 yeah.
18:33 Is the GPT app store rumor true? What
18:36 What's the GPT app store rumor that
18:39 they're going to have one?
18:42 Let's go see if we can find the GPT app
18:45 store rumor. I don't I do not know.
18:47 Let's see. GPT app store.
18:52 Uh uh uh uh uh uh uh
18:59 I see no rumor. Wait,
19:03 October 14th.
19:09 No, I see no rumor. Oh, that it
19:13 launches. Yeah. No, that rumor is not
19:15 true.
19:18 That was a good one. That was a good
19:20 one. Side hustle.
19:27 Yeah. No, that's that's not going to
19:30 happen. You know who didn't have a daily
19:33 practice? The the launcher of the GPT
19:37 store.
19:43 Um,
19:45 so
19:47 a potentially interesting thing
19:50 is November 18th, um, Gemini 3 is
19:54 apparently launching.
19:58 And Gemini 2.5, I mean, Gemini 2.5 is
20:03 the one that has nano banana associated
20:06 with it. It's the one that's got V3
20:08 associated with it. Um, it's really good
20:11 at coding. It's really good at a bunch
20:13 of stuff and apparently three is much
20:15 better. But
20:18 I don't know. I don't know. Google's
20:20 weird, man. Like Google
20:23 went two years just sucking wind and
20:26 just being awful and then they just kind
20:29 of showed up six, eight months ago and
20:31 started dropping.
20:32 Like, you know what's funny? You know
20:34 what I think really kicked it off for
20:36 Google? This is weird. Is I think it was
20:40 Notebook LM. I think they hit the the
20:42 the
20:44 kind of runaway hit that notebook LM it
20:47 it was just another app in
20:49 labs.google.com google.com
20:51 and for whatever reason it caught fire
20:55 and I I just feel like because of that
20:58 maybe they just put a different level of
21:00 employee on it, a different level of
21:03 product person, a different level of
21:05 engineer
21:06 and and ever since then like the past
21:09 year they've just their their game has
21:11 been upped.
21:13 Google invented Transformers and Titans.
21:15 Yeah, but but they they invented the
21:18 transformer in 2017
21:21 and they [ __ ] buried it. I mean, they
21:24 launched it. They they gave it out to
21:26 the world, thank God. But I think that
21:29 they recognized that if we go build
21:32 something on this and it actually works,
21:34 that's going to [ __ ] up our business.
21:37 And then Open AI went and figured it
21:40 out, right? Open open AI took it and
21:43 they're like, "Wait a minute. What if we
21:44 throw all the data and as much compute
21:46 as we can afford at this thing? And they
21:49 did GPT1, GPT2, and then GPT3
21:54 in 2020, no 2020, 2020.
21:59 GPT3 launched in 2020
22:02 and it made some noise where people
22:03 like, hey, this is this is actually kind
22:06 of a thing.
22:09 And then they realized they needed even
22:12 more compute. So they cut the deal with
22:14 Microsoft that started at a billion
22:16 dollars and then quickly went to 10.
22:20 And realizing that they needed to
22:24 have something to show off to talk about
22:26 their 10 billion dollars, they whipped
22:28 up a little chat GPT
22:31 like it's like a two-eek roll out.
22:35 Um
22:37 I think because Notebook LM had its own
22:39 limits or rather lack
22:43 lack of its part of it possibly.
22:50 I believe that Google isn't far off from
22:52 Titans now though. They have over two
22:56 million
22:58 token context. You know what I want to
23:01 know is where's all the [ __ ] that they
23:02 demoed when when they rolled out Gemini
23:05 Ultra
23:07 and they had all those they had all
23:08 these demos like one of the demos was a
23:11 dynamic
23:13 like when when the large language model
23:15 responded to you, it figured out if it
23:18 needed to generate an app or a text
23:20 response.
23:22 And if it needed to generate an app, it
23:24 would dynamically generate an
23:25 interactive app.
23:27 Where is that [ __ ] thing? That was
23:29 like two years ago they demoed that.
23:32 That looked really fun. Well, they
23:33 didn't demo it. They showed a video of
23:34 it.
23:39 Oh my god. Um, what do we want to do
23:43 tonight? What do we want to do tonight?
23:46 What do we want to do? Um,
23:54 slides in Gemini. That might not be bad.
23:56 All right, let's go do that. Gemini
24:00 gemini.google.com.
24:01 So the way we're going to get to this is
24:03 we're going to go into Canvas. Go into
24:06 Canvas. So Canvas is just like Canvas
24:09 and chat GPT.
24:11 New video generation just got better
24:13 with VO3.1. Great. Awesome.
24:17 Um, so now I'm going to say, um,
24:22 make me a presentation
24:26 about,
24:29 um,
24:30 1950s
24:33 train
24:35 based carnivals.
24:38 All right, let's see what we get. Am I
24:41 sharing my screen right? I am.
24:46 that you can put your own content in
24:48 Notebook LM is huge. Yeah, Notebook LM
24:51 is just really something like the fact
24:53 that you can put I think it's now 300
24:56 documents in it and just start
24:58 interacting with them. Stephen Johnson
25:00 is a writer, liberal arts guy on
25:02 notebook. Yeah, I know. I know Stephen
25:04 Johnson really well. I was doing Urban
25:06 Desires and he was doing um
25:10 he was doing Nerve Nerve Mag. So, we
25:12 both had art and culture magazines and
25:14 we had a sex section and he had a Nerve
25:17 was a sex magazine. Um, or sexuality
25:21 magazine, whatever it was. And, uh, so I
25:24 know him really well. He's he's a he's a
25:27 wicked smart dude. He's a wicked smart
25:29 dude. Good writer, wicked smart dude. He
25:32 he designed Notebook LM. It was the
25:34 thing he always wanted was the ability
25:36 to just dump a p bunch of unstructured
25:39 data into a notebook and be able to
25:41 interact with it. So he he he definitely
25:44 had a vision for that thing. You need to
25:47 specify the output at least personality
25:50 when using Gemini Pro. Okay. Creating
25:53 slides. 1950s carnival trains. I I
25:57 actually wrote that not knowing if they
25:59 were a thing.
26:07 The difference between a practice and a
26:09 habit. A habit is just make a
26:12 presentation of something nonsensical.
26:17 Actually, when I was in
26:20 When did I see Stephen Johnson? I was in
26:22 San Francisco.
26:27 No, I was in DC. Last time I was in DC,
26:30 six months ago, it was a Google event
26:33 and he was there and we we caught up. It
26:34 was nice to catch up.
26:37 You're using Flash. I'm using Yes. 2.5
26:41 Flash,
26:43 which is good.
26:48 I mean, I could use Pro in reasoning,
26:50 but I'm just making a slide show about
26:53 carnivals.
26:55 Carnival barkers
26:58 rolling wonder. All right, let's How do
27:00 we make this present? Is that the
27:02 present button?
27:07 Export to slides.
27:16 I only use Flash when I hit my limits.
27:20 It's better at everything. Pro is better
27:23 at everything. I guess it makes sense,
27:27 but it's generating an HTML app, not a
27:30 presentation right now. No, it's
27:31 generating a presentation.
27:34 Unless it wrote it in HTML, which it
27:37 probably did.
27:40 Did this export to slides?
27:47 Come on, Gemini. What are you doing?
27:54 Did I crash?
28:01 Open slides.
28:15 Did
28:24 Gemini still embrace the jank? I think
28:27 it is. This is This is pretty bad.
28:34 The iconic ferris wheel, the classic
28:36 carousel with giant women on tiny
28:38 horses, the tilt whirl. So, it's
28:42 definitely generating the images.
28:47 The performer's life was tough.
28:54 60 cars on a major show train. A single
28:58 show train was a massive undertaking.
29:01 Like the James Eye Straits show or World
29:03 of Mirthth moved 40 to 80 cars.
29:09 These trains carried hundreds of staff,
29:11 families, and all the equipment
29:14 needed to build a small enter small city
29:16 of entertainment overnight. All right,
29:18 so you know what we're going to do? No,
29:21 Kyle, tell me questions. Thank you for
29:24 your attention. We're going to go to Gen
29:26 Spark and do the same thing. Let me get
29:29 the exact prompt I use.
29:31 I used
29:34 make a presentation about 1950s
29:36 trainbased carnivals.
29:38 I don't think you are giving Gemini
29:43 enough credit. Tough.
29:49 I like picking on Google.
29:55 But let's let's go to Gemini. We're
29:57 gonna I'm doing apples for apples. We're
29:59 doing the exact same prompt in Gemini
30:02 or not Gemini. Uh what's it called?
30:04 Genpark. Genpark.
30:07 Tabs for tabs. You're correct, sir. All
30:10 right. GenSpark Super Agent. We're going
30:12 to do slides. So, we're going to go to
30:14 into their slide maker and I'm going to
30:17 do the exact same thing. Boom. All
30:20 right.
30:21 Sign in. Oh crap. It's gonna Please log
30:24 in to use this feature.
30:33 D.
30:55 I'll create a fascinating presentation
30:57 about 1950s trainbased carnivals for
31:00 you. Let me start by researching. See, I
31:03 like this better already. Like, like
31:05 Gemini just did all this in the
31:06 background. Initial research.
31:09 Excellent. I've gathered some
31:11 fascinating initial research on this
31:13 unique American slice of Americana. Let
31:17 me deep dive deeper into the most
31:19 valuable sources to craft a
31:21 comprehensive presentation about these
31:23 rolling wonders of the 1950s. I didn't
31:27 even know that these were a thing, but
31:29 they were rolling wonders.
31:33 Wish my wife didn't work nights at home.
31:35 It'd be easier to do these lives.
31:39 Listen, I'm really sorry I'm [ __ ] up
31:41 everyone's family life. I know these
31:44 lives are at an inconvenient time and
31:47 we're doing festivus in between the
31:49 holidays. By the way, if you don't know
31:50 about Festivus,
31:52 um Festivus happens December 26th and
31:55 27th. It's a Friday and Saturday, 9:00
31:58 a.m. to 9:00 p.m. for free.
32:02 24 hours over two days of
32:06 AI practitioners basically donating
32:09 their time. It's a remarkable event. So,
32:12 put that in your [ __ ] calendar and
32:14 tell your family you can't be there
32:16 unless you love them, broken as they
32:19 are.
32:24 No, I mean, I don't also go live and
32:26 duet. Oh, if she wasn't working. Oh,
32:28 yeah, that'd be cool. Um,
32:31 let's see.
32:34 I like to time them. It's usually after
32:36 my kids are in bed.
32:38 Oh, man.
32:41 Okay, let's see what's this doing.
32:46 Oh, look. It made images. Oh, see these
32:48 are cool images. Oh, it researched some
32:50 images, too. These aren't just
32:51 generated.
32:53 Yeah. See? You see? See? Yeah. That's
32:57 what I'm talking about. Oh, yeah.
33:02 Initialize slide project.
33:07 Okay. It's writing the slides.
33:10 I I hate to break it to you, Brandon,
33:13 like the slides are already the slides
33:15 in text form are already way better.
33:17 They're like researched.
33:21 Um, but you know, I know I don't give
33:24 Google a shot cuz I cuz
33:27 here, you want to know why I'm pissed
33:28 off at Google?
33:30 You want to know why? You want to know
33:32 why?
33:34 When they first started, they said,
33:36 "First, do no evil." And then one of the
33:39 first things they did was drop that
33:43 and then did evil
33:51 gamma slides in comparison. We can go to
33:53 gamma slides and do that too. Fine. All
33:56 right. What's this doing? Is this
33:58 perfect? I've generated a beautiful
34:00 vintage cover image. Okay, that's fine.
34:03 But
34:05 waiting for generation. Okay. So now
34:06 it's doing All I'm saying is that Gemini
34:12 is was was done faster.
34:19 Brandon,
34:22 let's go do what is it? Gamma. G MA. I
34:26 don't think Gamma does um research
34:28 though, does it? We'll find out, won't
34:31 we?
34:33 Transform your ideas into stunning
34:35 presentations. Make me a Okay, here we
34:37 go.
34:39 Going. Beautiful. Oh, now I got to pick
34:43 a [ __ ] slide design. Seriously,
34:48 I don't want to have to work.
34:51 AI is supposed to do this for me.
34:56 We don't We don't need no stinking
34:58 design templates.
35:00 Hey. Hey. Hey. Hey. Hey, man. Hey. Hey.
35:07 Um, none of these look like 1950s. We'll
35:10 we'll use that one. Future ports. Oh,
35:13 wait. You're not seeing this Tik Tok
35:15 pin.
35:17 Um,
35:19 Mary Mary. Oh my. I'm passing notes in
35:23 the back row of class. I better pay
35:25 attention now. Are you all chatting with
35:27 each other and not paying attention to
35:29 me? You're supposed to pay attention to
35:32 every single word I say. They're all
35:36 important. Every one of them. Deeply
35:38 important.
35:41 I thought you knew that.
35:44 All right, let's go back to Gen Spark.
35:46 So, Gen Spark is now writing HTML
35:48 slides.
35:56 I tried to use AI to build a slideshow
35:58 for my wife, but her work for Gemini
36:01 failed me. A Oh, for her work, but
36:04 Gemini failed me. Corey sent Kyle, I'm
36:07 asking, what does Gen Spark do? Okay, so
36:09 if you haven't played with GenSpark,
36:11 Corey, it's it's definitely worth a
36:13 shot. Um,
36:16 wait, did this
36:20 Oh, waiting for generation.
36:25 It's still building slides. Okay,
36:27 whatever. Um, let me do a new let me do
36:30 a new thing.
36:34 Um,
36:38 okay. So, this is genark.
36:41 So, Corey, genspark is very similar to
36:44 Manis
36:45 in that it is an agentic.
36:48 it it's an agent that will go off and
36:51 just do [ __ ] for you. What's cool about
36:54 Gen There's a couple of cool things
36:56 about GenSpark. One is
36:59 just the design aesthetic. Like even
37:01 just look at the like the card display.
37:04 It's in this arc and then when you
37:05 scroll down it fans out.
37:09 Um and they have all these different
37:10 things. They've got a super agent,
37:14 a super agent that will do everything.
37:17 Um, they've got a slide generator, a
37:20 spreadsheet generator, a a Google Doc
37:23 generator, um, a code generator, a
37:27 design generator. There's a thing called
37:29 Clip Genius, which is I can point Clip
37:33 Genius at a YouTube link and tell it to
37:37 analyze the video and cut me a 4-minute
37:40 super cut and put subtitles on it and
37:43 and chapter cards in between the
37:45 sections and it'll just go do that. Um,
37:50 so you can choose individual tools, but
37:52 you can also just say to the super
37:54 agent, go to this YouTube URL and make
37:57 me some [ __ ] and it'll figure out which
37:58 of the tools to use. It's pretty slick.
38:01 I think I have the GenSpark app. Oh. Oh
38:03 my. I have Manis as well. I I think both
38:06 of them
38:09 like Manis was the one that it was the
38:12 first of these agent apps to come out
38:14 that was really good, but like
38:18 like all of them right now, it was
38:20 really janky and the design [ __ ] that it
38:22 did was bad. GenSpark was the first one
38:24 that came out that had decent design and
38:26 its outputs. Um, but then GenSpark just
38:29 keeps releasing new tools, tool after
38:31 tool after tool after tool. They're
38:33 taking this idea of the agent and
38:35 they're just making all these custom
38:36 little basically these are like custom
38:38 GPTs for an autonomous agent. Oh yeah,
38:42 it's also Chinese. So anything you put
38:46 in there, assume that it's your data is
38:49 gone or or going to be used somewhere
38:51 else. Um, but it's pretty slick. So, if
38:54 you're doing something like I just did,
38:56 like,
38:58 you know, make me something about a
39:00 carnival in the 50s. Uh, it'll do that.
39:04 Okay. Are we done? Did it finish?
39:07 Why is that not uh What slide is that?
39:10 Five of 10.
39:13 So, these the these aren't actually
39:15 better than Gemini, like design-wise.
39:18 Anyway, let's see. View and export.
39:21 Advanced edit. How do I
39:25 view View and export? View and new
39:27 window.
39:29 Tab. Yes. Thank you.
39:36 The romance, spectacle, and logistics of
39:39 America's Rolling Midway.
39:44 A city that arrived overnight. Prestige
39:47 on rails. consistency and reach uniquely
39:50 American. Three car types like why do
39:53 these not have images
39:56 and like that one's missing an image
40:00 and that one was tall.
40:03 The hell this was bad.
40:06 Never mind. I thought this was going to
40:08 be good. It wasn't. Um let's see. Show
40:11 this tab instead.
40:14 Generate PowerPoint. Let's see what
40:16 gamma does.
40:18 Oh, come on.
40:23 This is annoying.
40:25 This is annoying. This is annoying.
40:36 It's kind of cool watching it build a
40:38 little slide deck for us.
40:44 Is it going to switch out the images or
40:46 is it going to use these crappy stock
40:48 images?
40:52 I created EDM music worthy of a
40:54 nightclub. I just had my double
40:56 Mallister moment.
40:58 Gravity jump. Oh. Oh, you just
41:00 downloaded. If you haven't played with
41:02 So, I'll tell you. You want to know
41:04 something that'll blow your mind? That
41:05 song I played earlier, the the really
41:07 pretty, you know, Twinky one. Um
41:11 the that started with a little guitar
41:13 riff and so you can record
41:17 yourself humming or singing or playing
41:22 into it and then have it write a song
41:24 from that. It's pretty astounding.
41:27 Claude on Canva chat GBT
41:30 to Canva. Oh, Claude to Canva. Chat GBT
41:33 to Canva. I don't know what you're
41:35 talking about, Chef Kelly, but that
41:37 sounds cool. Chef Kelly vibe coded up a
41:40 really cool app. I I I'm one of the
41:44 things I'm really excited about is
41:45 people have deep expertise and deep
41:47 passion for something.
41:51 Starting to realize all the different
41:53 areas they can aim that energy and that
41:56 expertise
41:57 where they couldn't have before without
41:59 a lot of resources is is a really
42:01 remarkable thing to watch.
42:04 Um, this thing says it's 97% done.
42:24 I think they all did poorly with a with
42:28 a shitty prompt. True. I gave a shitty
42:30 prompt. I'll give you that.
42:38 I will give you that.
42:41 I I did a solid nothing
42:45 to give these things a shot to be
42:47 successful.
42:50 Grock now remembers. Tell it something
42:52 in this post. Then when you talk to it
42:55 again, it will remember what you talked
42:57 about.
42:59 I Oh, wait. Let me share my tab.
43:02 I think that
43:10 puppies
43:12 are more tender than kittens.
43:23 Okay. So, now
43:27 I should be able to go to Grock
43:30 and say, um,
43:33 I recently
43:35 commented
43:36 on Robert
43:39 on a
43:42 at Scoilizer.
43:45 No, I'm gonna do Robert Scoble on a
43:47 Robert
43:49 Scoble
43:51 post.
43:55 What was I talking about?
44:02 Thinking
44:07 Robert Scoble is a well-known blogger.
44:14 No, it didn't find it.
44:17 I guess I guess you have to give it time
44:19 to get ingested.
44:29 Yeah. Grocon X.
44:54 It remembered something I said to Robert
44:56 Scoble. Whatever. All right. Whatever.
44:59 Let's just go look at what's going on on
45:02 the Twitter.
45:04 Gen Spark. Gen Spark done yet? You mean
45:06 Gamma?
45:08 No, Gamma's not done yet.
45:15 Suddenly Gemini is not looking so bad.
45:18 Brandon, I swear to God, Brandon is
45:20 being paid by Google. He's always like,
45:22 you know, Android phones are not really
45:25 as horrible as you think they are. You
45:27 should try Android sometime. Like, no,
45:30 no,
45:32 no.
45:34 Bad producer.
45:37 But we got Sora today. Yeah. By the way,
45:40 if you want to make a a uh a what you
45:43 call it, an avatar of Brandon, you can
45:46 go to Sora now. He's in there.
45:50 Um
45:55 I'm good. Just tired. Thanks for asking.
45:57 I wonder if Grock remembers you from
45:59 another PC. Yeah, probably because it's
46:01 just it goes into its big ass database.
46:04 All right. Um I'm going to get out of
46:05 here. I'm I am exhausted from tonight.
46:08 And Kelly Camp coming in here going, I'm
46:10 exhausted. I'm You know what, Kelly, I'm
46:12 with you.
46:16 Let's get the hell out of Dodge.
46:19 All right, let me share this tab. Let me
46:21 see if there's Let me see if there's any
46:23 inspiration in
46:25 in Twitter tonight. Wait, show skills.
46:28 Google since you've been dunking on them
46:31 all night. All right, fine.
46:34 Fine, Brandon. skills. If this doesn't
46:36 blow my socks off,
46:39 then we we're going to have an issue.
46:41 Here's the problem. If it does blow my
46:42 socks off, then we're going to be here
46:44 for another 45 minutes. Build AI skills
46:47 for tomorrow. Today, what do you want to
46:48 learn? I want to learn um how do I make
46:54 a movie
46:58 with VO
47:03 data engineer?
47:05 What?
47:11 It's a library of free training. Oh.
47:16 Oh. So I can't just ask it for [ __ ]
47:19 Wait, so this is like
47:23 Oh god, no. Okay, I am
47:28 hang on. If if you if you search for
47:30 generative AI, they've got a course on
47:32 there that's free where they're showing
47:34 you how to do generative AI and not the
47:38 way that we kind of play around with it.
47:41 They actually teach you like a ple
47:43 course but make it understandable for
47:46 the general public and it's all free and
47:47 that you can get certified and you know
47:49 >> you need to actually learn it. This is
47:52 this is exhausting to me.
47:56 >> Introduction to Gemini Pro
47:59 Generative AI with Vert.ex AI. But so so
48:02 let's go to one of these.
48:06 I got to sign up.
48:06 >> It might take you a while to on board.
48:09 >> Oh, really?
48:12 Oh, yeah. Because you're taking an
48:13 actual course. Oh, yeah. Okay. All
48:15 right. So, these are full-on courses.
48:17 That's kind of cool. That's kind of
48:18 cool. I just want I just want
48:23 I'm I'm lazy. I want to not have to
48:26 work. I'm tired of learning. I just want
48:29 it piped into my head. I at this point,
48:31 I'll do a Neurolink implant. Just tell
48:34 me what the latest AI tools are and just
48:37 show me how to use them all. I'll just
48:39 You can just run it right into my right
48:40 into my synapses. I'm good.
48:44 All right. Um, let me go back to X and
48:47 see if there's anything there. X.com.
48:50 I'm with Kyle. If it starts to feel like
48:53 work, I'm no longer interested. Yeah,
48:55 exactly. Neuralink's right around the
48:56 corner. I mean, the Neuralink stuff is
48:59 actually quite impressive. The like the
49:01 nine or 10 people that have it so far is
49:03 quite impressive.
49:05 Doing the work of a plumber will not be
49:07 difficult for AI. Once you have a world
49:10 model,
49:11 there's a brain and a plastic skull.
49:18 Google is now updating its maps logo.
49:21 Oh, thank God. Maybe they're going away
49:23 from flat design. I think the worst
49:25 thing that happened to the internet is
49:28 the the minute Apple gave up on
49:32 skuorphic design things that looked like
49:34 real buttons
49:36 and Google came in with their [ __ ]
49:39 flat design aesthetic that was like,
49:42 "Hey, I got an idea. Let's take anything
49:44 that's aesthetically interesting and
49:46 remove that from all interfaces. Let's
49:49 make it so that it's so flat you can't
49:52 even find a button if you're older than
49:55 seven.
49:58 [ __ ] I hated that. I still hate it. So,
50:02 the fact that they're going to
50:04 Appleesque gradients. Fine. It's about
50:07 time. You're 20 20 years too late. You
50:10 already [ __ ] up the whole internet.
50:13 You wonder why I'm bitter about Google?
50:22 Haha, legit flat UI. Um, what about
50:26 Apple's glass UI? I, you know, I just
50:29 updated to the to the like release
50:32 candidate for um for iOS 26 or whatever
50:36 it is. I don't hate it. I don't hate it.
50:39 Like the fact that it's got a little bit
50:41 of depth to it. Um, I haven't sort of
50:44 flipped it into pure glass mode. But
50:46 it's kind of weird. Yeah. But it's like
50:48 I it like it's at least an elegant
50:50 attempt to do something
50:54 um aesthetic again, like I said, and and
50:56 some of their icons are going back to a
50:58 bit of that like a like I would call it
51:01 like post skuorphic,
51:04 right? like it's they're they're not
51:06 going full-on, you know, 3D cartoon
51:09 models, but they're but they're, you
51:11 know, they're starting to put some
51:12 dimensionality back in the interface,
51:15 which great. Love it. Fantastic.
51:18 Fantastic. Bob,
51:21 um Google's going to launch um TPUs into
51:24 space. So that company this week when I
51:27 was at TED AI um there was a company
51:29 there that was launching uh H100s or an
51:32 H100 into space this week
51:36 or no last week. Last week they did it.
51:38 Um and Google just said they're going to
51:40 do that shortly. So GPUs in space.
51:47 Um it actually makes economic sense to
51:49 do it which is kind of cool.
51:52 Um,
52:02 all right. Yeah, nothing nothing super
52:04 interesting.
52:06 Nano Banana 2 is incoming.
52:10 Breaking. Google is preparing Nano
52:11 Banana 2 Gem Pix 2 for an upcoming
52:15 release.
52:18 A new announcement has been recently
52:19 added to the Gemini website, which means
52:22 we should expect a release within the
52:24 following weeks. All right. Well, that's
52:27 not really a scoop, but that's cool.
52:31 VO3.1 limit of 8 seconds through flow
52:34 and Gemini is a disadvantage compared to
52:36 Sora 2.
52:40 Someone's back.
52:44 All right, there's nothing. All right,
52:45 we're going to go. You all can go to
52:46 sleep now.
52:49 I'm going to sleep now. We're done. Um,
52:52 okay. Today,
52:56 today is Tuesday. So, tomorrow we've got
52:58 the AI Readiness Project podcast. So,
53:01 Ann Murphy and I are doing a wrapup of
53:04 what we're calling season zero. So,
53:06 season zero was we just threw the AI
53:09 readiness project up there as a podcast.
53:12 We just threw it out. We didn't really
53:13 think about it. We did, I think, 30
53:17 episodes in the first season. So, we're
53:19 going to be kicking off season 1 coming
53:21 up. So, tomorrow it's at 400 PM Mountain
53:25 time. Um, aire readiness project.com
53:30 and uh we're going to be streaming on
53:34 YouTube
53:35 and we're going to be just talking about
53:38 the our favorite moments from season
53:40 zero. So, you should come check that
53:42 out. All right. Beautiful. And then I'll
53:44 be back here tomorrow night at 8:00 p.m.
53:46 Mountain time. Same normal time tomorrow
53:50 night. Sorry we were a little late
53:51 tonight. I'm a little low energy, but
53:53 you know, you get what you get.
53:57 A shoe. No, I just got here. Sorry.
54:00 Sorry. It's been a grand day. Yeah, it
54:02 was a good day. Thank you, Kyle. Thanks,
54:04 mods. Yeah, thank you for thanking the
54:06 mods. I didn't get a nap today, so I can
54:08 nap now. Yeah, good. Let's all go nap
54:11 and then we'll come back tomorrow with
54:12 uh with freshness
54:15 and and some moisture.
54:18 Peace. I'll see you tomorrow.
54:22 All