Influencer Marketing Gets an AI Makeover

[00:00:00] [SPEAKER_02]: This is AI Inside Episode 34, recorded Wednesday, September 11th, 2024.

[00:00:06] [SPEAKER_02]: Influencer Marketing Gets an AI Makeover.

[00:00:11] [SPEAKER_02]: This episode of AI Inside is made possible by our wonderful patrons

[00:00:14] [SPEAKER_02]: at patreon.com slash AI Inside Show.

[00:00:18] [SPEAKER_02]: If you like what you hear, head on over and support us directly.

[00:00:20] [SPEAKER_02]: And thank you for making independent podcasting possible.

[00:00:31] [SPEAKER_02]: What's going on, everybody?

[00:00:33] [SPEAKER_02]: Welcome to another episode of AI Inside, the show where we take a look

[00:00:36] [SPEAKER_02]: at the AI that is layered like lasagna throughout everything in technology,

[00:00:41] [SPEAKER_02]: even an apple lasagna as we're going to talk today, because this is, after all,

[00:00:45] [SPEAKER_02]: Apple Week in the world of technology.

[00:00:47] [SPEAKER_02]: Joining me as always, my co-host, Jeff Jarvis.

[00:00:50] [SPEAKER_02]: How you doing, Jeff?

[00:00:51] [SPEAKER_02]: So if it's Apple, is it a strudel then?

[00:00:55] [SPEAKER_02]: It's a strudel, yes.

[00:00:57] [SPEAKER_02]: This week, it's an AI strudel.

[00:00:59] [SPEAKER_02]: Exactly. I love where you went with that.

[00:01:01] [SPEAKER_02]: That didn't occur to me on the fly, but I think that's the perfect analogy.

[00:01:05] [SPEAKER_02]: And, you know, both of us, not Apple users.

[00:01:09] [SPEAKER_02]: So, you know, I bet there's a lot of listeners and viewers that actually have

[00:01:14] [SPEAKER_02]: iOS devices, but because we're the weird ones.

[00:01:17] [SPEAKER_02]: Yeah.

[00:01:18] [SPEAKER_02]: Yeah.

[00:01:19] [SPEAKER_04]: Even though the majority in the world in this world were weird.

[00:01:23] [SPEAKER_02]: Yeah, totally.

[00:01:25] [SPEAKER_02]: Exactly.

[00:01:26] [SPEAKER_02]: We got a lot to talk about today.

[00:01:28] [SPEAKER_02]: It was an interesting AI Newsweek.

[00:01:30] [SPEAKER_02]: Not a whole lot of like groundbreaking announcements and pieces of news.

[00:01:34] [SPEAKER_02]: But we did find some fun stuff to talk about before we get there.

[00:01:37] [SPEAKER_02]: Big thank you to those of you who support us on a weekly and monthly basis via

[00:01:43] [SPEAKER_02]: Patreon, patreon.com slash.

[00:01:45] [SPEAKER_02]: That's the wrong lower third.

[00:01:47] [SPEAKER_02]: Let me put up the right one here.

[00:01:49] [SPEAKER_02]: Where is it?

[00:01:50] [SPEAKER_02]: There it is.

[00:01:51] [SPEAKER_02]: Before your very eyes.

[00:01:52] [SPEAKER_02]: Patreon.com slash AI Inside Show.

[00:01:55] [SPEAKER_02]: Philip is one of our supporters from the very beginning.

[00:01:58] [SPEAKER_02]: And, Philip, we appreciate you being on board with us on the Patreon so, so very

[00:02:03] [SPEAKER_02]: much.

[00:02:04] [SPEAKER_02]: Also, if you are watching live because each week our live numbers continue to

[00:02:08] [SPEAKER_02]: grow, I encourage you to subscribe to the show.

[00:02:12] [SPEAKER_02]: AI Inside.

[00:02:13] [SPEAKER_02]: Show.

[00:02:13] [SPEAKER_02]: You can go to our website.

[00:02:15] [SPEAKER_02]: You can subscribe to the podcast, put it into your pod role and whatever your

[00:02:19] [SPEAKER_02]: podcatcher is.

[00:02:20] [SPEAKER_02]: Is it Apple Podcasts?

[00:02:21] [SPEAKER_02]: Is it Pocket Cast?

[00:02:22] [SPEAKER_02]: Whatever it is, we don't care.

[00:02:24] [SPEAKER_02]: It doesn't matter.

[00:02:25] [SPEAKER_02]: As long as you subscribe, then you won't miss it.

[00:02:27] [SPEAKER_02]: Even if you happen to miss the live recording that we do each and every week.

[00:02:31] [SPEAKER_02]: Uh, and we thank you for being here.

[00:02:32] [SPEAKER_02]: Thank you.

[00:02:33] [SPEAKER_02]: All right.

[00:02:35] [SPEAKER_02]: So let's get into some of the news then, because we got some stuff to talk about.

[00:02:39] [SPEAKER_02]: And I think, you know, top of the heap has to be Apple this time around just

[00:02:44] [SPEAKER_02]: because Apple had its, you know, biggest product announcement of the year.

[00:02:49] [SPEAKER_02]: Does it around this time of year?

[00:02:50] [SPEAKER_02]: Every single year, we've got new iPhones, got new Apple watches.

[00:02:55] [SPEAKER_02]: And, um, it really, I, you know, I will say for the first time ever, I did live

[00:03:00] [SPEAKER_02]: coverage of the Apple event, which I, you know, I don't, I'm an analyst.

[00:03:06] [SPEAKER_02]: You know, I figured it was probably, probably the right thing to do.

[00:03:10] [SPEAKER_02]: It's not like Apple is an insignificant player in the, in the technology industry.

[00:03:14] [SPEAKER_02]: So, um, I wanted to kind of see it more.

[00:03:17] [SPEAKER_02]: Yeah.

[00:03:18] [SPEAKER_02]: More Android perspective on the, on the iOS and outside.

[00:03:23] [SPEAKER_04]: Apart from just generic Apple hand candy wonder.

[00:03:27] [SPEAKER_04]: Yeah.

[00:03:28] [SPEAKER_04]: Did you see anything?

[00:03:29] [SPEAKER_04]: This is AI inside episode 34, Wednesday, September 11th, 2020 that got me

[00:03:34] [SPEAKER_02]: influencer marketing gets an AI makeover.

[00:03:37] [SPEAKER_02]: And that was sort of AI inside is made possible by our wonderful patrons at

[00:03:42] [SPEAKER_02]: patreon.com slash g.

[00:03:43] [SPEAKER_02]: Guitars hanging on the lawn.

[00:03:44] [SPEAKER_02]: I'm a musician and they, this episode of AI inside is made possible by our

[00:03:49] [SPEAKER_02]: wonderful patrons at patreon.com slash AI inside show.

[00:03:53] [SPEAKER_02]: If you like what you hear head on over.

[00:03:54] [SPEAKER_02]: It's very niche, right?

[00:03:56] [SPEAKER_02]: So it's not something that everybody's going to benefit from necessarily.

[00:04:00] [SPEAKER_02]: But if I have an idea for a song and I pulled out my guitar and I use the Apple

[00:04:04] [SPEAKER_02]: notes, you know, as many musicians do, they rely on Apple notes to compile their

[00:04:09] [SPEAKER_02]: ideas on the fly.

[00:04:11] [SPEAKER_02]: This new feature allows you to basically do a multi-track recording on top of that

[00:04:16] [SPEAKER_02]: original recording and have them be separate recordings.

[00:04:19] [SPEAKER_02]: So essentially what that means for musicians is I can record my guitar part when I'm

[00:04:23] [SPEAKER_02]: inspired to do so.

[00:04:24] [SPEAKER_02]: And then later I'm in the kitchen, you know, cooking my pasta for lunch and I'm

[00:04:27] [SPEAKER_02]: like, Oh, this would be a cool vocal line over the top of it.

[00:04:30] [SPEAKER_02]: I can open it up.

[00:04:31] [SPEAKER_02]: I can hit record.

[00:04:32] [SPEAKER_02]: It plays the audio out into the kitchen.

[00:04:34] [SPEAKER_02]: I sing my part and the recording, I'm not sure exactly how Apple is doing it.

[00:04:40] [SPEAKER_02]: They're probably doing some sort of polarity reversal to remove the guitar

[00:04:44] [SPEAKER_02]: that's playing into the room from the recordings that is just your voice, but it

[00:04:49] [SPEAKER_02]: allows you to layer your voice on top of another track so you can kind of capture

[00:04:55] [SPEAKER_02]: more cohesive ideas on the fly.

[00:04:58] [SPEAKER_02]: And I just thought, you know, from, from a musician standpoint, I was like, Oh, I

[00:05:02] [SPEAKER_02]: wish I had that.

[00:05:03] [SPEAKER_02]: Is there an app on Android that does this?

[00:05:06] [SPEAKER_02]: Cause I would love that.

[00:05:07] [SPEAKER_04]: And then you could replace, so you could do a humming track and then replace that

[00:05:11] [SPEAKER_04]: with some with, with instrumental or something else later.

[00:05:14] [SPEAKER_02]: Sure.

[00:05:15] [SPEAKER_02]: And I mean, what they were saying also is like, you could take that, you know, that

[00:05:19] [SPEAKER_02]: isolated recording and, and use it in your project.

[00:05:22] [SPEAKER_02]: And I'm a little hesitant on that.

[00:05:24] [SPEAKER_02]: Like, okay, you know, is it really going to be that great?

[00:05:27] [SPEAKER_02]: Um, so great that you could actually use it in a project and it would sound good,

[00:05:30] [SPEAKER_02]: but still nonetheless, as far as like a prototyping tool for, for music, I think

[00:05:36] [SPEAKER_02]: that's, that's really neat.

[00:05:37] [SPEAKER_02]: There was another feature that they showed off with the, um, the camera, the

[00:05:42] [SPEAKER_02]: video camera, and again, with sound where if you're recording something in like a

[00:05:48] [SPEAKER_02]: noisy environment, it has these different profiles of sound processing.

[00:05:53] [SPEAKER_02]: So, and for the, so the example they gave were there, they were, it was almost like

[00:05:57] [SPEAKER_02]: a movie was being shot in this cafeteria and these two guys were sitting at the

[00:06:01] [SPEAKER_02]: counter talking to each other and the audio recording did have the noise of the

[00:06:05] [SPEAKER_02]: cafeteria.

[00:06:06] [SPEAKER_02]: It had their noise and through the processing, they could have different, um,

[00:06:11] [SPEAKER_02]: different ways that that background noise was processed.

[00:06:16] [SPEAKER_02]: Was it removed entirely?

[00:06:17] [SPEAKER_02]: Was it spread out into a stereo field?

[00:06:20] [SPEAKER_02]: And then the dialogue was the center piece, you know, all these different things to

[00:06:24] [SPEAKER_02]: make it sound more cinematic or whatever.

[00:06:26] [SPEAKER_02]: And I just really wonder from a curiosity standpoint, how those sound like it's a

[00:06:31] [SPEAKER_02]: neat parlor trick, but is it useful?

[00:06:33] [SPEAKER_04]: You know, as you're talking, what I find it, cause I'm, I'm just a writer, so I just

[00:06:37] [SPEAKER_04]: type in letters and that's all I do.

[00:06:39] [SPEAKER_04]: Yeah.

[00:06:39] [SPEAKER_04]: Right.

[00:06:39] [SPEAKER_04]: Whereas in music, you have tracks and in, um, digital imagery, uh, since, uh,

[00:06:49] [SPEAKER_04]: Adobe tools, you have layers.

[00:06:52] [SPEAKER_04]: And, you know, I looked at a tool, I saw a tool to BDMI event some, some time ago

[00:06:57] [SPEAKER_04]: where the, um, cartoon artist was using it to create the behind layer of the

[00:07:02] [SPEAKER_04]: scene, right?

[00:07:04] [SPEAKER_04]: And you could animate and you could do other things, right?

[00:07:06] [SPEAKER_04]: Music.

[00:07:06] [SPEAKER_04]: You have, you have the tracks and layers.

[00:07:08] [SPEAKER_04]: It strikes me as that's, that's an obvious, um, an opening for AI collaboration.

[00:07:16] [SPEAKER_04]: I do these layers.

[00:07:17] [SPEAKER_04]: Oh yeah.

[00:07:18] [SPEAKER_04]: I'll just let that let the AI do that one.

[00:07:20] [SPEAKER_04]: You know, it's boring to do the rhythm track.

[00:07:22] [SPEAKER_04]: It's boring to do this.

[00:07:23] [SPEAKER_04]: It's boring to do something else or, or, or, um, uh, hear what I did and then,

[00:07:30] [SPEAKER_04]: and then nominate something, try something and figure out something else there.

[00:07:33] [SPEAKER_04]: It's interesting that that as opposed to me, we're either I write it or it writes

[00:07:37] [SPEAKER_04]: it.

[00:07:38] [SPEAKER_04]: It's either or that's not the case of the world you work in, right?

[00:07:42] [SPEAKER_02]: Yeah.

[00:07:42] [SPEAKER_02]: Yeah, for sure.

[00:07:43] [SPEAKER_02]: No, the collaborative kind of aspect of things is, is what really excited about

[00:07:47] [SPEAKER_02]: this technology because it is.

[00:07:48] [SPEAKER_02]: Yeah.

[00:07:49] [SPEAKER_02]: And the, and the layer enables that, right?

[00:07:50] [SPEAKER_02]: Exactly.

[00:07:51] [SPEAKER_02]: So it's not all or nothing like some of the stuff, you know, we've talked about

[00:07:55] [SPEAKER_02]: like some of the music generation services like UDO and Suno and stuff and what

[00:08:00] [SPEAKER_02]: they're creating is a single audio file that has all the information in it.

[00:08:05] [SPEAKER_02]: And, and then they're allowing you to later go in and process out the stems so

[00:08:11] [SPEAKER_02]: you can have them isolated.

[00:08:12] [SPEAKER_02]: It doesn't sound as good as if they were just isolated to begin with, right?

[00:08:16] [SPEAKER_02]: Something like what Apple has here kind of seems like it has it more isolated to

[00:08:21] [SPEAKER_02]: begin with.

[00:08:21] [SPEAKER_02]: Um, and you know, that gives you a little bit more flexibility, I guess, in that

[00:08:25] [SPEAKER_02]: collaborative process.

[00:08:26] [SPEAKER_02]: And we'll also, what you're saying really kind of, um, speak, you know, what comes

[00:08:31] [SPEAKER_02]: to mind for me too, is in some of the work that I do from a visual perspective,

[00:08:35] [SPEAKER_02]: like the thumbnail work that I do for the YouTube channel and stuff, I often find

[00:08:39] [SPEAKER_02]: myself in a position where I'm like, I know exactly what I want in the foreground,

[00:08:42] [SPEAKER_02]: but the background, I just need something that's interesting or colorful or

[00:08:47] [SPEAKER_02]: abstract or whatever.

[00:08:49] [SPEAKER_02]: And AI is great for that.

[00:08:51] [SPEAKER_02]: I, you know, I go into ideogram and I give it a standard prompt that I have now.

[00:08:56] [SPEAKER_02]: And it gives me some really great, um, options for backgrounds and you know,

[00:09:00] [SPEAKER_02]: it's, it's fine.

[00:09:01] [SPEAKER_02]: It does the trick.

[00:09:02] [SPEAKER_02]: It satisfies.

[00:09:03] [SPEAKER_02]: It's a nice collaboration.

[00:09:06] [SPEAKER_02]: So anything else from Apple that interested you for sure?

[00:09:09] [SPEAKER_02]: So the, so those were the, the, the like standout things that I was like, if I had

[00:09:14] [SPEAKER_02]: an iPhone, I really can't be right now.

[00:09:17] [SPEAKER_02]: Um, but then you've got the more standard stuff kind of, um, the things like what

[00:09:21] [SPEAKER_02]: they call visual intelligence, which is anything on the screen can be used to gain

[00:09:27] [SPEAKER_02]: more context with the AI.

[00:09:30] [SPEAKER_02]: So we have that on Android, you know, Gemini is, is certainly doing that.

[00:09:34] [SPEAKER_02]: So that feels kind of like, okay, we're seeing this, you know, on everybody's

[00:09:38] [SPEAKER_02]: kind of doing a little bit of this right now.

[00:09:40] [SPEAKER_02]: Um, the, one of the examples that they give in their kind of media marketing

[00:09:44] [SPEAKER_02]: around this feature was photo of a flyer and, you know, use the AI to add the event

[00:09:50] [SPEAKER_02]: details to your calendar.

[00:09:52] [SPEAKER_02]: And that sounds pretty familiar from the, from the Google event a couple of weeks

[00:09:56] [SPEAKER_02]: ago.

[00:09:56] [SPEAKER_02]: Also from did that.

[00:09:58] [SPEAKER_04]: Yeah.

[00:09:58] [SPEAKER_04]: Interesting.

[00:09:59] [SPEAKER_04]: That was also open AI's thing too, wasn't it?

[00:10:02] [SPEAKER_04]: Yeah.

[00:10:02] [SPEAKER_04]: Oh, I don't remember if open AI did that as well.

[00:10:04] [SPEAKER_04]: But it's, yeah.

[00:10:05] [SPEAKER_04]: It's trying to integrate with your apps, which is interesting.

[00:10:07] [SPEAKER_04]: Yeah.

[00:10:08] [SPEAKER_04]: Yeah.

[00:10:08] [SPEAKER_04]: It's getting to be agentive.

[00:10:11] [SPEAKER_02]: Yes.

[00:10:12] [SPEAKER_02]: Yes, indeed.

[00:10:12] [SPEAKER_02]: And, and not just something you talk or type to, but something that you can kind

[00:10:17] [SPEAKER_02]: of point the camera at or, or show the contents of a screen and interact with

[00:10:22] [SPEAKER_02]: that.

[00:10:22] [SPEAKER_02]: That is definitely something you can do on Android devices with Gemini at this

[00:10:28] [SPEAKER_02]: point.

[00:10:28] [SPEAKER_02]: They also showed, you know, uh, demonstrated the example of using a photo

[00:10:32] [SPEAKER_02]: of a restaurant and then interpreting that and then finding the details of the

[00:10:37] [SPEAKER_02]: menu and the hours and stuff.

[00:10:39] [SPEAKER_02]: So when you're walking out and about, you just, Oh, there's the restaurant.

[00:10:42] [SPEAKER_02]: Take a picture.

[00:10:43] [SPEAKER_02]: And it brings you right into all the information you need to know that sort of

[00:10:46] [SPEAKER_02]: stuff.

[00:10:47] [SPEAKER_02]: So that's, you know,

[00:10:48] [SPEAKER_02]: Which all goes to where glasses are going to be soon too.

[00:10:50] [SPEAKER_02]: Yeah.

[00:10:51] [SPEAKER_04]: Yeah.

[00:10:51] [SPEAKER_02]: Right.

[00:10:51] [SPEAKER_02]: Yeah.

[00:10:53] [SPEAKER_02]: Soon.

[00:10:54] [SPEAKER_02]: Define soon.

[00:10:55] [SPEAKER_02]: Yeah.

[00:10:55] [SPEAKER_02]: We'll, we'll see how soon it happens.

[00:10:57] [SPEAKER_04]: So I didn't, I didn't get into detail on this cause I am an Android guy and

[00:11:01] [SPEAKER_04]: live, love you to Google.

[00:11:02] [SPEAKER_04]: Uh, but, uh, I sensed that there was, this is going to be a long rollout of

[00:11:06] [SPEAKER_04]: Apple AI stuff.

[00:11:07] [SPEAKER_02]: Is that, that there was October is what it.

[00:11:10] [SPEAKER_02]: Yeah.

[00:11:10] [SPEAKER_02]: It sounds like October is when many of these features are going to be released.

[00:11:14] [SPEAKER_01]: You know, I did all right.

[00:11:15] [SPEAKER_02]: Um, yeah, I'm not sure exactly to what degree, like, like the things that I, I

[00:11:20] [SPEAKER_02]: believe are coming out in October, uh, better search or like an improved search

[00:11:24] [SPEAKER_02]: experience using AI for your photos, your email, your texts on your device,

[00:11:29] [SPEAKER_02]: writing tools.

[00:11:31] [SPEAKER_02]: So the standard, you know, feature set there, summary, rewriting, proofreading

[00:11:34] [SPEAKER_02]: tone, all that kind of stuff.

[00:11:36] [SPEAKER_02]: Uh, something they call cleanup, which is removing something from an image.

[00:11:40] [SPEAKER_02]: So that's kind of standard stuff.

[00:11:42] [SPEAKER_02]: A memory movie, which is you ask photos to create a movie out of the photos that

[00:11:48] [SPEAKER_02]: particular, uh, fit into a certain query.

[00:11:52] [SPEAKER_02]: So that's, yeah, I guess.

[00:11:54] [SPEAKER_02]: And I think a lot of these are coming also updates to Siri, which is an, I

[00:11:58] [SPEAKER_02]: think is an important one.

[00:12:00] [SPEAKER_02]: So Siri, you know, becomes a little bit more of that voice that you can converse

[00:12:04] [SPEAKER_02]: with.

[00:12:04] [SPEAKER_02]: It can understand when you stumble, it's more of a conversational context.

[00:12:08] [SPEAKER_02]: It can carry out tasks, um, such as adding something to your calendar based on,

[00:12:14] [SPEAKER_02]: you know, details from the conversation, that sort of stuff.

[00:12:17] [SPEAKER_04]: So, um, as Google tried to emphasize how much is going to happen locally, uh,

[00:12:23] [SPEAKER_04]: cause you have this powerful tensor chip on your phone.

[00:12:26] [SPEAKER_04]: Uh, was Apple also pushing that localness to the computation?

[00:12:33] [SPEAKER_02]: Yeah, they, um, you know, Apple is often very, um, vocal about privacy.

[00:12:40] [SPEAKER_02]: Except where I was going.

[00:12:41] [SPEAKER_02]: Exactly.

[00:12:42] [SPEAKER_02]: Yeah.

[00:12:43] [SPEAKER_02]: Yeah.

[00:12:43] [SPEAKER_02]: Yeah.

[00:12:43] [SPEAKER_02]: And what they were talking about, I, I didn't feel like I caught Apple too

[00:12:48] [SPEAKER_02]: often speaking specifically about on device.

[00:12:52] [SPEAKER_02]: Like all of this stays on your device, but they did talk about something and I

[00:12:55] [SPEAKER_02]: did not write down the name of this particular thing, but they essentially

[00:12:59] [SPEAKER_02]: were talking about how a portion of their cloud, they've done the engineering

[00:13:03] [SPEAKER_02]: to ensure it's privateness, you know, to ensure that they don't see the thing.

[00:13:09] [SPEAKER_02]: So yes, some of this stuff does rely on the cloud, but Apple's kind of

[00:13:13] [SPEAKER_02]: sprinkling it's it's Apple privacy magic on it.

[00:13:16] [SPEAKER_02]: And you know, whatever that detail is, I wish I could remember the name and I

[00:13:21] [SPEAKER_02]: don't want to spend the time trying to Google search to find it.

[00:13:23] [SPEAKER_02]: But, um, they did talk that up for a moment and kind of highlight that.

[00:13:28] [SPEAKER_04]: Cause my argument about Apple and privacy has been that they turned a bug

[00:13:32] [SPEAKER_04]: into a feature in that Apple never succeeded at the ad business.

[00:13:34] [SPEAKER_04]: So they didn't need your privacy data.

[00:13:36] [SPEAKER_04]: So they said, Oh, we're going to, all right, then we're the privacy company.

[00:13:40] [SPEAKER_04]: Right?

[00:13:40] [SPEAKER_04]: Yeah.

[00:13:40] [SPEAKER_04]: Now, when you get into wanting to, to capture and analyze and feed back all

[00:13:46] [SPEAKER_04]: of this, all of this personal action, uh, it puts a very different cast on

[00:13:50] [SPEAKER_04]: privacy for Apple and it'd be interesting to see whether they really do have

[00:13:54] [SPEAKER_04]: those muscles well-developed or whether it was just a marketing ploy.

[00:13:59] [SPEAKER_02]: Right.

[00:13:59] [SPEAKER_02]: Right.

[00:14:00] [SPEAKER_02]: Uh, private cloud compute, I believe.

[00:14:02] [SPEAKER_02]: And, um, realizing right now that this is not the first time that they've

[00:14:06] [SPEAKER_02]: mentioned private cloud compute, but that's what they were kind

[00:14:09] [SPEAKER_02]: of highlighting yesterday.

[00:14:10] [SPEAKER_02]: It looks like June 10th is, uh, when they first announced that kind of,

[00:14:17] [SPEAKER_02]: uh, security and privacy, um, effort for their, for their cloud

[00:14:21] [SPEAKER_02]: computing related to AI.

[00:14:22] [SPEAKER_02]: So that's what it is.

[00:14:24] [SPEAKER_02]: Good.

[00:14:24] [SPEAKER_02]: But they did take time to focus on that a little bit.

[00:14:26] [SPEAKER_04]: Were you the least bit tempted to buy an Apple phone?

[00:14:29] [SPEAKER_02]: Well, I mean, if I had all the money in the world, absolutely.

[00:14:32] [SPEAKER_02]: I would just to have one and be able to play around with these features, but

[00:14:36] [SPEAKER_02]: I do not have all the money in the world or even a small portion of the money

[00:14:40] [SPEAKER_02]: in the world that would require be required to buy one of these devices.

[00:14:42] [SPEAKER_02]: See discussion of, uh, Kickstarter folks.

[00:14:45] [SPEAKER_02]: Yes.

[00:14:46] [SPEAKER_02]: Yeah.

[00:14:46] [SPEAKER_02]: Of, uh, yeah.

[00:14:47] [SPEAKER_02]: The Patreon.

[00:14:47] [SPEAKER_02]: Get this poor guy an Apple.

[00:14:50] [SPEAKER_02]: I mean, I do.

[00:14:51] [SPEAKER_02]: I don't, I don't know if I mentioned this, but I do have an iPhone.

[00:14:55] [SPEAKER_02]: Um, because we switched to T-Mobile not too long ago and through it.

[00:15:01] [SPEAKER_02]: Um, they were like, well, you know, we can get you an iPhone 13 for free.

[00:15:05] [SPEAKER_02]: You just pay tax.

[00:15:06] [SPEAKER_02]: And I was like, well, okay.

[00:15:08] [SPEAKER_02]: I mean, at least then I have something in hand.

[00:15:11] [SPEAKER_02]: If some of these features, you know, I want to play around with and everything,

[00:15:13] [SPEAKER_02]: but I mean, at this point it's three years old, even though it's brand new,

[00:15:16] [SPEAKER_02]: it's three years old.

[00:15:17] [SPEAKER_02]: So, you know, some of the, most of these AI features probably not coming to the

[00:15:21] [SPEAKER_02]: iPhone 13 would be my guess.

[00:15:23] [SPEAKER_02]: I could be surprised, but, um, so I don't know that it's going to do me a whole lot

[00:15:27] [SPEAKER_02]: of good.

[00:15:27] [SPEAKER_02]: I could trade it in may, maybe.

[00:15:29] [SPEAKER_04]: Well, that's, I was just thinking, yeah, I looked, I looked for the, for the,

[00:15:31] [SPEAKER_04]: um, not that we're getting a little off AI here, but I looked for the, uh,

[00:15:35] [SPEAKER_04]: pixel nine and my pixel six, I think that they were offering $450 trade in.

[00:15:40] [SPEAKER_02]: Yeah.

[00:15:41] [SPEAKER_02]: I think the iPhone 13, I looked it up yesterday and they're offering somewhere

[00:15:44] [SPEAKER_02]: around like $250 trade in.

[00:15:46] [SPEAKER_02]: So it's not a, it's not a huge amount of trade in, but it's something, you know,

[00:15:50] [SPEAKER_02]: it's a, it all comes down to money for me.

[00:15:52] [SPEAKER_02]: Um, Apple, Apple isn't, isn't the company that just sprinkles out review

[00:15:57] [SPEAKER_02]: devices to anyone who requests them, you know, but you know, I might consider

[00:16:02] [SPEAKER_02]: picking one up for on a temporary basis and playing around with it, sending it

[00:16:06] [SPEAKER_02]: back or, you know, returning it or something just to like get a sense of it.

[00:16:09] [SPEAKER_02]: Um, I don't know.

[00:16:10] [SPEAKER_02]: I don't know if I'm going to do that or not, but, but I'm curious.

[00:16:13] [SPEAKER_02]: I am legitimately curious to see because Apple is such a major player in mobility

[00:16:18] [SPEAKER_02]: and in the smartphone world and Google has been doing the AI thing for awhile.

[00:16:23] [SPEAKER_02]: I think one thing that did stand out to me that I'm curious to see how this

[00:16:27] [SPEAKER_02]: works is it really seemed like what Apple was pushing with the advancements in its

[00:16:33] [SPEAKER_02]: Siri, um, offering and you know, the way that they were showing things off,

[00:16:38] [SPEAKER_02]: it seemed like Apple was going the direction of AI is embedded into the OS now.

[00:16:45] [SPEAKER_02]: And that AI that's embedded there reaches into all these apps in these ways.

[00:16:50] [SPEAKER_02]: And I feel like Google does it a little differently.

[00:16:53] [SPEAKER_02]: Google isn't embedding AI into the OS.

[00:16:56] [SPEAKER_02]: It's making sure that you have the Gemini app installed and kind of putting, uh,

[00:17:01] [SPEAKER_02]: some of these features into apps themselves like Google photos or whatever.

[00:17:05] [SPEAKER_02]: And so I'm wondering, like curiosity is if it's embedded into the OS in an

[00:17:10] [SPEAKER_02]: integrated way, the way Apple has actually proven to be really good at,

[00:17:14] [SPEAKER_02]: cause it controls its whole ecosystem.

[00:17:17] [SPEAKER_02]: Um, is that going to be more enticing for people to actually, um, you know,

[00:17:24] [SPEAKER_02]: integrate them, their own use cases with these features?

[00:17:28] [SPEAKER_02]: Are they going to have more success than Google's going to have on a case by case

[00:17:32] [SPEAKER_02]: basis?

[00:17:33] [SPEAKER_04]: You know, you raise an interesting ecosystem question,

[00:17:35] [SPEAKER_04]: which is whether in an app ecosystem,

[00:17:41] [SPEAKER_04]: you don't want to make it so complex that app developers can't develop apps

[00:17:44] [SPEAKER_04]: because they can't, you know, AI-ify them. Right?

[00:17:48] [SPEAKER_04]: So in that case,

[00:17:49] [SPEAKER_04]: it's got to happen in essence at the central location of Gemini that you can,

[00:17:53] [SPEAKER_04]: you can do a bridge to Gemini and API to Gemini and you can figure out stuff,

[00:17:57] [SPEAKER_04]: but you don't want to have to make a AI stuff in your own app cause it's too

[00:18:01] [SPEAKER_04]: complicated.

[00:18:02] [SPEAKER_04]: So I don't know how that works in terms of the future of app stores and stuff.

[00:18:06] [SPEAKER_02]: Yeah. Yeah. I don't know either. That is a really good question.

[00:18:10] [SPEAKER_04]: Does AI actually becomes useful or not? We'll see.

[00:18:12] [SPEAKER_02]: Yeah. Yeah, totally. I mean, but they all see the benefit of it.

[00:18:16] [SPEAKER_02]: And I think that is, that is telling as well, right? Like if,

[00:18:20] [SPEAKER_02]: if there was no there,

[00:18:21] [SPEAKER_02]: there Apple might not be playing the same game that everybody else is,

[00:18:26] [SPEAKER_02]: but it really seems like they are. They're,

[00:18:28] [SPEAKER_02]: they're putting the summaries in there. They're putting the, you know,

[00:18:31] [SPEAKER_02]: AI eraser tool, whatever they call it in there.

[00:18:35] [SPEAKER_02]: A lot of that, those features are what we've already seen.

[00:18:39] [SPEAKER_02]: What we've already heard is just now on an iPhone. So that,

[00:18:42] [SPEAKER_02]: that gives some credence to the value or what the technology companies believe

[00:18:47] [SPEAKER_02]: to be the value of these tools.

[00:18:49] [SPEAKER_02]: It just still really remains to be seen whether people actually care as much as

[00:18:52] [SPEAKER_02]: they think they do.

[00:18:54] [SPEAKER_04]: Yeah. And that's the, that's the retail level of AI.

[00:18:58] [SPEAKER_04]: And there's the, you know, cause we're here on the show,

[00:19:00] [SPEAKER_04]: we have to talk about B2B and B2C and how much,

[00:19:03] [SPEAKER_04]: what's going to drive the business in the future. We just don't know yet.

[00:19:08] [SPEAKER_02]: Yeah. Yeah, totally. That's all being, uh,

[00:19:11] [SPEAKER_02]: being laid out in front of, in front of our very eyes. Um, well,

[00:19:15] [SPEAKER_02]: speaking of the AI on an app by app basis,

[00:19:18] [SPEAKER_02]: Google also had an announcement this week. What's this, this week? No,

[00:19:22] [SPEAKER_02]: this was the fifth. So this was a late last week, but, um,

[00:19:26] [SPEAKER_02]: they rolled out some new or actually they rolled out one new feature to Google

[00:19:30] [SPEAKER_02]: photos. So this totally relates to what we're just talking about, how app,

[00:19:33] [SPEAKER_02]: you know, integrates into me or AI integrates into the app itself and,

[00:19:37] [SPEAKER_02]: and not the OS layer. But anyways, they rolled out a one feature right now,

[00:19:42] [SPEAKER_02]: which is new and improved search in the app. So for example, you know,

[00:19:47] [SPEAKER_02]: I always felt like Google photos was actually pretty good. If I was like,

[00:19:50] [SPEAKER_02]: you know, red, if I put in a query for red,

[00:19:52] [SPEAKER_02]: it would pull back all the photos that had like red as a predominant color. Um,

[00:19:56] [SPEAKER_02]: used to do that better before than it did in recent,

[00:20:00] [SPEAKER_02]: in the last couple of years, which is interesting to me.

[00:20:02] [SPEAKER_02]: I remember at one time it was really good at that.

[00:20:04] [SPEAKER_02]: And then I'd tried in the last couple of years and didn't have as much success.

[00:20:08] [SPEAKER_02]: Well now they're kind of expanding on that and allowing you to do things like,

[00:20:13] [SPEAKER_02]: you know, there are examples,

[00:20:14] [SPEAKER_02]: Alice and me laughing or kayaking on the lakes surrounded by mountains or Emma

[00:20:19] [SPEAKER_02]: painting in the backyard.

[00:20:20] [SPEAKER_02]: I actually opened up Google photos and tried Jason at Christmas and it worked

[00:20:25] [SPEAKER_02]: really well. It didn't only just pull photos of me on Christmas day.

[00:20:29] [SPEAKER_02]: It pulled up photos of me during the month of December eating at a restaurant

[00:20:34] [SPEAKER_02]: that had Christmas decorations in it. You know what I mean?

[00:20:37] [SPEAKER_02]: Like it understood the context of what I was looking for there. And uh,

[00:20:42] [SPEAKER_02]: yeah, I think that's, that's, that's an interesting capability.

[00:20:46] [SPEAKER_02]: So I tried a few different things.

[00:20:47] [SPEAKER_04]: I asked him for parties and it returned, um,

[00:20:51] [SPEAKER_04]: some things that were like a demonstration,

[00:20:52] [SPEAKER_04]: but it was a lot of people look festive or festival,

[00:20:56] [SPEAKER_04]: but also my father's birthday and a cake.

[00:20:59] [SPEAKER_04]: So it knew that concept of party in a broader sense. It's interesting. Um,

[00:21:05] [SPEAKER_04]: uh, I asked it, uh, you could ask it for things,

[00:21:08] [SPEAKER_04]: favorite places you went to vacation. So I asked it, um,

[00:21:12] [SPEAKER_04]: where have I had schnitzel? Nothing. But if I just searched on schnitzel,

[00:21:16] [SPEAKER_04]: I have five pictures of schnitzels here, which is making me hungry.

[00:21:19] [SPEAKER_04]: Plus a restaurant I forgot that I ate at in 2017 called schnitzel love.

[00:21:24] [SPEAKER_02]: Oh, you know, but it wasn't able to identify that.

[00:21:28] [SPEAKER_02]: Like it knew what schnitzel was.

[00:21:29] [SPEAKER_02]: No, I couldn't answer that question. Interesting.

[00:21:33] [SPEAKER_04]: That's weird. Yeah. Cause I was looking at the story about, you know,

[00:21:36] [SPEAKER_02]: trying to, trying to figure out how to place it one way or the other. Um,

[00:21:41] [SPEAKER_02]: well, okay. So, uh, yes, because there is another part to this story.

[00:21:44] [SPEAKER_02]: So the second part is ask photos and I'm wondering if maybe this is why that's

[00:21:48] [SPEAKER_02]: why I don't think you have it yet. Well, I mean, if,

[00:21:50] [SPEAKER_02]: if what you're talking about is, is from the ask photos kind of feature set,

[00:21:55] [SPEAKER_02]: this is open to select users right now. It's,

[00:21:58] [SPEAKER_02]: it's the conversational search in Google. So the examples they have,

[00:22:02] [SPEAKER_02]: which sounds similar to what you're talking about is what did we eat at the

[00:22:05] [SPEAKER_02]: hotel in Stanley or, you know, the top 10 things we saw in Idaho?

[00:22:09] [SPEAKER_02]: Maybe not though. I mean, it's, it's kind of, they're, they're kind of similar.

[00:22:13] [SPEAKER_02]: I mean, Jason at Christmas seems similar to the query that you put in there.

[00:22:17] [SPEAKER_02]: So I don't know why it wouldn't have worked that way. But, um, anyways,

[00:22:20] [SPEAKER_02]: ask photos not rolled out to everyone yet, but coming soon. Um,

[00:22:24] [SPEAKER_04]: what did I do at Disneyland? It gave me Disneyland,

[00:22:26] [SPEAKER_04]: but it recognized the word Disneyland. Yeah. I put in for happy me,

[00:22:33] [SPEAKER_04]: um, happy me, happy me.

[00:22:36] [SPEAKER_04]: And it gave me two pictures that were happy,

[00:22:38] [SPEAKER_04]: but one or I'm just looking like this. I don't know.

[00:22:44] [SPEAKER_02]: Oh, now I'm so curious. Happy. Yeah. See, I haven't been putting me,

[00:22:48] [SPEAKER_02]: I've been more, uh, putting Jason, you know,

[00:22:52] [SPEAKER_04]: just because I know that I'm asking for something about, uh,

[00:22:55] [SPEAKER_02]: happy daughters. Uh, Oh, that's an interesting,

[00:23:00] [SPEAKER_02]: an interesting idea. Sorry. I'm not showing everybody this.

[00:23:03] [SPEAKER_02]: You don't want to show them that. No, it's my photo roll. Um,

[00:23:07] [SPEAKER_02]: happy daughters. I'm seeing pictures of my daughters smiling. Yeah. All right.

[00:23:11] [SPEAKER_02]: Like, I don't know that I'm seeing everyone.

[00:23:14] [SPEAKER_02]: I don't think that it's necessarily a cohesive and comprehensive collection

[00:23:17] [SPEAKER_02]: of all the photos of my daughter smiling. Cause I, cause you know, there's a,

[00:23:21] [SPEAKER_02]: there's a handful of photos dating between now and July. Right.

[00:23:25] [SPEAKER_02]: And I'm sure there's more of them in there. Oh, that's July of last year. Yeah.

[00:23:29] [SPEAKER_02]: Definitely not comprehensive. Uh, but nonetheless, it gives me a,

[00:23:33] [SPEAKER_02]: a few options here and there, I suppose.

[00:23:36] [SPEAKER_04]: That's an, and there's my absolutely most boring picture collection ever done.

[00:23:40] [SPEAKER_04]: Whiteboards

[00:23:43] [SPEAKER_02]: compile and make a compile and summarize all the whiteboards.

[00:23:48] [SPEAKER_04]: Oh, we're going to save this. Like it's history and you got to hold onto it.

[00:23:52] [SPEAKER_04]: So I've got tons of pictures of whiteboards in here, but,

[00:23:55] [SPEAKER_04]: but the time you don't do that, Jeff is the time.

[00:23:57] [SPEAKER_04]: I know the germ of brilliant idea. Great invention, a patent that'll make a

[00:24:02] [SPEAKER_04]: fortune. Yeah.

[00:24:04] [SPEAKER_02]: I mean, you could store those in your screenshots app now on the pixel and,

[00:24:08] [SPEAKER_02]: and it will, uh, you know,

[00:24:10] [SPEAKER_02]: allow you to contextualize that to a certain degree,

[00:24:13] [SPEAKER_02]: depending on the complexity of it for sure. So you've got options now.

[00:24:18] [SPEAKER_02]: Uh, anyways, interesting stuff. I've,

[00:24:19] [SPEAKER_02]: I've always really enjoyed the Google photos app.

[00:24:22] [SPEAKER_02]: It's definitely one of my favorite apps that Google has been working on over the

[00:24:26] [SPEAKER_02]: years. A large reason aside from the fact that it stores all my photos in the

[00:24:31] [SPEAKER_02]: cloud with, you know, very easily,

[00:24:33] [SPEAKER_02]: but it's because of this contextual stuff, finding things, you know,

[00:24:37] [SPEAKER_02]: when our data sets in this case,

[00:24:39] [SPEAKER_02]: our photos are hundreds of thousands of photos, you know, it's,

[00:24:44] [SPEAKER_02]: it's really hard to find the thing that you need or you want in any given moment.

[00:24:47] [SPEAKER_02]: So having a really effective and capable search, uh,

[00:24:52] [SPEAKER_02]: that allows you to get granular with it, I think is really interesting.

[00:24:56] [SPEAKER_02]: It seems like Google's moving in that direction.

[00:24:58] [SPEAKER_04]: And then what else can you do with that? It's, it's, it's, it's,

[00:25:00] [SPEAKER_04]: if it could recognize if all these things we saw in an Android because of the AI,

[00:25:05] [SPEAKER_04]: you know,

[00:25:05] [SPEAKER_04]: you can recognize the text in an image and find out that's a date and put that in

[00:25:09] [SPEAKER_04]: your calendar. Right. Yeah. Um,

[00:25:11] [SPEAKER_04]: this goes to the discussion we had before where you're doing that with

[00:25:13] [SPEAKER_04]: screenshots. This strikes me as the, um,

[00:25:18] [SPEAKER_04]: bridge to photos and eventually also glasses. Like, like,

[00:25:23] [SPEAKER_04]: um, what did I have? What did we have for dinner or that restaurant? We liked

[00:25:28] [SPEAKER_02]: the image of it. Can it use the words from it? Yeah.

[00:25:32] [SPEAKER_02]: Or, Oh, I know this person. Dang. What's when's the last person time I saw this

[00:25:36] [SPEAKER_02]: person. Oh yeah. Oh yeah. Oh yeah. Then I can be like, Oh, Hey,

[00:25:41] [SPEAKER_02]: we had such a good time at six flags or whatever the case may be.

[00:25:46] [SPEAKER_02]: That can also go in weird directions too. Cause people have, um,

[00:25:50] [SPEAKER_02]: people definitely feel strangely about, you know,

[00:25:53] [SPEAKER_02]: using something like that to like identify other people, you know, that's right.

[00:25:57] [SPEAKER_02]: Goes in bad, bad places pretty quickly. But anyways, yeah,

[00:26:00] [SPEAKER_02]: I think you're absolutely right.

[00:26:02] [SPEAKER_02]: Interesting direction of the technology there. Speaking of Google, uh,

[00:26:07] [SPEAKER_02]: that new research by a research consultancy called a Thorough,

[00:26:11] [SPEAKER_02]: a Thorough toss, a Thorough toss, uh, showed that AI overview.

[00:26:16] [SPEAKER_02]: That's probably an AI made up company name, but

[00:26:20] [SPEAKER_02]: just reminds me of, uh, who is it? Is it someone from South park?

[00:26:26] [SPEAKER_02]: Anyways, uh, somebody out there got that, uh, showed that AI overviews.

[00:26:30] [SPEAKER_02]: So this is Google search product where you put in a search and in their normal

[00:26:35] [SPEAKER_02]: search product and up at the top of the search is the little AI kind of summary

[00:26:39] [SPEAKER_02]: or collection of information from some of the results.

[00:26:43] [SPEAKER_02]: They were looking at kind of the, the,

[00:26:46] [SPEAKER_02]: the presence of that product one, because if you remember,

[00:26:50] [SPEAKER_02]: this was rolled out, you know,

[00:26:52] [SPEAKER_02]: a couple of months ago and it was rolled out widely and then Google really peeled

[00:26:57] [SPEAKER_02]: it back because it was offering pretty poor information up there.

[00:27:01] [SPEAKER_02]: Um, in some cases dangerous information.

[00:27:04] [SPEAKER_02]: And so Google really kind of peeled it back. Well, according to this research,

[00:27:10] [SPEAKER_02]: now it's being in their study of it anyways,

[00:27:13] [SPEAKER_02]: it's being offered around 70% or sorry,

[00:27:16] [SPEAKER_02]: 17% of the time for queries in the UK and the U S currently.

[00:27:21] [SPEAKER_02]: Um, and then, uh, let's see here. They,

[00:27:25] [SPEAKER_02]: they found that AI answers are not likely to be given with current events queries.

[00:27:30] [SPEAKER_02]: So it happens sometimes, but not all the time. But on a deeper level,

[00:27:35] [SPEAKER_02]: it seems to impact publisher visibility in search,

[00:27:40] [SPEAKER_02]: which is a really big question and concern for,

[00:27:43] [SPEAKER_02]: for businesses who are creating content.

[00:27:45] [SPEAKER_02]: That one example they gave is that Martha Stewart article on how to kill wasps,

[00:27:50] [SPEAKER_02]: um, is in the, the,

[00:27:53] [SPEAKER_02]: the AI rewrite of that happens to be in the top position when you do that query.

[00:27:57] [SPEAKER_02]: The original article appears on the third full page scroll down when you're on

[00:28:03] [SPEAKER_02]: mobile. So it's pulling that information from that article, but it's,

[00:28:07] [SPEAKER_02]: it's putting that article way down in the search results versus the AI overview

[00:28:12] [SPEAKER_02]: that's up at the very top.

[00:28:13] [SPEAKER_04]: So I thought about that one as an example, right?

[00:28:15] [SPEAKER_04]: It's one matter if you have exclusive reporting about something,

[00:28:18] [SPEAKER_04]: but Martha is not the only source for knowing how to kill wasps as cruel as

[00:28:22] [SPEAKER_04]: Martha may be to buzzing insects. She ain't the only one. Yeah.

[00:28:27] [SPEAKER_04]: And so if you think about this, not from the publisher's perspective,

[00:28:30] [SPEAKER_04]: but from the user's perspective,

[00:28:32] [SPEAKER_04]: all you want to know as the thing is buzzing around you and you're getting

[00:28:35] [SPEAKER_04]: stung, how do I kill wasps? Right. Yeah. And it's information.

[00:28:39] [SPEAKER_04]: It's not treatment. There's not copyright. There's not, it's I, I,

[00:28:45] [SPEAKER_04]: I think we've got to recognize that that kind of commodified information,

[00:28:49] [SPEAKER_04]: um, is not where you find your value anymore. And, and yes, this ruins it.

[00:28:54] [SPEAKER_04]: It happens to be that this kind of search ruins it, but, um, so be it. I mean,

[00:28:59] [SPEAKER_04]: I could just as easily call the library librarian could go look up what you used

[00:29:03] [SPEAKER_04]: to do with the old days.

[00:29:04] [SPEAKER_04]: Nice librarians would go look up a book and come back on the phone and tell you

[00:29:07] [SPEAKER_04]: how to kill the wasp.

[00:29:08] [SPEAKER_04]: And they're probably not giving you a citation that this is a wonderful

[00:29:13] [SPEAKER_02]: magazine. It's right.

[00:29:14] [SPEAKER_02]: Found in the book by author or blah, blah, blah.

[00:29:17] [SPEAKER_04]: And all you care about is the information, right? Interesting. So, yeah.

[00:29:21] [SPEAKER_04]: Um, a publisher's are going to scream bloody murder,

[00:29:24] [SPEAKER_04]: insist their own all kinds of money because of this. But, um,

[00:29:28] [SPEAKER_04]: if all they're giving you is commodity information,

[00:29:30] [SPEAKER_02]: it's going to be treated as a commodity.

[00:29:32] [SPEAKER_02]: So one question I have that's super interesting. Um,

[00:29:35] [SPEAKER_02]: I'm happy that you said that, that,

[00:29:37] [SPEAKER_02]: what that brings to mind for me is one question that I have here is how in the

[00:29:43] [SPEAKER_02]: court, in the course of this study,

[00:29:45] [SPEAKER_02]: are they saying that Martha Stewart's information was used primarily because

[00:29:49] [SPEAKER_02]: that's the reference that Google gives or is that AI generated blurb a,

[00:29:55] [SPEAKER_02]: an amalgamation of many sources that say the same thing and they're just,

[00:30:01] [SPEAKER_02]: and Martha Stewart was just one of them.

[00:30:04] [SPEAKER_02]: And so they're linking to just one of them because they're not using all of

[00:30:06] [SPEAKER_04]: them. Right? No. But I logically,

[00:30:08] [SPEAKER_04]: I would have to believe it's the latter in the sense that how, unless it,

[00:30:12] [SPEAKER_04]: unless the system knows that Martha Stewart is a tremendous, uh,

[00:30:17] [SPEAKER_04]: resource and an emology. Yeah. Right. That's the wrong word.

[00:30:21] [SPEAKER_02]: And trustworthy. Like it always, always gives the correct information.

[00:30:26] [SPEAKER_04]: What it's going to value is the commonality. Yes.

[00:30:30] [SPEAKER_04]: Kill and wasps includes these elements in enough places cause it's all

[00:30:33] [SPEAKER_04]: relational, right? It's all about it's, it's, it's a, it's a, um, uh,

[00:30:40] [SPEAKER_04]: a chart of the relationships of words. That's what the train is, right?

[00:30:44] [SPEAKER_04]: So it's got to have found those words in common relationships efficiently to

[00:30:49] [SPEAKER_04]: pump up those tokens to say, we're going to say wasp,

[00:30:55] [SPEAKER_04]: kill soap, whatever it said here.

[00:30:59] [SPEAKER_02]: How do you go? Whatever the solution was now I'm kind of curious cause I don't

[00:31:03] [SPEAKER_04]: like soap and water is the first way. And other way is vinegar.

[00:31:06] [SPEAKER_02]: Other way is wasp spray. Well, yeah. Okay. Oh, go buy it.

[00:31:10] [SPEAKER_02]: That makes a lot of sense. Yeah. Let me tell you the trick with wasps,

[00:31:13] [SPEAKER_02]: what you do is you use wasp killing spray.

[00:31:18] [SPEAKER_02]: Exactly. Keep it, keep it a secret.

[00:31:20] [SPEAKER_02]: We don't want everyone to know the solution here. Right. Um, yeah. So that's,

[00:31:24] [SPEAKER_02]: that's really interesting. Um, okay. Well,

[00:31:27] [SPEAKER_02]: I do also think that this is interesting from the perspective of the fact that

[00:31:32] [SPEAKER_02]: Google has said already in the past that AI overviews get more clicks for

[00:31:36] [SPEAKER_02]: sites. Um, not less.

[00:31:40] [SPEAKER_02]: They have not provided my understanding anyways,

[00:31:43] [SPEAKER_02]: any evidence to support that assertion other than saying this is,

[00:31:47] [SPEAKER_02]: this is what we've found. Um, so yeah, I think jury's still out. I mean,

[00:31:52] [SPEAKER_02]: the, the, you know,

[00:31:53] [SPEAKER_02]: part of this report was really just trying to make it the case for the fact that

[00:31:58] [SPEAKER_02]: search algorithms and what, you know,

[00:32:01] [SPEAKER_02]: the signals that impact search rankings are going to be widely impacted because

[00:32:07] [SPEAKER_02]: of AI features like, uh, like this,

[00:32:11] [SPEAKER_02]: like AI overviews and people are just going to have to figure out how they play

[00:32:15] [SPEAKER_02]: the game differently. But that's kind of also always the case, right?

[00:32:18] [SPEAKER_02]: Because Google search algorithms really do change over time. They're,

[00:32:22] [SPEAKER_02]: they're never, uh, stationary and stagnant for very, for too long. Anyways.

[00:32:27] [SPEAKER_02]: I mean,

[00:32:27] [SPEAKER_04]: the great example of this is that publishers scream bloody murder over,

[00:32:30] [SPEAKER_04]: we've talked about this before about perplexities discovery. Yeah. And you know,

[00:32:34] [SPEAKER_04]: if I've any given the story, I just pulled up the Google loses appeal,

[00:32:38] [SPEAKER_04]: $2.7 billion in EU antitrust fund. Fine.

[00:32:41] [SPEAKER_04]: So they do that as a collaborated story like the kind of box on Google.

[00:32:46] [SPEAKER_04]: Uh, but they do cite and the four stories they cite are not identical,

[00:32:52] [SPEAKER_04]: but all the same information is there on top.

[00:32:55] [SPEAKER_04]: So you can have a human being rewriting each other,

[00:32:57] [SPEAKER_04]: which is what we do as journalists or the machine in turn does it because they do

[00:33:02] [SPEAKER_02]: it. And what's worse, I'm not so sure.

[00:33:07] [SPEAKER_02]: What's worse or what's even different there.

[00:33:10] [SPEAKER_02]: Well, and then we'll figure out about the machine doing the rewriting thing.

[00:33:14] [SPEAKER_02]: But I mean, if humans are doing it, like what is the difference here? You know,

[00:33:17] [SPEAKER_02]: other than it's automatic.

[00:33:19] [SPEAKER_04]: Right. And what's better here is that it can then, so ask a question.

[00:33:23] [SPEAKER_04]: What are the main arguments Apple used? Well,

[00:33:26] [SPEAKER_04]: now it's going to think and it's going to give me, um,

[00:33:28] [SPEAKER_04]: um, four answers, right? Compliance with Irish tax laws,

[00:33:33] [SPEAKER_04]: economic contributions, state aid, misinterpretation,

[00:33:36] [SPEAKER_04]: and international taxation principles. Okay.

[00:33:38] [SPEAKER_04]: I got that information a lot faster than I would reading a 20 inch narrative of

[00:33:44] [SPEAKER_04]: the whole story where I've got to get all the background and I got everything

[00:33:46] [SPEAKER_04]: else. So that's the other problem is that,

[00:33:48] [SPEAKER_04]: is that we're just not built for that world. Yeah.

[00:33:53] [SPEAKER_04]: Yeah. Interesting. It's not like it's trying to kill us.

[00:33:56] [SPEAKER_04]: It's like consumers will want a different path to getting information they want.

[00:34:02] [SPEAKER_04]: And if we're not part of that path, well,

[00:34:06] [SPEAKER_04]: tough cookies as they used to say on Saturday night live. Yep. Yep.

[00:34:10] [SPEAKER_02]: Left behind. Uh, and then one more before we take a break,

[00:34:15] [SPEAKER_02]: a new treaty called the framework convention on artificial intelligence signed

[00:34:20] [SPEAKER_02]: by the U S the UK and the EU and uh,

[00:34:25] [SPEAKER_02]: few other places in there as well. But I think all,

[00:34:28] [SPEAKER_02]: all contained within the EU if I'm not mistaken. Um,

[00:34:31] [SPEAKER_02]: it ensures that artificial intelligence aligns with quote human rights,

[00:34:36] [SPEAKER_02]: democracy and the rule of law and each signatory must,

[00:34:41] [SPEAKER_02]: and this is a quote adopt or maintain appropriate legislative administrative or

[00:34:46] [SPEAKER_02]: other measures that are reflected by the framework.

[00:34:50] [SPEAKER_02]: So I don't know that there's necessarily a like must or what?

[00:34:55] [SPEAKER_02]: They're just making the promise. Is there any sort of, you know,

[00:34:59] [SPEAKER_04]: yeah, I don't see here. No,

[00:35:01] [SPEAKER_04]: I mean even though they say it's legally enforceable and all that,

[00:35:03] [SPEAKER_04]: it's a treaty. It just means that it's a framework wherein governments can deal

[00:35:08] [SPEAKER_04]: with these kinds of issues. I think that's okay.

[00:35:10] [SPEAKER_04]: I don't have a problem with that.

[00:35:11] [SPEAKER_04]: The problem I have is things like the California and legislation now on the

[00:35:14] [SPEAKER_04]: governor's desk. Um, is it gets so specific in what it's doing?

[00:35:19] [SPEAKER_04]: I think frameworks as a beginning point are okay. So I don't,

[00:35:22] [SPEAKER_04]: I don't object to that. And the AI law in the EU is more of a framework than it

[00:35:26] [SPEAKER_04]: is a statute. I think that's okay.

[00:35:30] [SPEAKER_02]: Yeah. But you know, it's the first, uh,

[00:35:33] [SPEAKER_02]: legally enforceable AI treaty as a, as the verge points out.

[00:35:37] [SPEAKER_02]: So notable, I suppose, and in that regard, and you know,

[00:35:41] [SPEAKER_02]: potentially a model for other, other things that will come after it.

[00:35:47] [SPEAKER_02]: Um, how much impact it has in the grand scheme of things,

[00:35:50] [SPEAKER_02]: I guess remains to be seen, but there we are. All right.

[00:35:55] [SPEAKER_02]: Let's take a quick break. And then in the latter half of the episode,

[00:36:00] [SPEAKER_02]: we've got some really interesting and it really does seem like the format of this

[00:36:06] [SPEAKER_02]: show is becoming like the first half is like the, Oh, what is the,

[00:36:10] [SPEAKER_02]: what is the news? The gruffy news. And then the last half is like,

[00:36:13] [SPEAKER_02]: Oh my goodness, check this out, you know? And I'm okay with it. I like it.

[00:36:17] [SPEAKER_02]: I like having a little bit of that in there. That's coming up.

[00:36:19] [SPEAKER_04]: We also, we also have some stuff coming up in the comments.

[00:36:22] [SPEAKER_04]: If I can just report quickly,

[00:36:23] [SPEAKER_04]: Jesse Scott came in the comments and said that notebook LM is rolling out audio

[00:36:27] [SPEAKER_04]: overview, a feature that turns your documents,

[00:36:30] [SPEAKER_04]: slides and charts into engaging audio discussions.

[00:36:34] [SPEAKER_04]: When you open a notebook,

[00:36:35] [SPEAKER_04]: I'm reading from nine to five Google now and go into a high level guide.

[00:36:39] [SPEAKER_04]: It already lets you create a fact table of contents and so on. Um,

[00:36:43] [SPEAKER_04]: but now they can summarize your added sources and make connections between

[00:36:46] [SPEAKER_04]: topics and banter. So it sounds like we're replaced, Jason. Yeah,

[00:36:51] [SPEAKER_04]: it's a podcast. All we have to do is give it the run notebook,

[00:36:54] [SPEAKER_04]: allow the rundown and there's it. We've got to show.

[00:36:57] [SPEAKER_02]: Oh my goodness. Like I'm, I'm half tempted and half not like just,

[00:37:02] [SPEAKER_02]: just to see what it, what would come out of that.

[00:37:04] [SPEAKER_02]: But then I've also heard podcasts that are the, Oh, well this podcast is done by

[00:37:10] [SPEAKER_02]: AI and that's the cool thing about it. I'm like, Oh no, thank you.

[00:37:14] [SPEAKER_02]: So as, as tempted as I am to explore that, uh, I doubt it'll be released.

[00:37:18] [SPEAKER_04]: Okay. Jesse, you try to replace us.

[00:37:22] [SPEAKER_02]: But now we're coming back after that.

[00:37:25] [SPEAKER_02]: That's right. I love it though. I think that was a,

[00:37:28] [SPEAKER_02]: that was a good story to sneak in there. Thank you, Jesse. All right,

[00:37:30] [SPEAKER_02]: quick break then we'll come back. All right. So, uh,

[00:37:37] [SPEAKER_02]: I thought this was interesting. You put in a, uh, a story or not a story,

[00:37:42] [SPEAKER_02]: more like a report by Waymo

[00:37:46] [SPEAKER_04]: between transparency and PR. Yeah, yeah, yeah, yeah.

[00:37:50] [SPEAKER_02]: So what is this? It's a safety impact page where they're really, you know,

[00:37:55] [SPEAKER_02]: at least trying to show data that they've compiled about their vehicle fleet

[00:38:00] [SPEAKER_02]: ultimately in an effort to show, Hey, you know, we're doing pretty good.

[00:38:04] [SPEAKER_02]: Basically they say that Waymo vehicles are,

[00:38:07] [SPEAKER_02]: have been reported to be in around 73% fewer injury causing crashes compared to

[00:38:15] [SPEAKER_02]: human drivers, which is, which is, this is a,

[00:38:18] [SPEAKER_04]: this is like an SAT test question.

[00:38:20] [SPEAKER_04]: So it's 73% fewer injury causing crashes or yes,

[00:38:25] [SPEAKER_04]: compared to what and given what raw number.

[00:38:28] [SPEAKER_04]: So then it has a down arrow and says 46 fewer,

[00:38:32] [SPEAKER_04]: 46 fewer than what's the total.

[00:38:35] [SPEAKER_04]: I'm trying to get a sense of the larger image here, right?

[00:38:39] [SPEAKER_04]: And I don't know much here. Yeah. Well it's, and,

[00:38:43] [SPEAKER_02]: and so a little extra context compared to a human driver over the same 22 mile

[00:38:49] [SPEAKER_02]: distance in Phoenix and San Francisco,

[00:38:53] [SPEAKER_02]: the Waymo driver had, and then 73% fewer.

[00:38:57] [SPEAKER_04]: But I don't know is how many crashes injury causing crashes did it have?

[00:39:01] [SPEAKER_02]: Yes. Right. Well,

[00:39:03] [SPEAKER_02]: and especially when you consider how many like actual vehicles are driving in

[00:39:10] [SPEAKER_02]: these areas, which is not that many, I don't think.

[00:39:14] [SPEAKER_04]: Like that was the frustration. I put this in here,

[00:39:15] [SPEAKER_04]: but that's the frustration I had was I couldn't get a sense of

[00:39:21] [SPEAKER_02]: raw numbers. Yeah. Yeah. I mean it's a percentage, but yeah,

[00:39:26] [SPEAKER_02]: you want, you want to be able to dial into that a little bit more.

[00:39:28] [SPEAKER_02]: Waymo vehicles have now driven 22 million miles since launch in LA,

[00:39:35] [SPEAKER_02]: San Francisco and Phoenix with a fleet of around 300 vehicles.

[00:39:38] [SPEAKER_02]: So that's a small number of vehicles that you know,

[00:39:42] [SPEAKER_02]: and I guess the question that I have is,

[00:39:45] [SPEAKER_02]: is this information done in a way where that,

[00:39:49] [SPEAKER_02]: that small amount of vehicles is truly representative of a bigger picture or,

[00:39:55] [SPEAKER_02]: you know, and, and I mean,

[00:39:56] [SPEAKER_02]: obviously from what I can tell,

[00:39:58] [SPEAKER_02]: we don't have that information to know that for sure.

[00:40:00] [SPEAKER_04]: I mean, one matter, the other thing to be fair to them, um,

[00:40:03] [SPEAKER_04]: is they could have picked Des Moines,

[00:40:06] [SPEAKER_04]: San Francisco and first you're Angeles are challenging places to drive.

[00:40:11] [SPEAKER_02]: Yeah. Really good point. Really good point. Yeah. I don't, I mean, yes,

[00:40:16] [SPEAKER_02]: LA is,

[00:40:17] [SPEAKER_02]: is a challenging place to drive because there's so much traffic and everything

[00:40:20] [SPEAKER_02]: like that. San Francisco is just a unique geography that is, Oh,

[00:40:24] [SPEAKER_02]: it's so complicated. It's complicated for humans driving,

[00:40:27] [SPEAKER_02]: driving around San Francisco. I'm so happy.

[00:40:29] [SPEAKER_02]: I don't have to deal with that every day if I'm honest. Um, yeah,

[00:40:32] [SPEAKER_02]: they also said 48% fewer police reported crashes,

[00:40:36] [SPEAKER_02]: 84% fewer airbag deployment crashes.

[00:40:40] [SPEAKER_02]: So, you know, by that metric or by the,

[00:40:44] [SPEAKER_02]: by those examples seemingly safer than humans, right.

[00:40:48] [SPEAKER_02]: Still going to run up against the,

[00:40:50] [SPEAKER_02]: the ongoing kind of struggle and, and,

[00:40:54] [SPEAKER_02]: and difficulty that I think autonomous vehicle makers run into,

[00:40:58] [SPEAKER_02]: which is just the fact that it, for, for whatever reason, be it right or wrong,

[00:41:04] [SPEAKER_02]: we as humans have a hard time accepting any sort of

[00:41:08] [SPEAKER_02]: bad outcome from an autonomous vehicle versus

[00:41:13] [SPEAKER_02]: the human side of things where we are more accepting of the fact that humans are

[00:41:18] [SPEAKER_02]: flawed,

[00:41:18] [SPEAKER_02]: but we are not accepting of the fact that computers and systems like this are

[00:41:23] [SPEAKER_02]: also flawed. They're not going to work 100% perfectly.

[00:41:26] [SPEAKER_04]: You know my friend David Weinberger who's written a whole bunch of books

[00:41:29] [SPEAKER_04]: starting as coauthor of Clue Train Manifesto. Um,

[00:41:33] [SPEAKER_04]: and everything's miscellaneous and another wonderful, wonderful books.

[00:41:36] [SPEAKER_04]: He's a philosopher. Um,

[00:41:39] [SPEAKER_04]: he made the point that an accident is not an accident.

[00:41:42] [SPEAKER_04]: An accident is something that we can't explain.

[00:41:45] [SPEAKER_04]: Right. And I think that's what throws us off in human terms is,

[00:41:49] [SPEAKER_04]: Oh, that was just an accident. Um, no,

[00:41:53] [SPEAKER_04]: it had some cause and effect that went on and make it happen. Right.

[00:41:56] [SPEAKER_04]: There's something behind it. Yeah.

[00:41:57] [SPEAKER_04]: And it's interesting that you can't necessarily explain that when it comes to us

[00:42:01] [SPEAKER_04]: because we're just strange random beings.

[00:42:04] [SPEAKER_04]: But here,

[00:42:05] [SPEAKER_04]: what causes accidents and why is going to yield more data.

[00:42:09] [SPEAKER_04]: It'll be interesting to kind of understand the causes,

[00:42:11] [SPEAKER_04]: not only on the way most side, but also on the other vehicle side,

[00:42:15] [SPEAKER_04]: the human side of a crash. Was it at way most fault or not?

[00:42:19] [SPEAKER_04]: Is there such a thing as fault here? It's just interesting.

[00:42:21] [SPEAKER_04]: I'm curious, Jason, when you think about a self-driving car in your head,

[00:42:25] [SPEAKER_04]: in the automatic categorization in your head,

[00:42:28] [SPEAKER_04]: does that fall under robotics to you?

[00:42:32] [SPEAKER_02]: Yeah, I think to a certain degree it does. I mean, I do have a Tesla. Uh,

[00:42:37] [SPEAKER_02]: we have a Tesla here, a model Y that we got, you know, at this point,

[00:42:40] [SPEAKER_02]: I think we got it like three years ago almost. And, um,

[00:42:44] [SPEAKER_02]: not too long ago,

[00:42:46] [SPEAKER_02]: Tesla rolled out a like a month free trial for everyone to use their fully

[00:42:52] [SPEAKER_02]: autonomous, um, driving capability or the, you know, the,

[00:42:56] [SPEAKER_02]: the one that they have, whatever they call it, I can't remember.

[00:42:58] [SPEAKER_02]: And so that was my first like personal firsthand experience with it.

[00:43:03] [SPEAKER_02]: And it took a little trust. It took a little time for me to be like, okay,

[00:43:06] [SPEAKER_02]: I feel comfortable with this now. Um,

[00:43:09] [SPEAKER_02]: I'm still didn't get in completely comfortable with it.

[00:43:12] [SPEAKER_02]: It felt pretty robotic to me. I mean, you know, it's, it's changing lanes.

[00:43:16] [SPEAKER_02]: It's, it's,

[00:43:18] [SPEAKER_02]: it's deciding that now's the right time to go into the other lane or,

[00:43:22] [SPEAKER_02]: or speed up or slow down. You know, it's making a lot of decisions that feels,

[00:43:25] [SPEAKER_02]: that autonomy feels robotic to me when it's controlling the vehicle like that.

[00:43:30] [SPEAKER_02]: So I would say yes. What about you?

[00:43:32] [SPEAKER_04]: Yeah, I think so. I just don't, you, we, we tend to have a mind's eye of robots.

[00:43:37] [SPEAKER_04]: And, um, uh, a car does not seem to be there,

[00:43:41] [SPEAKER_04]: but I think it's absolutely part of that. Yeah. I mean,

[00:43:42] [SPEAKER_04]: the difference between that and Amazon warehouse thing that's going around on

[00:43:46] [SPEAKER_04]: wheels, uh, it's all robotic. Yeah,

[00:43:49] [SPEAKER_02]: absolutely. Yeah. And it's driven by the same type of system.

[00:43:52] [SPEAKER_02]: Just a different format just happens to be carrying around human passengers

[00:43:56] [SPEAKER_02]: versus carrying around, you know, a big box from the top shelf in the warehouse.

[00:44:01] [SPEAKER_02]: Um, I would absolutely. Yeah. The more that I think about this,

[00:44:05] [SPEAKER_02]: the more I realize, yes, I think the answer for me is yes.

[00:44:08] [SPEAKER_04]: I think our good friend Ozone asks, says that, uh, that he, um,

[00:44:14] [SPEAKER_04]: tends to distrust internal reports from any company. Uh,

[00:44:18] [SPEAKER_04]: this is to answer your question. This is a Waymo issued report.

[00:44:22] [SPEAKER_04]: I don't think it's independent. Oh, 100%. Yeah.

[00:44:26] [SPEAKER_02]: This is not an impartial report. This is Waymo reporting on its own, you know,

[00:44:31] [SPEAKER_02]: insights on its fleet and everything.

[00:44:33] [SPEAKER_02]: And so you have to take that with a grain of salt and they,

[00:44:36] [SPEAKER_04]: then they report their methodology and so on.

[00:44:40] [SPEAKER_04]: So it'd be interesting to dig into it. Yeah. And this is a nightmare.

[00:44:44] [SPEAKER_04]: This drives me nuts that, that especially Tesla's is not regulated.

[00:44:49] [SPEAKER_04]: Waymo, Waymo,

[00:44:50] [SPEAKER_04]: I think it's more regulated and has to go through a job through more hoops

[00:44:53] [SPEAKER_04]: because there's no human being in the car. But yeah, the self-driving,

[00:44:57] [SPEAKER_04]: so-called self-driving from, from Tesla,

[00:45:00] [SPEAKER_04]: the fact that that didn't undergo a lot more regulation,

[00:45:02] [SPEAKER_04]: a lot more research with a lot more data just drives me crazy.

[00:45:06] [SPEAKER_04]: It's out there running at 80 miles an hour and can kill people. Okay.

[00:45:10] [SPEAKER_04]: That's a few year boy. You all want to do it. Okay. Yeah. Yeah.

[00:45:15] [SPEAKER_02]: Very interesting. That's a good point. Um, well in less lethal,

[00:45:21] [SPEAKER_02]: uh, news potentially, uh,

[00:45:23] [SPEAKER_02]: I can't think of any lethal news that would be attached to this,

[00:45:26] [SPEAKER_02]: but audible is inviting a small group of audio book narrators to use its,

[00:45:32] [SPEAKER_02]: uh, AI to create replicas of their voice.

[00:45:34] [SPEAKER_02]: I feel like we've talked about this to a certain degree in the past about the

[00:45:37] [SPEAKER_02]: possibility of this happening and sorry,

[00:45:40] [SPEAKER_02]: I don't have a Bloomberg subscription so I can't show you the, uh,

[00:45:43] [SPEAKER_02]: the unencumbered version of this page, but, um,

[00:45:46] [SPEAKER_02]: use their AI to create replicas of their voices to be used on the audio book

[00:45:52] [SPEAKER_02]: creation exchange.

[00:45:54] [SPEAKER_02]: So the ACE authors can use the exchange to select the,

[00:46:00] [SPEAKER_02]: the voices that they want to read their books.

[00:46:03] [SPEAKER_02]: So it's a way to really close the gap on books that have no audio book component.

[00:46:07] [SPEAKER_02]: Also a way for authors to potentially, you know,

[00:46:11] [SPEAKER_02]: line up the voices of some of their favorite narrators without those narrators

[00:46:16] [SPEAKER_02]: being like, Oh, I don't, you know, no, I've, I call this much.

[00:46:20] [SPEAKER_02]: My wonderful dulcet tones, you know, I, I am very selective.

[00:46:25] [SPEAKER_02]: They still have those, those audio book narrators still have the, the,

[00:46:30] [SPEAKER_02]: you know, the choice to either accept or decline or whatever.

[00:46:33] [SPEAKER_02]: But I guess they just don't have to go through the work, the time, which,

[00:46:37] [SPEAKER_02]: you know, you would know better than I,

[00:46:39] [SPEAKER_02]: I'm sure it takes a lot of time and effort and so good.

[00:46:43] [SPEAKER_04]: I've got a few things here. Good. Repentance took me,

[00:46:47] [SPEAKER_04]: they scheduled four days. I did it in three days. It was three hard days,

[00:46:51] [SPEAKER_04]: hard days. Yeah.

[00:46:53] [SPEAKER_04]: Do you keep your voice going that whole time? Like I would imagine that's part

[00:46:57] [SPEAKER_04]: of it. Plus the other thing that happens.

[00:46:58] [SPEAKER_04]: So I do this podcast right after I've had lunch cause that's what time it is

[00:47:01] [SPEAKER_04]: here. And I tend to be a little, pardon me for this more sharing.

[00:47:05] [SPEAKER_04]: You want a little Flemme here for this show that I am for twig because the time

[00:47:10] [SPEAKER_04]: of the day I couldn't figure out why am I always kind of just yucky and I realized

[00:47:13] [SPEAKER_04]: I just had lunch. Um, uh, so yeah, the different times of day,

[00:47:18] [SPEAKER_04]: which you can hear off and as for you hear it,

[00:47:19] [SPEAKER_04]: especially if you listen to audio book and somebody had to come in later to do

[00:47:22] [SPEAKER_04]: an insert.

[00:47:23] [SPEAKER_02]: Oh, you definitely hear, you hear the tone.

[00:47:25] [SPEAKER_02]: Sometimes the energetic miss of them shifts and I'm like, Oh, that,

[00:47:30] [SPEAKER_02]: this is a new recording. I can totally tell.

[00:47:32] [SPEAKER_04]: So I'm in multiple minds about this.

[00:47:33] [SPEAKER_04]: It took me forever to get them to agree to do the Gutenberg parenthesis in

[00:47:39] [SPEAKER_04]: audio book. Pardon me there. I am doing it.

[00:47:42] [SPEAKER_04]: And um, I wanted to do it myself and I did it myself.

[00:47:47] [SPEAKER_04]: The web we weave, which is out next month, still no deal on an audio book.

[00:47:51] [SPEAKER_04]: Just pissing me off. On the other hand, if this existed in easy form,

[00:47:56] [SPEAKER_04]: um, on the one hand I'm thinking, okay,

[00:47:58] [SPEAKER_04]: if it made the audio book more possible and we'd sell more books and more people

[00:48:01] [SPEAKER_04]: would hear it. Okay, do this. On the other hand,

[00:48:05] [SPEAKER_04]: if you think it's me and it's not really my intonation of the meaning,

[00:48:12] [SPEAKER_04]: um, no, I don't like that. If it's me,

[00:48:15] [SPEAKER_04]: I gotta be me. I have to be me. Um,

[00:48:19] [SPEAKER_04]: and so, you know, as a, as an author and an audio book

[00:48:25] [SPEAKER_04]: reader, uh, I'm not sure what I think about this.

[00:48:28] [SPEAKER_02]: Yeah. Yeah. I mean it does,

[00:48:31] [SPEAKER_02]: it does give a narrator another path with which to,

[00:48:36] [SPEAKER_02]: you know, make, continue to make a living, you know, cause time is money.

[00:48:41] [SPEAKER_02]: And as you say, reading one of these books takes a lot,

[00:48:46] [SPEAKER_02]: you know, you've got a, you've got to carve out the time for it.

[00:48:49] [SPEAKER_02]: It takes a lot of effort versus, you know,

[00:48:52] [SPEAKER_02]: is a narrator in a position where they can be like, you know,

[00:48:56] [SPEAKER_02]: the really big major projects I'm willing to do in person,

[00:49:00] [SPEAKER_02]: but I'm also willing to make kind of like a side,

[00:49:03] [SPEAKER_02]: almost like a side hustle income or a separate tangent income of my business on

[00:49:09] [SPEAKER_02]: these smaller ones that I'm happy to have my voice lent to,

[00:49:12] [SPEAKER_02]: that I don't actually have to put in the time and the work for.

[00:49:15] [SPEAKER_02]: Does that though impact the value of the work that they do get paid for on a

[00:49:22] [SPEAKER_02]: bigger scale?

[00:49:22] [SPEAKER_04]: There were reputations. So when I did record the book,

[00:49:24] [SPEAKER_04]: I recorded a studio in New York and there were professional audio book

[00:49:27] [SPEAKER_04]: narrators there, right? And part of what this story is, well, they could expand,

[00:49:31] [SPEAKER_04]: they could scale, right? They could loan their,

[00:49:34] [SPEAKER_04]: because there are some that are really well known now.

[00:49:36] [SPEAKER_04]: They could loan their voice to that. But once again,

[00:49:38] [SPEAKER_04]: if I'm listening to the book made by AI and it just doesn't have the same

[00:49:42] [SPEAKER_04]: je ne sais quoi, then that affects the narrator's brand.

[00:49:47] [SPEAKER_02]: Yes, it does. So absolutely. Yeah. And I haven't,

[00:49:50] [SPEAKER_02]: and I haven't really heard an AI voice like replica or whatever that has me sold

[00:49:57] [SPEAKER_02]: 100%. You know what I mean?

[00:49:59] [SPEAKER_02]: Like there's always something that becomes evident.

[00:50:02] [SPEAKER_02]: Either it's a mispronounced word or it's the fact that their intonation and

[00:50:08] [SPEAKER_02]: inflection is always kept in the same place and you never kind of steers out of

[00:50:13] [SPEAKER_02]: that point. And you know, at a certain point,

[00:50:15] [SPEAKER_02]: like I could not listen to a 12 hour book like that.

[00:50:17] [SPEAKER_02]: Like it would really get grating at a certain point.

[00:50:20] [SPEAKER_02]: So the talents are still there.

[00:50:22] [SPEAKER_04]: Oh, they absolutely are. I went to a presentation at BDMI,

[00:50:25] [SPEAKER_04]: which is the barrel spin investment arm by now, probably a year ago.

[00:50:30] [SPEAKER_04]: And there was a company there that had a AI voice,

[00:50:34] [SPEAKER_04]: but it also had the ability to control variation in that.

[00:50:40] [SPEAKER_04]: And so what I speculated about at the time is I can see the day pretty soon

[00:50:45] [SPEAKER_04]: where you end up with a voice markup language. Oh yeah. Right.

[00:50:50] [SPEAKER_04]: So, so it's like punctuation and bold and italic. I think we have. Yeah. Yeah.

[00:50:54] [SPEAKER_04]: And so I could imagine that if, all right, this is,

[00:50:57] [SPEAKER_04]: I can then edit the AI of my voice. That's a joke.

[00:51:03] [SPEAKER_04]: Pause here. Pause is pretty easy, but, but I could you know, get excited,

[00:51:08] [SPEAKER_04]: get sad. Are there ways that you could,

[00:51:10] [SPEAKER_04]: you could change the markup of how the voice is going to operate? Also,

[00:51:15] [SPEAKER_04]: by the way, teaching the AI,

[00:51:18] [SPEAKER_04]: the relationships of certain things about the structure of the language to the

[00:51:22] [SPEAKER_04]: intonation of it would make the whole of AI voice better.

[00:51:28] [SPEAKER_04]: Yeah. But to do that, you've got to give me the power to fix it and train it.

[00:51:34] [SPEAKER_01]: Yeah. You're like, you need to give me a, you need to give me a language for doing

[00:51:37] [SPEAKER_01]: that. Yeah. Oh, make you excited and really dull.

[00:51:43] [SPEAKER_01]: Build up the intensity and oh sarcasm.

[00:51:48] [SPEAKER_02]: Totally. It's like a voice conductor. Super interesting. Yeah.

[00:51:52] [SPEAKER_02]: I wonder if that's the direction,

[00:51:54] [SPEAKER_02]: although that sounds kind of tedious in its own right. It does. But if,

[00:51:58] [SPEAKER_04]: if it's sort of having to read the book for three or four days,

[00:52:01] [SPEAKER_04]: I could spend half a day editing it going through this. That's fine.

[00:52:05] [SPEAKER_04]: That's fine. No, that was a little bit off. How do I tweak that?

[00:52:08] [SPEAKER_04]: And if I had the tools to tweak it, which again would,

[00:52:10] [SPEAKER_04]: would mean a markup language then I could see it being at least an

[00:52:15] [SPEAKER_02]: improvement. Yeah. Yeah, totally. To what we have already.

[00:52:20] [SPEAKER_02]: Okay. This, this kind of relates in a strange way. You had put in,

[00:52:25] [SPEAKER_02]: excuse me, you put in a link to a new site. I don't know if it's new,

[00:52:29] [SPEAKER_02]: but it's new to me. A site called icon.me,

[00:52:33] [SPEAKER_02]: icon.me and this is just really,

[00:52:37] [SPEAKER_02]: really weird. I'm super curious to hear what you think about this. Quote,

[00:52:41] [SPEAKER_02]: create winning influencer ads with AI.

[00:52:45] [SPEAKER_02]: You can partner with creators and turn one video into 20 videos with AI,

[00:52:51] [SPEAKER_02]: AB testing messaging and find winning ads they say.

[00:52:56] [SPEAKER_02]: So the creator offers a simple video for the sponsor. So, you know,

[00:53:01] [SPEAKER_02]: in this case, like this woman is holding up a, a, a, a green,

[00:53:05] [SPEAKER_02]: a super foods container from a company called bloom.

[00:53:09] [SPEAKER_02]: And I think this is actually an AI generated version.

[00:53:13] [SPEAKER_02]: Let me see if we've got audio here.

[00:53:15] [SPEAKER_00]: Hey, it's Alex as a pro in chess and poker. I need to stay focused.

[00:53:20] [SPEAKER_00]: When my energy is low, I turn to bloom nutrition's greens.

[00:53:23] [SPEAKER_02]: They help keep my energy. Okay. And so, you know, it's an AI generated that,

[00:53:28] [SPEAKER_02]: that I believe is the output. That is one of the AI generated outputs.

[00:53:32] [SPEAKER_02]: But is it based on a real Alex? It is based on a real video.

[00:53:37] [SPEAKER_02]: So she, as the influencer would upload her one take,

[00:53:42] [SPEAKER_02]: you know, and I'm imagining the,

[00:53:44] [SPEAKER_02]: the script and all that kind of stuff might,

[00:53:46] [SPEAKER_02]: might come from the company or be approved through them or whatever.

[00:53:50] [SPEAKER_02]: She would upload her video and then that would then be turned into 20 different

[00:53:56] [SPEAKER_02]: videos that are based on, on the video that she uploaded as the source.

[00:54:01] [SPEAKER_02]: And so you end up as this as a potential sponsor,

[00:54:05] [SPEAKER_02]: you end up with all these options. I thought this was funny.

[00:54:08] [SPEAKER_02]: Turn one video into 20 videos with AI,

[00:54:10] [SPEAKER_02]: no more waiting weeks for creators to make videos.

[00:54:15] [SPEAKER_04]: Pesky humans.

[00:54:17] [SPEAKER_02]: They need all that time. Creators are just so lazy.

[00:54:21] [SPEAKER_02]: They take weeks to create the videos. Um, anyways, you know,

[00:54:25] [SPEAKER_02]: like which really, what it really enables for Jitnish is AB testing.

[00:54:30] [SPEAKER_04]: Yeah, that's true. That's true.

[00:54:32] [SPEAKER_04]: It's just, you can try different things and see what the conversion is,

[00:54:35] [SPEAKER_04]: without making the creator have to create 20 videos. So in that sense,

[00:54:40] [SPEAKER_04]: in an otherwise obnoxious field of, of, of influencers, um,

[00:54:44] [SPEAKER_04]: I see the sense in it. But once again, I,

[00:54:48] [SPEAKER_04]: it just makes all those stuff less believable.

[00:54:52] [SPEAKER_02]: Yeah, it's funny. I'm watching this one and if you watch his hands,

[00:54:55] [SPEAKER_02]: his hands get a little weird at times.

[00:54:57] [SPEAKER_02]: They'd like sprout an extra finger and then a finger goes away right there.

[00:55:00] [SPEAKER_02]: Oh, yep. That was a little weird. Um, yeah, I mean what,

[00:55:05] [SPEAKER_02]: what comes up for me on this really ties into, you know,

[00:55:08] [SPEAKER_02]: the audible story that we were just talking about,

[00:55:09] [SPEAKER_02]: which is trust and authenticity. Um, you know,

[00:55:14] [SPEAKER_02]: if there is any tell whatsoever that this is AI generated,

[00:55:18] [SPEAKER_02]: how does the viewer actually feel about that? Or even beyond that,

[00:55:22] [SPEAKER_02]: are we moving into a time where people don't care as much,

[00:55:26] [SPEAKER_02]: where this becomes so,

[00:55:28] [SPEAKER_02]: I want my influence to be influenced. I don't know. I don't know.

[00:55:32] [SPEAKER_02]: You know, people might just be like, yeah,

[00:55:34] [SPEAKER_02]: that's how influencers make money or that's how audio book narrators make

[00:55:38] [SPEAKER_02]: money. And that's fine. Whatever. You got to make the bills.

[00:55:41] [SPEAKER_04]: But at some point if somebody is recommending something to you,

[00:55:45] [SPEAKER_04]: you want to feel that there's some authenticity there.

[00:55:47] [SPEAKER_04]: I think that's super important. I totally agree. Right. I totally agree.

[00:55:51] [SPEAKER_04]: I attended a world economic forum, uh, Davos, uh, AI governance meeting yesterday,

[00:55:56] [SPEAKER_04]: uh, virtually. And, um,

[00:56:00] [SPEAKER_04]: it was about implementation and ethics of AI and all kinds of questions.

[00:56:04] [SPEAKER_04]: And it was interesting.

[00:56:05] [SPEAKER_04]: It was some corporate people there and they were talking about their structure

[00:56:09] [SPEAKER_04]: and how they have an AI committee and how they have approval of rules and how

[00:56:12] [SPEAKER_04]: they have this and how they have that. And there was a professor there, uh,

[00:56:17] [SPEAKER_04]: who's really smart and I've talked to before and she's kind of in the middle of

[00:56:21] [SPEAKER_04]: it. She said, it sounds like all you're trying to do is check boxes.

[00:56:24] [SPEAKER_04]: You're trying to, you're trying to create a, um, you know, cover your ass,

[00:56:29] [SPEAKER_04]: um, risk abatement, uh,

[00:56:34] [SPEAKER_04]: structure rather than trying to implement ethics.

[00:56:37] [SPEAKER_04]: And so the question for all this info, I mean,

[00:56:39] [SPEAKER_04]: influencers how ethical is it as a field? I think, I think people, you know,

[00:56:44] [SPEAKER_04]: when you do a commercial that you care about,

[00:56:45] [SPEAKER_04]: you're only going to do things you actually care about. Leo all the time.

[00:56:48] [SPEAKER_04]: He says, I won't, I won't advertise certain things. Yes,

[00:56:50] [SPEAKER_04]: there are ethics in here. And so how this is implemented in what ways? Uh,

[00:56:55] [SPEAKER_04]: now it says here that the influencers can approve all the videos made of them.

[00:57:01] [SPEAKER_04]: Right. So if you put words in Jason's mouth saying,

[00:57:05] [SPEAKER_04]: I prefer Apple phones to Android, you know, it's not Jason.

[00:57:09] [SPEAKER_04]: I know and love, uh, and Jason could disapprove that. Um,

[00:57:14] [SPEAKER_04]: so I get it. It makes life easier. You can do a B testing, but yeah,

[00:57:19] [SPEAKER_04]: we both, I think when I put this in the run down,

[00:57:21] [SPEAKER_04]: you and I both have the same reaction of, uh, are we sure about this? Yeah.

[00:57:26] [SPEAKER_02]: Yeah. The authenticity piece is the thing that really screams to me because I

[00:57:29] [SPEAKER_02]: mean, at the end of the day, I feel like what,

[00:57:32] [SPEAKER_02]: what really drives the influencer kind of world and the influencer economy is

[00:57:37] [SPEAKER_02]: that direct connection to person and that authentic kind of, you know,

[00:57:42] [SPEAKER_02]: the, the,

[00:57:43] [SPEAKER_02]: the fact that a lot of people are watching some,

[00:57:45] [SPEAKER_02]: some of these influencers and whether it's true or not,

[00:57:49] [SPEAKER_02]: believe that what they see is that person's authentic life,

[00:57:53] [SPEAKER_02]: that they are living an authentic life.

[00:57:55] [SPEAKER_02]: And you start getting into AI generated and,

[00:57:58] [SPEAKER_02]: and start to see the tells of, of, Oh, well,

[00:58:01] [SPEAKER_02]: that was a little weird what happened to the fingers there. Or,

[00:58:03] [SPEAKER_02]: or there's that like weird kind of star, you know, a distant gaze that,

[00:58:08] [SPEAKER_02]: uh, some of these things tend to tend to give, they're getting better there,

[00:58:12] [SPEAKER_02]: but you know, these things kind of break that,

[00:58:15] [SPEAKER_02]: that level of authenticity and connection from a human level.

[00:58:18] [SPEAKER_04]: Yeah. Well, AI has already ruined online reviews.

[00:58:22] [SPEAKER_04]: You cannot be assured that any review,

[00:58:25] [SPEAKER_04]: whether it's on a website or whether it's an Amazon review or anywhere,

[00:58:28] [SPEAKER_04]: they are generated in huge number now by AI.

[00:58:32] [SPEAKER_04]: And so the authenticity of, of, of consumer reviews,

[00:58:35] [SPEAKER_04]: which was our best hope for replacing the likes of a consumer reports is now

[00:58:40] [SPEAKER_04]: ruined. Thank you. AI.

[00:58:43] [SPEAKER_02]: Thanks AI. Thanks for all the fish. Um,

[00:58:47] [SPEAKER_02]: this week in AI getting it wrong.

[00:58:50] [SPEAKER_02]: If we've had a couple of those probably today already,

[00:58:53] [SPEAKER_02]: but ESPN did an AI generated recap of a soccer game.

[00:58:57] [SPEAKER_02]: This just caught my attention and, and uh,

[00:58:59] [SPEAKER_02]: I thought it was interesting between North Carolina courage and San Diego wave

[00:59:03] [SPEAKER_02]: FC two teams. I mean, this is a sport that I do not watch personally.

[00:59:06] [SPEAKER_02]: So my apologies,

[00:59:07] [SPEAKER_02]: but they did an AI generated recap of this as ESPN apparently does on some of,

[00:59:13] [SPEAKER_02]: you know, some games from different divisions around the,

[00:59:15] [SPEAKER_02]: around the world and around the country.

[00:59:16] [SPEAKER_02]: The original version of the recap cause it's been corrected since, uh,

[00:59:20] [SPEAKER_02]: was around 215 words listed all of the pertinent details of the game except for

[00:59:26] [SPEAKER_02]: the reason most people saw it as important, which was it was,

[00:59:29] [SPEAKER_02]: it had the two time world cup winner and Olympic gold medalist,

[00:59:32] [SPEAKER_02]: Alex Morgan, they're playing her last professional match. Uh, you know,

[00:59:37] [SPEAKER_02]: it had a tearful goodbye moment with the stadium chanting her name and her

[00:59:42] [SPEAKER_02]: wishing the crowd and her team goodbye and wishing everyone well and everything.

[00:59:46] [SPEAKER_02]: It was this very kind of like emotional moment for this game and for fans,

[00:59:50] [SPEAKER_02]: my understanding, a really big reason to actually watch the game.

[00:59:53] [SPEAKER_02]: Of course the AI generated recap didn't know any of that or didn't, didn't,

[00:59:58] [SPEAKER_02]: you know,

[00:59:59] [SPEAKER_02]: I had no way to know understanding.

[01:00:00] [SPEAKER_04]: And some of the characters in the machine had no, it has no news judgment.

[01:00:04] [SPEAKER_04]: It doesn't know it's another damn game. Okay. I'll give it to you.

[01:00:07] [SPEAKER_02]: Yeah. So I'll give you the score and I'll give you the highlights of the points

[01:00:11] [SPEAKER_02]: made and everything like that.

[01:00:12] [SPEAKER_02]: I think her name wasn't even mentioned at all in the recap initially.

[01:00:16] [SPEAKER_02]: Now it's there. Like I checked the article and of course they have a little blurb

[01:00:20] [SPEAKER_02]: there, which, you know,

[01:00:21] [SPEAKER_02]: I'm assuming one of the editors typed up or maybe they fed it more information

[01:00:25] [SPEAKER_02]: and had it regenerate. But anyways, I thought that was interesting.

[01:00:29] [SPEAKER_02]: An interesting example of how, you know, these things, like you just said,

[01:00:33] [SPEAKER_02]: they don't,

[01:00:33] [SPEAKER_02]: they're not going to automatically know or infer all of the same things that a

[01:00:38] [SPEAKER_02]: human, you know, might without the,

[01:00:40] [SPEAKER_02]: without the pertinent information in the dataset there to begin with,

[01:00:44] [SPEAKER_02]: it's not just going to automatically know this stuff.

[01:00:46] [SPEAKER_02]: It's not part of the oxygen.

[01:00:48] [SPEAKER_04]: So not unless a human tells them or unless it happens to read 20 stories online

[01:00:52] [SPEAKER_04]: and recognize it. Right, right. Yeah. Right. It has to have that.

[01:00:57] [SPEAKER_02]: Yeah, exactly that. And finally, uh,

[01:01:01] [SPEAKER_02]: because apparently we end this show with robots just kind of,

[01:01:05] [SPEAKER_02]: they keep happening. Robots keep happening.

[01:01:07] [SPEAKER_02]: Apparently the show's going to be AI and robots eventually, but a new one,

[01:01:13] [SPEAKER_02]: Isaac is its name by weave robots. And, uh,

[01:01:17] [SPEAKER_02]: this just doesn't even look real right now,

[01:01:19] [SPEAKER_02]: but a personal robot made in California expected to ship to first

[01:01:24] [SPEAKER_02]: consumers in fall of 2025. So not that long, you know,

[01:01:28] [SPEAKER_02]: in a year from now and you know, and once again, I'll believe it when I see it,

[01:01:32] [SPEAKER_02]: this thing is tall. It's not humanoid. It definitely looks like a robot.

[01:01:36] [SPEAKER_02]: It's meant to accomplish home tasks.

[01:01:40] [SPEAKER_02]: So things like folding laundry, tidying, organizing spaces,

[01:01:45] [SPEAKER_02]: bringing you glasses of wine while you sit on the couch, uh,

[01:01:49] [SPEAKER_02]: you know, really important things like that. Um, can take, Oh,

[01:01:54] [SPEAKER_02]: bring you over a cup of coffee, um,

[01:01:57] [SPEAKER_02]: has voice or text commands that you can issue to it.

[01:02:00] [SPEAKER_02]: The camera that's on the device actually folds away when it's not used. So,

[01:02:04] [SPEAKER_02]: you know, the robot isn't watching you and it costs $59,000,

[01:02:10] [SPEAKER_02]: uh, to the first 30 customers.

[01:02:14] [SPEAKER_04]: There you go. You know, I looked at it as a toy for very rich people who can do

[01:02:18] [SPEAKER_04]: this. I think limited, um, abilities,

[01:02:21] [SPEAKER_04]: the kind of variety of tasks you do, it'd be very hard for just all of them.

[01:02:26] [SPEAKER_04]: But what really hit me was for accessibility. If someone were handicapped,

[01:02:32] [SPEAKER_04]: um, and can't do some of these tasks, then I can see real value in this,

[01:02:37] [SPEAKER_04]: but there's not a big business in that because people who are handicapped don't

[01:02:40] [SPEAKER_04]: have as much money. And that's the shame. That's the,

[01:02:44] [SPEAKER_04]: the highest utility I could imagine would be for somebody who has limited

[01:02:47] [SPEAKER_04]: mobility, limited site. Um, right. Um,

[01:02:53] [SPEAKER_04]: limited, um, you know, things like Parkinson's, uh,

[01:02:57] [SPEAKER_04]: limited surety of movement and that's who should be using this,

[01:03:01] [SPEAKER_04]: but insurance is gonna pay for it. And so this isn't going to do the good that

[01:03:06] [SPEAKER_04]: it could do.

[01:03:07] [SPEAKER_02]: Yeah. I mean the question that I have in watching the examples that the video

[01:03:12] [SPEAKER_02]: kind of gave, you know, the bringing over the cup of coffee,

[01:03:16] [SPEAKER_02]: bringing over the glasses of wine to the couple that's sitting on the couch.

[01:03:20] [SPEAKER_02]: Is the robot just a deliverer or did the robot actually prepare any of that?

[01:03:25] [SPEAKER_02]: You know, like I don't make it myself a glass of wine. Yeah.

[01:03:29] [SPEAKER_02]: That glass of wine comes with me.

[01:03:30] [SPEAKER_02]: I don't just like leave it on the counter and go sit down on the couch and wait

[01:03:33] [SPEAKER_02]: for someone to bring it to me. You know, it's a little weird there.

[01:03:37] [SPEAKER_02]: Like the cup of coffee, it's just sitting on the countertop.

[01:03:39] [SPEAKER_02]: Did a human make the cup of coffee and then set it there and be like, all right,

[01:03:42] [SPEAKER_02]: now I'm going to ask my robot to bring it to me. You know? So that, I don't know.

[01:03:46] [SPEAKER_02]: I could be wrong.

[01:03:47] [SPEAKER_02]: It could actually make you coffee or pour you a glass of wine. But uh,

[01:03:52] [SPEAKER_02]: interesting nonetheless.

[01:03:53] [SPEAKER_04]: I can buy a lot of really good wine for $60,000 though and a good cork screw.

[01:04:00] [SPEAKER_02]: That's true. That's true.

[01:04:02] [SPEAKER_02]: Do you want to potentially sacrifice that,

[01:04:06] [SPEAKER_02]: that liquid gold that's in that glass by asking your robot to deliver it to

[01:04:10] [SPEAKER_02]: you? Uh, I don't know. I don't know that you do.

[01:04:13] [SPEAKER_04]: Now all I can afford is for it to bring me Mountain Dew. Yeah. Bad choice.

[01:04:17] [SPEAKER_04]: Right. Just hopefully it doesn't crush the can and the process.

[01:04:22] [SPEAKER_02]: Well, that's it. We've reached the end of this episode of AI inside.

[01:04:25] [SPEAKER_02]: A lot of fun talking about this week in the world of artificial intelligence.

[01:04:29] [SPEAKER_04]: You have so many watched live next week.

[01:04:34] [SPEAKER_04]: We're going to be, yeah, so that's right. That's right.

[01:04:37] [SPEAKER_02]: We are going to be going early.

[01:04:38] [SPEAKER_02]: I have a trip to New York for another podcast that I do Android faithful.

[01:04:43] [SPEAKER_02]: Uh, we're going to droid con in New York Wednesday through Friday.

[01:04:46] [SPEAKER_02]: So our normal recording of this show is Wednesdays at 10 AM Pacific 1 PM

[01:04:51] [SPEAKER_02]: Eastern. Next week, however, we're going to be doing it and Tuesday.

[01:04:57] [SPEAKER_02]: Uh, what, so 11, 24, it's the, what was that? September 17th,

[01:05:01] [SPEAKER_02]: Tuesday, September 17th at 9 30 AM Pacific 12 30 PM Eastern.

[01:05:07] [SPEAKER_02]: So, uh, you know, if you just go to the textbook,

[01:05:10] [SPEAKER_02]: if you want to watch it live, go to the tech splitter, YouTube channel,

[01:05:13] [SPEAKER_02]: youtube.com slash at tech splitter.

[01:05:17] [SPEAKER_02]: And I will have on there a little placeholder for the live event so you can,

[01:05:22] [SPEAKER_02]: you can add it or, you know, many of you are watching live through Twitter,

[01:05:27] [SPEAKER_02]: through, through Jeff's kind of rebroadcast of this through Twitter.

[01:05:30] [SPEAKER_02]: So you'll just see it in Twitter where I'm sure you live much of your life as

[01:05:34] [SPEAKER_02]: you're doing your work day anyway. So, so just look for that. But yes,

[01:05:39] [SPEAKER_02]: different day. And I think the podcast is going to publish on a Tuesday.

[01:05:42] [SPEAKER_02]: There's no reason to hold it for Wednesday.

[01:05:43] [SPEAKER_02]: So I'll just publish it on the Tuesday once, but once it's all recorded,

[01:05:46] [SPEAKER_02]: right. Um, and then of course, Gutenberg parenthesis.com.

[01:05:51] [SPEAKER_02]: If you want to find all things, Jeff Jarvis for now,

[01:05:54] [SPEAKER_04]: and my page isn't up yet, but if you want the web, we weave, uh,

[01:05:58] [SPEAKER_04]: search for that and go to the Heshet basic book site and use the code web 20

[01:06:03] [SPEAKER_04]: cap WB 20 for a 20% discount.

[01:06:09] [SPEAKER_02]: Excellent. Definitely do that.

[01:06:11] [SPEAKER_02]: Looking forward to the day when it'll be on your page so I can show it off and

[01:06:14] [SPEAKER_02]: get my son. Yeah. And, and,

[01:06:15] [SPEAKER_02]: and flash it on screen for everybody to see with their eyes. Um,

[01:06:20] [SPEAKER_02]: and then finally, uh, we've, we've said it once before, we'll say it again.

[01:06:23] [SPEAKER_02]: You can support us on Patreon. And if you do that,

[01:06:26] [SPEAKER_02]: you get a lot of extra stuff in the process and not to mention our,

[01:06:30] [SPEAKER_02]: our appreciation and gratitude. Uh,

[01:06:34] [SPEAKER_02]: it's patreon.com slash AI inside show.

[01:06:37] [SPEAKER_02]: We have ad free shows early access to AI related videos on the textbook or

[01:06:41] [SPEAKER_02]: YouTube channel, discord community, uh,

[01:06:43] [SPEAKER_02]: hangouts with the community with Jeff and I at times.

[01:06:47] [SPEAKER_02]: And you can also get an AI inside t-shirt if you become an executive producer

[01:06:52] [SPEAKER_02]: of this show, like Dr. Do Jeffrey Marachini WPVM 103.7 in Asheville,

[01:06:58] [SPEAKER_02]: North Carolina and Paul Langs. We've got four.

[01:07:03] [SPEAKER_02]: Yay.

[01:07:03] [SPEAKER_02]: On the show bosses told us what to do. Jeez. I know. And we,

[01:07:07] [SPEAKER_02]: and we'd love to have more. I'd love to be able to pull out my thumb.

[01:07:10] [SPEAKER_02]: We want to show off how many we have. So you know,

[01:07:14] [SPEAKER_02]: patreon.com slash AI inside show and uh,

[01:07:18] [SPEAKER_02]: throw your support and we really do appreciate when you do that.

[01:07:21] [SPEAKER_02]: So thank you very much.

[01:07:23] [SPEAKER_02]: Thank you everybody for watching and listening to this show.

[01:07:26] [SPEAKER_02]: Find everything you need to know at AI inside dot show.

[01:07:29] [SPEAKER_02]: That is the single place on the web where everything that you need to know can

[01:07:33] [SPEAKER_02]: be found. And uh, we appreciate you.

[01:07:35] [SPEAKER_02]: We'll see you next time on another episode of AI inside. Bye everybody.