Connected

00:00:07 ◼   ► Hello and welcome to Connected episode 62.

00:00:13 ◼   ► It's made possible by our sponsor sponsors Scribe and Century.

00:00:17 ◼   ► I am your annual chairman, Stephen Hackett, and I have the pleasure of being

00:00:21 ◼   ► joined by Mike Hurley.

00:00:23 ◼   ► Hello, it's me, and we also both have the pleasure of being joined by

00:00:27 ◼   ► the keynote chairman himself, Veter Federico Vicici.

00:00:29 ◼   ► Federico, do you have any middle names?

00:00:32 ◼   ► We've talked about this before many, many years ago.

00:00:35 ◼   ► I'm sure.

00:00:36 ◼   ► I I do.

00:00:37 ◼   ► They're kind of embarrassing.

00:00:39 ◼   ► What are they?

00:00:40 ◼   ► Well it's the the after So th they were so middle names are not

00:00:45 ◼   ► really important in Italy.

00:00:47 ◼   ► Okay.

00:00:48 ◼   ► Right?

00:00:48 ◼   ► You just r register them when when your child is born.

00:00:54 ◼   ► Um they were uh chosen after my grandfather.

00:00:58 ◼   ► Um, it's Dante and Livio.

00:01:01 ◼   ► Livio?

00:01:02 ◼   ► Yes.

00:01:03 ◼   ► Dante's pretty cool though.

00:01:04 ◼   ► I don't know why you're embarrassed about that.

00:01:07 ◼   ► Well I don't know.

00:01:08 ◼   ► I don't it's it's you know.

00:01:10 ◼   ► Nobody's called Dante anymore.

00:01:14 ◼   ► Incredibly Kate in the Discord wrote down.

00:01:17 ◼   ► Remembered, yes.

00:01:18 ◼   ► How do people it's like I know I know this because I'm also a

00:01:22 ◼   ► podcast listener, right?

00:01:23 ◼   ► That I am very aware of things and remember things about the podcast hosts

00:01:28 ◼   ► that I listen to.

00:01:30 ◼   ► But I also I don't know how me as the podcaster cannot remember the

00:01:35 ◼   ► things that the listeners can remember.

00:01:37 ◼   ► Yeah.

00:01:38 ◼   ► You know?

00:01:38 ◼   ► It's it's incredible, isn't it?

00:01:40 ◼   ► Yeah, it's it's it's very difficult.

00:01:42 ◼   ► Um speaking of things that are difficult, I have a quick story before follow-up.

00:01:46 ◼   ► All right and I just wanted to put it here instead of uh later

00:01:50 ◼   ► because I want to get your like rawest reactions, so Is it embarrassing?

00:01:56 ◼   ► I have a high school daughter and homecoming is coming up.

00:02:00 ◼   ► Big fall dance.

00:02:02 ◼   ► She's gonna go.

00:02:04 ◼   ► And she's gonna go to a party afterwards with some friends and she was

00:02:07 ◼   ► telling us about this asking if she could go.

00:02:09 ◼   ► And uh they have a theme for the party.

00:02:12 ◼   ► It's a retro party.

00:02:14 ◼   ► I love I love where this is going.

00:02:15 ◼   ► Okay.

00:02:16 ◼   ► Uh the year that they picked to dress up for and to remember is

00:02:21 ◼   ► the year twenty sixteen.

00:02:24 ◼   ► What?

00:02:27 ◼   ► Ten years ago.

00:02:28 ◼   ► Why would you do this?

00:02:29 ◼   ► What?

00:02:30 ◼   ► Yeah.

00:02:31 ◼   ► That's not a year to remember.

00:02:32 ◼   ► It's like yesterday.

00:02:34 ◼   ► And I I that's what I said.

00:02:35 ◼   ► And then I asked, What

00:02:39 ◼   ► what does twenty sixteen mean?

00:02:41 ◼   ► Like what do you and what uh they were like ah the good old

00:02:45 ◼   ► days.

00:02:46 ◼   ► Yeah, that's right.

00:02:47 ◼   ► Uh but their view of twenty sixteen is basically how I think about like

00:02:52 ◼   ► two thousand eight or nine.

00:02:54 ◼   ► It's like skinny jeans and you know

00:02:58 ◼   ► basically described like sort of emo guys with bags kind of scene.

00:03:04 ◼   ► Which I mean some of us were screen.

00:03:06 ◼   ► Guys with bags?

00:03:07 ◼   ► What does that mean?

00:03:08 ◼   ► You know what I mean?

00:03:08 ◼   ► Like we were I don't know what you mean.

00:03:10 ◼   ► Bags?

00:03:11 ◼   ► Like bags to wear?

00:03:13 ◼   ► Like bags under their eyes?

00:03:14 ◼   ► Like you know, we all have like, you know, main purses.

00:03:18 ◼   ► Maybe just me.

00:03:19 ◼   ► What?

00:03:20 ◼   ► Just me?

00:03:20 ◼   ► What?

00:03:21 ◼   ► What were emote guys like in America Steven?

00:03:24 ◼   ► I don't remember having like a bag phase.

00:03:27 ◼   ► Okay, well I did.

00:03:28 ◼   ► Um have you ever left your bag face?

00:03:31 ◼   ► I think you're still very much in your bag face.

00:03:33 ◼   ► I do love a good bag.

00:03:35 ◼   ► I'm just googling it.

00:03:36 ◼   ► But man.

00:03:37 ◼   ► Oh, I see what you mean.

00:03:39 ◼   ► I see what you mean.

00:03:41 ◼   ► Okay.

00:03:42 ◼   ► Like with we're hmm.

00:03:44 ◼   ► Let me Google it too.

00:03:45 ◼   ► Emo guys with bags.

00:03:47 ◼   ► Yeah, you gotta Google emo guys.

00:03:49 ◼   ► I'm gonna see if I can find a picture of myself.

00:03:51 ◼   ► Uh uh, I mean this is just emo guys with bags.

00:03:54 ◼   ► This doesn't particularly suggest to me any kind of trend.

00:03:57 ◼   ► It's just like I think it Maybe it was just me and my sad

00:04:00 ◼   ► friends, but

00:04:02 ◼   ► the the the the the the twenty sixteen part, forget what they think twenty

00:04:07 ◼   ► sixteen was, like, but man, I I have never turned to dust more quickly

00:04:13 ◼   ► in a conversation.

00:04:15 ◼   ► Just absolutely brutal.

00:04:17 ◼   ► It doesn't it just I don't know.

00:04:20 ◼   ► I'm like I'm surprised that the the party would be a year that they

00:04:27 ◼   ► were alive in.

00:04:28 ◼   ► Yeah.

00:04:29 ◼   ► I would figure you would go like to a time that you d you

00:04:33 ◼   ► just didn't know.

00:04:34 ◼   ► You know, like we're gonna do a nineties party or like we're gonna do

00:04:38 ◼   ► a two thousand and one party, like a twenty sixteen party.

00:04:42 ◼   ► That is a surprise to me that that's the year they would that they

00:04:45 ◼   ► would go for.

00:04:46 ◼   ► I mean, look, for your daughter's age, ten years ago is is like a

00:04:52 ◼   ► very long time.

00:04:52 ◼   ► Like I I understand that.

00:04:54 ◼   ► But I'm just I'm just surprised that it's like a year that she could

00:04:57 ◼   ► remember or would remember is the one that they would that they would go

00:05:02 ◼   ► for.

00:05:02 ◼   ► But yeah.

00:05:03 ◼   ► Kids today, am I right?

00:05:05 ◼   ► I know.

00:05:05 ◼   ► I found a picture of myself in two thousand nine.

00:05:08 ◼   ► You got a bag?

00:05:09 ◼   ► I got a bag.

00:05:10 ◼   ► Now look Discord.

00:05:12 ◼   ► I'm gonna put this where y'all can see it.

00:05:15 ◼   ► Uh-huh.

00:05:16 ◼   ► It has to stay within the Discord.

00:05:19 ◼   ► Remember the rules?

00:05:19 ◼   ► There was a Discord rule which you cannot share things from the Discord without

00:05:23 ◼   ► consenting.

00:05:24 ◼   ► That's right.

00:05:24 ◼   ► And if you're not a member, uh joining the time to sign up.

00:05:29 ◼   ► Now is the time to sign up.

00:05:30 ◼   ► Yeah, all right.

00:05:31 ◼   ► So I'm gonna paste this in the live stream.

00:05:35 ◼   ► Connectedpro dot co.

00:05:37 ◼   ► This was September two thousand and nine.

00:05:42 ◼   ► Import important content.

00:05:43 ◼   ► Okay.

00:05:44 ◼   ► Uh-huh.

00:05:45 ◼   ► Well this this is not in any way an embarrassing image.

00:05:48 ◼   ► It just looks like you now, but with a bag

00:05:51 ◼   ► There isn't even a bag.

00:05:53 ◼   ► You can't even see the bag.

00:05:54 ◼   ► This is just a guy with a t shirt and a stap.

00:05:56 ◼   ► You don't look any different.

00:05:58 ◼   ► Okay.

00:05:58 ◼   ► This is just you.

00:06:00 ◼   ► What are we looking at here, man?

00:06:01 ◼   ► It's like a picture of you with a bag.

00:06:04 ◼   ► This is hilarious.

00:06:05 ◼   ► All right, let me let me

00:06:07 ◼   ► let me go back a year.

00:06:09 ◼   ► Okay.

00:06:10 ◼   ► Hang on.

00:06:11 ◼   ► I love that this is all so image is so embarrassing.

00:06:14 ◼   ► Steven of this is very embarrassing.

00:06:16 ◼   ► Please dis doesn't leave Discord, okay.

00:06:19 ◼   ► You even have a Saint Jude t shirt on.

00:06:22 ◼   ► Like there is nothing different about This is a picture from yesterday.

00:06:25 ◼   ► This is a picture from yesterday.

00:06:26 ◼   ► That actually is uh is is at St.

00:06:28 ◼   ► Jude.

00:06:29 ◼   ► Alright here is one two thousand eight.

00:06:31 ◼   ► Are you sure you filtered by date?

00:06:33 ◼   ► Yes.

00:06:34 ◼   ► Okay.

00:06:35 ◼   ► Two thousand eight.

00:06:37 ◼   ► I mean this one is much more embarrassing because it's there's just a l

00:06:41 ◼   ► there's just a bad photo of you.

00:06:43 ◼   ► Yeah, I know it is.

00:06:44 ◼   ► Yeah.

00:06:44 ◼   ► Tiny glasses for some reason.

00:06:46 ◼   ► I this is an unflattering photo of you, dude.

00:06:48 ◼   ► Yeah.

00:06:48 ◼   ► And you are wearing quite a watch.

00:06:51 ◼   ► Oh yeah.

00:06:53 ◼   ► That's a that's a watch.

00:06:54 ◼   ► The watch is a bracelet, like a big thick bracelet.

00:06:59 ◼   ► I think all of our glasses were tiny in two thousand eight.

00:07:02 ◼   ► Yeah, yeah.

00:07:04 ◼   ► So did did you crack that column that you're

00:07:08 ◼   ► strong.

00:07:11 ◼   ► Are you wearing your genius training t-shirt in this moment?

00:07:14 ◼   ► Uh yes, I am.

00:07:16 ◼   ► Were you a genius then, or you just went to the city?

00:07:18 ◼   ► I was for the Okay, you weren't trying to steal Valor.

00:07:20 ◼   ► You do have a bag on again though, I will give you that.

00:07:23 ◼   ► Uh huh.

00:07:24 ◼   ► Yeah.

00:07:24 ◼   ► Do you mean messenger bags?

00:07:26 ◼   ► Is this what we're talking about?

00:07:27 ◼   ► Okay.

00:07:30 ◼   ► Yeah.

00:07:30 ◼   ► Like a small it's not like a laptop messenger bag.

00:07:33 ◼   ► It's like a In that one, I probably had like a notebook to write

00:07:37 ◼   ► poems in and some closed cigarettes in my wallet, you know?

00:07:43 ◼   ► There is, by the way.

00:07:44 ◼   ► Federico knows what's up.

00:07:46 ◼   ► If you are a relatively new listener of this show, like There is like

00:07:49 ◼   ► a whole different version of Steven that used to be the guy we know.

00:07:54 ◼   ► Like Yeah.

00:07:56 ◼   ► He used to drink, he used to smoke.

00:07:58 ◼   ► Like he was a very different person to to the one that we have

00:08:02 ◼   ► now.

00:08:03 ◼   ► Grew up.

00:08:05 ◼   ► Well, I wouldn't say growing up.

00:08:06 ◼   ► You just changed.

00:08:09 ◼   ► You just changed.

00:08:10 ◼   ► Okay.

00:08:11 ◼   ► So anyways, there's there's those things.

00:08:13 ◼   ► Anyways, twenty sixteen, not eligible for a retro party in my mind.

00:08:18 ◼   ► No, I don't think so.

00:08:19 ◼   ► No, I agree.

00:08:19 ◼   ► I agree.

00:08:20 ◼   ► I almost told her she couldn't party because of it.

00:08:22 ◼   ► That's not fair, but I understand.

00:08:24 ◼   ► No, she's going.

00:08:25 ◼   ► I understand the feeling.

00:08:27 ◼   ► Follow up.

00:08:28 ◼   ► Mm-hmm.

00:08:28 ◼   ► Follow up.

00:08:30 ◼   ► Dilpreet wrote in, just had a question for Mike.

00:08:33 ◼   ► He mentioned that he picked up his iPhone in America and got it repaired

00:08:37 ◼   ► in London.

00:08:38 ◼   ► So that's I guess is that your orange phone that you smashed?

00:08:42 ◼   ► Yeah.

00:08:42 ◼   ► My seventeen Pro, yeah.

00:08:43 ◼   ► Okay.

00:08:44 ◼   ► Presumably via Apple Care.

00:08:46 ◼   ► So was it via Apple Care?

00:08:47 ◼   ► You had Apple Care on that phone.

00:08:48 ◼   ► Okay.

00:08:49 ◼   ► Mm-hmm.

00:08:49 ◼   ► So was he able to buy UK Apple coverage all okay?

00:08:53 ◼   ► I'm from Australia and will be in the US for a trip and I'm

00:08:56 ◼   ► going to pick up the duo.

00:08:58 ◼   ► But I want to know that if when I get home I can still

00:09:01 ◼   ► buy the Australian Apple Care.

00:09:03 ◼   ► Yeah, I just bought it through Sattings.

00:09:05 ◼   ► So it just does whatever it does.

00:09:07 ◼   ► And my assumption is it would be it was tied to the UK.

00:09:11 ◼   ► Mm-hmm.

00:09:11 ◼   ► It's like I didn't buy it at purchase for this reason.

00:09:14 ◼   ► I I bought it through the very helpful badge that they put in the

00:09:18 ◼   ► settings app to remind you that you need to get Apple Care.

00:09:21 ◼   ► Um and I used I did that.

00:09:23 ◼   ► And then I was able to get it repaired here with not a any

00:09:27 ◼   ► real problem.

00:09:28 ◼   ► You may remember I had I couldn't go through the normal channel.

00:09:31 ◼   ► I had to kind of like speak to a support specialist and they helped

00:09:35 ◼   ► me out.

00:09:35 ◼   ► Like it it wasn't completely smooth, but it wasn't either.

00:09:39 ◼   ► It it was just the system wouldn't allow it, so they had to to

00:09:42 ◼   ► kind of help.

00:09:43 ◼   ► One thing I have noticed, and I think it's a similar thing, in now

00:09:47 ◼   ► that I have my uh eighteen pro, I'm I'm preparing some products for trade

00:09:52 ◼   ► in to have credit for the doer, right?

00:09:54 ◼   ► Mm-hmm.

00:09:55 ◼   ► And And the automatic kind like the automated system for trading in an iPhone

00:10:02 ◼   ► does not accept my serial number of my seventeen pro.

00:10:06 ◼   ► So that is just something to note.

00:10:08 ◼   ► I am sure that Apple would take this phone for trade in, but I

00:10:12 ◼   ► again similarly I just don't think I can go through all of the regular

00:10:16 ◼   ► systems to do this.

00:10:17 ◼   ► Um what I'm actually going to do is keep that phone, make that phone

00:10:21 ◼   ► my webcam and then trade in the fifteen pro that I have as the

00:10:25 ◼   ► webcam in this for for recording podcasts and stuff with.

00:10:28 ◼   ► So that's I'm gonna make because I wanna I my thoughts, oh, I wanna

00:10:32 ◼   ► keep that orange one around for a while.

00:10:34 ◼   ► Yeah.

00:10:34 ◼   ► So I'll do that.

00:10:35 ◼   ► But yeah, that so that's just something to note.

00:10:38 ◼   ► I'll say that like m there is nothing is as straightforward as it would

00:10:43 ◼   ► be otherwise, but it is absolutely possible to do.

00:10:47 ◼   ► And if I was gonna be in the US, I would do the exact

00:10:51 ◼   ► same thing and would get Apple Care on the Duo.

00:10:54 ◼   ► Um the duo feels like a phone you want Apple Care on, I think.

00:10:57 ◼   ► Yes.

00:10:57 ◼   ► For sure.

00:10:58 ◼   ► Just in general.

00:11:00 ◼   ► For sure.

00:11:02 ◼   ► Um I'm gonna be you know, in the pro show we talked about travel

00:11:06 ◼   ► and how I don't go anywhere.

00:11:08 ◼   ► Yeah, I actually have like two trips back to back coming up.

00:11:10 ◼   ► I'm gonna be out of town.

00:11:12 ◼   ► I'm gonna be flying home from a trip the day duo pre orders open.

00:11:17 ◼   ► Like I'm gonna be ordering it from the airport.

00:11:19 ◼   ► You hope.

00:11:20 ◼   ► I hope.

00:11:21 ◼   ► And then when it comes out, I'm gonna be on a second trip and

00:11:25 ◼   ► I'm I'm gonna order it for pickup in an Apple store in the city

00:11:28 ◼   ► I will be in.

00:11:29 ◼   ► Like it is the most I'm You are stacking problems on top of each

00:11:34 ◼   ► other.

00:11:34 ◼   ► I I am.

00:11:35 ◼   ► It's gonna be uh an adventure.

00:11:38 ◼   ► And the the problem will be is if that second trip falls through for

00:11:40 ◼   ► some reason, uh I still have to go to Missouri to get my phone.

00:11:47 ◼   ► Oh look, I don't want bad things to happen to you.

00:11:50 ◼   ► I really want that to happen because that is so funny.

00:11:53 ◼   ► Ten hour round trip to get to get my my duo.

00:11:57 ◼   ► God, that would be so funny.

00:11:58 ◼   ► Uh would I I would do it for the show, you know what I

00:12:01 ◼   ► mean?

00:12:01 ◼   ► Yeah.

00:12:02 ◼   ► Oh oh I mean it would be just incredible stuff.

00:12:06 ◼   ► That's that's amazing.

00:12:07 ◼   ► It would be something good.

00:12:09 ◼   ► Uh I got sent this by a listener.

00:12:11 ◼   ► Um quote Apple says that on the iPhone eighteen pro or later, the new

00:12:15 ◼   ► under display face ID technology uses facial images in addition to data from the

00:12:20 ◼   ► infrared sensor to authenticate a user.

00:12:22 ◼   ► Now, I note this because one this is close to what I was talking

00:12:27 ◼   ► about, right?

00:12:28 ◼   ► Of like how they could maybe have face ID on on the duo.

00:12:32 ◼   ► Um and but really it ultimately seems like the beginning of a change to

00:12:36 ◼   ► allow the face ID technology to work in other ways, right?

00:12:39 ◼   ► Like you They need to train a system, and this is part of how

00:12:44 ◼   ► you do it.

00:12:44 ◼   ► You know, maybe this is in the duo two or the home part of

00:12:46 ◼   ► a screen or whatever.

00:12:48 ◼   ► I will note on my 18 Pro face ID is less reliable.

00:12:52 ◼   ► Absolutely is less reliable.

00:12:54 ◼   ► Now I don't know if it's because it's a new phone, right?

00:12:57 ◼   ► Sometimes this stuff can take a bit to to to r kind of relearn

00:13:00 ◼   ► new.

00:13:01 ◼   ► Um, but I am noticing at the moment, a couple of weeks in, lots

00:13:06 ◼   ► of face ID failures.

00:13:07 ◼   ► Like I'm trying to pay for something and it's like I'll actually put your

00:13:10 ◼   ► passcode in.

00:13:11 ◼   ► So I'm not surprised if they're trying a new system that this is happening.

00:13:16 ◼   ► But also at the same time, there is also some kind of bug in

00:13:19 ◼   ► the face ID system.

00:13:21 ◼   ► Um where if a face ID fails for some users, it reboots their phone,

00:13:26 ◼   ► which is hilarious.

00:13:27 ◼   ► Um Apple has confirmed that they're gonna fix this with a with an update.

00:13:31 ◼   ► But I'm also on the twenty-seven point two beta, so maybe they've already fixed

00:13:35 ◼   ► it by uh by the time that came out.

00:13:38 ◼   ► Because I haven't had that happen.

00:13:39 ◼   ► But I have s I have noticed some unreliability um in in the system

00:13:44 ◼   ► itself, I think.

00:13:46 ◼   ► There is another issue that is popping up for some eight iPhone 18 Pro

00:13:51 ◼   ► users.

00:13:52 ◼   ► I don't like where this is going.

00:13:53 ◼   ► Some iPhone eighteen users are already seeing green and pink lines in the display.

00:13:57 ◼   ► No.

00:13:58 ◼   ► From phone arena.

00:13:59 ◼   ► Not true.

00:14:00 ◼   ► Phone arena.

00:14:01 ◼   ► We're sourced.

00:14:02 ◼   ► Complete fabrication.

00:14:03 ◼   ► It's on Reddit.

00:14:04 ◼   ► Have you ev have we ever sourced phone arena before?

00:14:07 ◼   ► Uh like fifteen people sent this to me.

00:14:09 ◼   ► I guess Green Gate is my brand.

00:14:12 ◼   ► Why would that be?

00:14:13 ◼   ► Would it be because you won't stop talking about it?

00:14:16 ◼   ► Because uh it just keeps happening.

00:14:18 ◼   ► There's one piece of journalist.

00:14:19 ◼   ► There's one video.

00:14:20 ◼   ► There's one video in Fonarina and it shows a pink line, not a green

00:14:24 ◼   ► one.

00:14:25 ◼   ► They're very similar.

00:14:27 ◼   ► No.

00:14:29 ◼   ► There was something on threads too, but uh someone had sent me, but yeah.

00:14:34 ◼   ► It's not really an issue.

00:14:35 ◼   ► It's

00:14:37 ◼   ► could be an issue.

00:14:38 ◼   ► Yeah, if you want to make it one.

00:14:40 ◼   ► You love this, don't you?

00:14:41 ◼   ► Uh last week we spoke about kind of how code mining, I d I

00:14:45 ◼   ► think that might be the phrase I'm that's the phrase I'm using, uh is

00:14:48 ◼   ► really like ramping up where people are using AI models to and LLMs or

00:14:54 ◼   ► whatever to dig through releases of uh OS updates to find things.

00:14:59 ◼   ► Right.

00:15:00 ◼   ► We were talking about how it seems like this is increasing rapidly.

00:15:05 ◼   ► I have three things for you today that have come out of the point

00:15:08 ◼   ► two betas, I think.

00:15:10 ◼   ► Um these are all from 9 to 5 Mac.

00:15:12 ◼   ► I'll put links in the show notes.

00:15:13 ◼   ► Uh I'm gonna read this one out because it's detailed and I think it's

00:15:16 ◼   ► interesting.

00:15:16 ◼   ► According to new code seen by nine to five Mac, the feature this feature

00:15:20 ◼   ► called Auto Lock internally can use the following signs to determine whether an iPhone

00:15:26 ◼   ► may have been taken from its owner.

00:15:27 ◼   ► So this is to prevent snatching, right?

00:15:31 ◼   ► A trend of theft where people will will be on a bike or some

00:15:35 ◼   ► kind of motorized scooter and they will snatch your phone and get out of

00:15:40 ◼   ► there, right?

00:15:41 ◼   ► So it can detect it with either a sharp acceleration consistent with a phone

00:15:45 ◼   ► being snatched, a paired Apple Watch remaining unreachable for a certain period, loss of

00:15:49 ◼   ► connectivity to a paired watch, prolonged loss of network activity, or the device remaining

00:15:53 ◼   ► unlocked beyond a certain interval.

00:15:55 ◼   ► New references include safeguards meant to prevent false positives such as biometric authentication and

00:16:01 ◼   ► all these kinds of things.

00:16:02 ◼   ► So they're like they're weighing up essentially.

00:16:04 ◼   ► There's like this internal system method of voting as like positives and negatives as

00:16:10 ◼   ► to whether it believes that a phone has been snatched out of someone's hands.

00:16:15 ◼   ► Fascinating thing that they're developing, um, to try and it obviously well, I mean

00:16:21 ◼   ► it could end up stopping it from happening because if the phones will just

00:16:24 ◼   ► all be rendered useless or completely locked and then what are you gonna do

00:16:27 ◼   ► with them?

00:16:28 ◼   ► So yeah.

00:16:28 ◼   ► I thought this was an interesting thing that Apple's doing, but it's only been

00:16:32 ◼   ► found because people are digging through code.

00:16:34 ◼   ► Uh also uh Apple music release countdowns.

00:16:37 ◼   ► So if someone's got a new album coming out, there's like a ticker that

00:16:40 ◼   ► counts down.

00:16:41 ◼   ► Third party action didn't describe a shortcut have been found.

00:16:45 ◼   ► And a new I I expected that this was found through um looking through

00:16:51 ◼   ► these systems because I don't know how anyone would find it.

00:16:53 ◼   ► This is actually currently in 27.2 on the watch.

00:16:58 ◼   ► The app switcher is back.

00:17:00 ◼   ► So you know where you would be able to to see the apps that

00:17:03 ◼   ► you had open and cycle through them and quit them.

00:17:06 ◼   ► But the way that you activate it is by long pressing on the bottom

00:17:10 ◼   ► of the watch screen when an app is open.

00:17:12 ◼   ► It is strange.

00:17:15 ◼   ► It is a incredibly weird gesture.

00:17:19 ◼   ► I don't think this is the gesture they're gonna settle on, and I actually

00:17:22 ◼   ► don't know if they're gonna have this feature available at all, 'cause it's like

00:17:26 ◼   ► that doesn't make any sense, right?

00:17:29 ◼   ► Like as a as a watch OS kind of interaction.

00:17:34 ◼   ► But it works on twenty seven point two, watch off twenty seven point two,

00:17:37 ◼   ► right now, you can do it if you have the beta on your watch,

00:17:39 ◼   ► long press on the bottom of the screen and the app switcher will come

00:17:42 ◼   ► up.

00:17:43 ◼   ► W why can't you just like double press the digital crown?

00:17:47 ◼   ► That seems so much simpler.

00:17:49 ◼   ► Because that's gonna be doing the what did that person say thing.

00:17:53 ◼   ► Oh yeah, they should not have that feature.

00:17:56 ◼   ► Replace it with this.

00:17:57 ◼   ► What I actually, you know, I completely forgot about that.

00:18:01 ◼   ► Uh-huh.

00:18:01 ◼   ► I'm putting my Apple Watch on right now.

00:18:05 ◼   ► How do you activate it?

00:18:06 ◼   ► Do I have access to it?

00:18:07 ◼   ► Hang on.

00:18:10 ◼   ► The audio Oh that's later.

00:18:12 ◼   ► Yeah, it's not even in beta, I don't think.

00:18:15 ◼   ► Yeah, that's coming later this year the that and the like getting um basically

00:18:19 ◼   ► high level meeting notes of your day kind of thing.

00:18:22 ◼   ► Are we willing to bet if that feature actually ships or not?

00:18:26 ◼   ► I feel like they're they're already getting pushback on it.

00:18:29 ◼   ► Here's my feeling on this.

00:18:30 ◼   ► I think they made a mistake by announcing it because I think that these

00:18:35 ◼   ► features aren't gonna be as creepy as they seem because they're not gonna be

00:18:39 ◼   ► very detailed.

00:18:40 ◼   ► And I think once they ship, people are gonna be like, Ah actually this

00:18:43 ◼   ► isn't very detailed.

00:18:44 ◼   ► Like the meeting notes thing or like the kind of notes of your day.

00:18:47 ◼   ► I think I think the problem they had is announcing something that they didn't

00:18:51 ◼   ► have ready because it enables it allows people to kind of like build an

00:18:55 ◼   ► idea of to like how could how creepy it's gonna be and I think

00:18:58 ◼   ► it's not I believe it's not gonna be as detailed as people fear and

00:19:02 ◼   ► it's just not really gonna be like much of a thing but we'll see.

00:19:05 ◼   ► I think they're gonna release it.

00:19:06 ◼   ► I absolutely think they're gonna release it.

00:19:08 ◼   ► And but it will just be about we'll see how people react when they

00:19:12 ◼   ► do.

00:19:12 ◼   ► Yeah, we'll see.

00:19:14 ◼   ► I saw Marquez made a video about it and I haven't watched it yet,

00:19:17 ◼   ► but it's MAQ.

00:19:19 ◼   ► I do agree that announcing early was definitely the wrong the wrong move.

00:19:23 ◼   ► Um 'cause it just an allows people to just paint an image in their

00:19:26 ◼   ► mind, whether right or wrong, of like what this feature is gonna be.

00:19:30 ◼   ► Um and they they absolutely did not need to announce it with the Apple

00:19:33 ◼   ► Watch.

00:19:33 ◼   ► They they could have announced it when it was ready and I think it

00:19:35 ◼   ► would have been fine.

00:19:36 ◼   ► They could have just made a separate video about it or whatever.

00:19:38 ◼   ► Um, but now it's kind of like I think this is one of these

00:19:42 ◼   ► scenarios where the story starts to get away from them.

00:19:45 ◼   ► Um and I think that's where they are right now.

00:19:48 ◼   ► Big news.

00:19:50 ◼   ► It's September.

00:19:51 ◼   ► Do you guys know that?

00:19:53 ◼   ► September?

00:19:53 ◼   ► I don't think so.

00:19:54 ◼   ► September is Childhood Cancer Awareness Month.

00:19:57 ◼   ► And each September, those of us at Relay and our friends at ATP and

00:20:01 ◼   ► a bunch of other folks come together to raise money for St.

00:20:04 ◼   ► Jude Children's Research Hospital.

00:20:06 ◼   ► Uh as we're recording right now, this year we have raised just over $450,000,

00:20:14 ◼   ► which is absolutely incredible.

00:20:16 ◼   ► That puts us at a mind-blowing five point two million dollars raise together since

00:20:22 ◼   ► twenty nineteen.

00:20:24 ◼   ► If you go donate in the next couple of days, you can help uh

00:20:28 ◼   ► pick Mike's cowboy name for the Buck Acathon, we have Moonlight Mike and Time

00:20:33 ◼   ► Track and Hurley.

00:20:34 ◼   ► Mm-hmm.

00:20:35 ◼   ► Currently that one is winning by quite a large margin.

00:20:37 ◼   ► But as it should.

00:20:38 ◼   ► If you got if you got a big donation to make, which you should

00:20:41 ◼   ► go make, and you want Moonlight Mike, now's the time.

00:20:44 ◼   ► You know, just swing it the other way.

00:20:46 ◼   ► Yeah, I don't like the phrase if you would like moonlight mic.

00:20:50 ◼   ► I don't think I like that as just like a thing to say in

00:20:53 ◼   ► that in the words in that order.

00:20:54 ◼   ► I want moonlight mic.

00:20:56 ◼   ► Well, I'm coming to Memphis.

00:20:57 ◼   ► Yeah, buddy You got it if you want it.

00:21:01 ◼   ► Yes.

00:21:02 ◼   ► Uh so how do you do this?

00:21:04 ◼   ► You go to Saint Jew dot org and you can donate.

00:21:08 ◼   ► A donation of sixty dollars and more gets you a really cool digital bundle

00:21:11 ◼   ► that includes a macOS screensaver that our friend James Thompson works on.

00:21:16 ◼   ► There's link in the show notes to a sizzle reel that he put together

00:21:19 ◼   ► of the screensaver.

00:21:22 ◼   ► It

00:21:23 ◼   ► has gone so much further than it should have this year.

00:21:27 ◼   ► Like I I've been unable to rein him in, honestly.

00:21:29 ◼   ► It's a

00:21:32 ◼   ► absolutely fantastic stuff.

00:21:34 ◼   ► Uh w wallpapers, if you are in that digital bundle as well.

00:21:38 ◼   ► They're stickers.

00:21:38 ◼   ► And if you raise If you start your own fundraiser to raise money, like

00:21:42 ◼   ► within your social circles, it unlocks a whole nother category of incentives, including a

00:21:47 ◼   ► really awesome hat, a very fancy leather cord organizer, uh a bandana, a bunch

00:21:53 ◼   ► of other stuff.

00:21:54 ◼   ► It's all Western themed.

00:21:55 ◼   ► You know, we're on theme this year.

00:21:57 ◼   ► So that's what it is.

00:21:59 ◼   ► But why would you do this?

00:22:00 ◼   ► Why why why do we spend time each September for eight years talking about

00:22:05 ◼   ► Saint Jude?

00:22:06 ◼   ► Uh it's because St.

00:22:07 ◼   ► Jude does really incredible work in the world of children's cancer research and treatment.

00:22:16 ◼   ► In June 2025, 31 students graduated from the St.

00:22:21 ◼   ► Jude Graduate School of Biomedical Sciences.

00:22:25 ◼   ► Now look, I like to think I'm a relatively smart person.

00:22:30 ◼   ► Um, these people truly next level.

00:22:33 ◼   ► They are doing incredible work.

00:22:35 ◼   ► This school offers a PhD in biomedical science and MS degrees in clinical investigations,

00:22:43 ◼   ► global child health, biomedical sciences, and and applied biomedical data sciences.

00:22:50 ◼   ► I don't know what some of those words are.

00:22:51 ◼   ► I don't know what it means.

00:22:52 ◼   ► Don't know what it means.

00:22:55 ◼   ► You know it's important.

00:22:56 ◼   ► And what's what's cool about this, this graduate school, St.

00:22:59 ◼   ► Jude is investing in these students so that they can do great work in

00:23:05 ◼   ► these fields.

00:23:06 ◼   ► The graduate stud the graduate school has now awarded degrees to more than a

00:23:10 ◼   ► hundred and thirty students in the last 10 years, equipping the next generation of

00:23:15 ◼   ► leaders in research, medicine, and global health.

00:23:20 ◼   ► One reason I absolutely love this is that this is a multi-generational fight to

00:23:26 ◼   ► figure and figure this out with childhood cancer and other catastrophic diseases.

00:23:30 ◼   ► And there's been incredible progress since St.

00:23:33 ◼   ► Jude started in nineteen sixty two.

00:23:37 ◼   ► The ch overall survival rate in the US has gone from twenty percent to

00:23:41 ◼   ► Eighty percent.

00:23:42 ◼   ► But one, that's not true worldwide.

00:23:44 ◼   ► And secondly, eighty percent is not good enough.

00:23:47 ◼   ► We've got to move that number up.

00:23:48 ◼   ► So St.

00:23:49 ◼   ► Jude is investing on the people in the people who are doing this work.

00:23:54 ◼   ► Saint Jude treats families like mine.

00:23:56 ◼   ► My oldest child was a St.

00:23:57 ◼   ► Jude patient when he was young.

00:23:59 ◼   ► He's getting ready to turn 18 and he's doing great because of the work

00:24:02 ◼   ► at St.

00:24:02 ◼   ► Jude.

00:24:03 ◼   ► St.

00:24:03 ◼   ► Jude didn't bill us for his treatment.

00:24:06 ◼   ► If we had to travel to St.

00:24:07 ◼   ► Jude, if we had to do lodging.

00:24:09 ◼   ► Meals.

00:24:09 ◼   ► St.

00:24:10 ◼   ► Jude pays for all that stuff because they want families to be able to

00:24:13 ◼   ► focus on their child and their family and getting through this and and healing.

00:24:19 ◼   ► And that's only possible.

00:24:20 ◼   ► That graduate school is only possible.

00:24:23 ◼   ► Caring for families without billing them is only possible.

00:24:25 ◼   ► All this research is only possible because regular people like me and like you

00:24:30 ◼   ► support Saint Jude.

00:24:32 ◼   ► So we would urge you to go to Saint Jude.org/slash relay and make a

00:24:36 ◼   ► donation.

00:24:37 ◼   ► Uh now is the time.

00:24:38 ◼   ► We're doing this through the end of the month, but we uh we want

00:24:41 ◼   ► to come into the pocastathon next week strong.

00:24:44 ◼   ► Mike, you want to talk a little bit about the show?

00:24:46 ◼   ► Yeah, so we're gonna be live streaming uh on Friday, October second from twelve

00:24:53 ◼   ► PM US Eastern time for twelve hours, so until twelve AM.

00:24:57 ◼   ► Uh there w we will all be hanging out all day, playing games, having

00:25:01 ◼   ► fun, getting up to no good uh and getting up to lots of good

00:25:04 ◼   ► at the same time.

00:25:05 ◼   ► Uh it's all gonna be on YouTube.

00:25:06 ◼   ► There's a link in the show notes.

00:25:07 ◼   ► You can go now and like press notify me and it will remind you

00:25:10 ◼   ► that it's there.

00:25:11 ◼   ► Uh but we'll be sharing it too.

00:25:13 ◼   ► Uh there's actually gonna be no episode of Connected next week because we'll be

00:25:16 ◼   ► in all day rehearsal.

00:25:17 ◼   ► Um, so we'll be we're not gonna be uh recording next week.

00:25:21 ◼   ► So if you really miss Steven and I, uh you can um you can

00:25:24 ◼   ► tune into the podcast of Thun and if you miss Federico, you could go

00:25:27 ◼   ► to Rome.

00:25:28 ◼   ► Um maybe you'll find him.

00:25:29 ◼   ► That's what that's what I recommend.

00:25:31 ◼   ► I don't know Federica thinks about that, but that's what I recommend.

00:25:33 ◼   ► Uh I mean, but look, you wouldn't miss him if you find him.

00:25:36 ◼   ► That's what I'm gonna say.

00:25:38 ◼   ► Uh so yeah, we're gonna be hanging out all day.

00:25:39 ◼   ► It's gonna be a great time.

00:25:40 ◼   ► Bunch of your favorite really hosts are gonna be there.

00:25:43 ◼   ► Um, maybe some of your least favorite relay hosts are gonna be there.

00:25:46 ◼   ► You know, I don't speak for who you like and who you don't, but

00:25:49 ◼   ► um you can come and hang out with us uh over at YouTube and

00:25:52 ◼   ► there's a link in the show notes for that.

00:25:53 ◼   ► Wow.

00:25:54 ◼   ► That's it went a a certain direction.

00:25:57 ◼   ► Uh different different directions.

00:25:59 ◼   ► A lot of different directions.

00:26:00 ◼   ► I will say my garage is full of set pieces for the live show

00:26:05 ◼   ► that I am finishing putting the finishing touches on.

00:26:09 ◼   ► I had to buy a metal disc, a cutting disc for my cutoff wheel

00:26:14 ◼   ► this morning because I got to cut some metal this afternoon.

00:26:16 ◼   ► Okay.

00:26:17 ◼   ► It's getting serious down there.

00:26:18 ◼   ► Sounds like it.

00:26:19 ◼   ► So saintjew.org slash relay, go make a donation, go sign up to fundraise and

00:26:25 ◼   ► stay tuned next Friday for the podcast a thon.

00:26:27 ◼   ► And thank you all who have donated, who have supported this campaign this year

00:26:31 ◼   ► in the past.

00:26:32 ◼   ► We've done some truly amazing work together in this.

00:26:35 ◼   ► And it it it just blows me away.

00:26:38 ◼   ► Apple owns a music venue now.

00:26:41 ◼   ► It's called Apple Music Hall.

00:26:43 ◼   ► Nailed it.

00:26:44 ◼   ► It's in the basement of uh the Black Sea Power Station where Apple has

00:26:49 ◼   ► its uh I think European headquarters, I think, is the Uh is London.

00:26:54 ◼   ► Ironic.

00:26:56 ◼   ► Stephen, I was thinking about this the other day.

00:26:58 ◼   ► Are you aware of the fact that the United Kingdom is in the continent

00:27:01 ◼   ► of Europe?

00:27:02 ◼   ► I mean it's an island.

00:27:03 ◼   ► It's like what?

00:27:05 ◼   ► Next door.

00:27:06 ◼   ► I don't want to have this conversation anymore because I don't even know what's

00:27:09 ◼   ► happening.

00:27:09 ◼   ► Uh so th this is a very beautiful venue.

00:27:12 ◼   ► Um they're actually I think tomorrow as we record this, they are announcing who

00:27:16 ◼   ► the first acts are gonna be that the play.

00:27:19 ◼   ► They're being a bit cagey about this I think purposefully.

00:27:22 ◼   ► They're trying to put assignment.

00:27:24 ◼   ► It's us.

00:27:25 ◼   ► I have had the thought of like, what does it take to get a

00:27:28 ◼   ► podcast recording in this place?

00:27:29 ◼   ► Like I know it's the Apple Music Hall, but come on now.

00:27:32 ◼   ► We do a real good job here.

00:27:34 ◼   ► Uh it's a relatively small place.

00:27:36 ◼   ► It's a capacity of six hundred.

00:27:39 ◼   ► Um and they basically with an artist, what you're gonna get out of it

00:27:44 ◼   ► is a video, a special audio recording mix.

00:27:47 ◼   ► This is where they will do what used to be the iTunes sessions, right?

00:27:52 ◼   ► Like and do you remember the they did the thing at the roundhouse?

00:27:55 ◼   ► What was that called?

00:27:56 ◼   ► Was it the iTunes Festival?

00:27:58 ◼   ► Yeah.

00:27:59 ◼   ► That kind of energy is gonna be all be in this place.

00:28:02 ◼   ► So they're gonna be, I'm sure, hosting emerging artists and big artists and whoever

00:28:09 ◼   ► wins that award, you know, that they do every year will probably do a

00:28:13 ◼   ► set there or something.

00:28:14 ◼   ► It's probably gonna be Billie Eilish again.

00:28:16 ◼   ► Yeah, I guarantee Billy Eilish will play the the Apple Music Hall.

00:28:21 ◼   ► Uh yeah, it's it's beautiful.

00:28:23 ◼   ► The whole thing, the inside and out, is like stunning.

00:28:26 ◼   ► Which doesn't Battersea is beautiful.

00:28:30 ◼   ► Um so yeah.

00:28:31 ◼   ► They've also got some um

00:28:33 ◼   ► they call them the Apple Music Radio Studios as well in there.

00:28:38 ◼   ► So they're they're doing they they've they're consolidating a lot of stuff uh into

00:28:42 ◼   ► this place, which is yeah really nice looking.

00:28:44 ◼   ► A couple of things came to mind looking at these pictures, and it is

00:28:47 ◼   ► incredible.

00:28:48 ◼   ► If you haven't seen these, I mean it's just it's a it's what you

00:28:50 ◼   ► would think.

00:28:51 ◼   ► This is an expensive space.

00:28:52 ◼   ► This is like this costs a lot of money.

00:28:53 ◼   ► This costs a lot of money.

00:28:55 ◼   ► I like that all of the uh I guess box seating.

00:29:00 ◼   ► It looks like you're sitting inside a Mac Mini.

00:29:03 ◼   ► Which is very funny.

00:29:04 ◼   ► Yeah uh to me.

00:29:05 ◼   ► A little handrail there.

00:29:06 ◼   ► Not a lot of seats.

00:29:07 ◼   ► It's it's it's an intimate venue, it looks like.

00:29:10 ◼   ► Um and there's a glass wall at the back, which It's confusing to me

00:29:14 ◼   ► for acoustic purposes, but that's okay.

00:29:16 ◼   ► I like that you I like that you think you spotted something.

00:29:20 ◼   ► It's like, oh come on, gang.

00:29:21 ◼   ► Did you not notice the clouds?

00:29:22 ◼   ► Did you not notice it?

00:29:23 ◼   ► It's brick and glass.

00:29:24 ◼   ► Yeah.

00:29:25 ◼   ► What are you doing?

00:29:25 ◼   ► No, that it looks awesome.

00:29:27 ◼   ► I'm convinced that they've taken care of the acoustics.

00:29:29 ◼   ► I mean, hopefully.

00:29:31 ◼   ► I mean we've recorded some doozies with Apple in the past, acoustic wise, you

00:29:36 ◼   ► know.

00:29:36 ◼   ► Yeah.

00:29:37 ◼   ► Maybe they've improved.

00:29:38 ◼   ► Maybe.

00:29:39 ◼   ► Uh I want to see this place so badly.

00:29:41 ◼   ► Like I I I want to see it so badly.

00:29:46 ◼   ► Let me So what we're saying is we will record an episode of Connected

00:29:50 ◼   ► Call.

00:29:50 ◼   ► Yeah.

00:29:51 ◼   ► Yeah.

00:29:51 ◼   ► Yeah.

00:29:52 ◼   ► We'll make it happen.

00:29:54 ◼   ► This episode of Connected is brought to you by Scribe.

00:29:58 ◼   ► Knowledge transfer is tricky to get right.

00:30:01 ◼   ► If you ever onboarded new employees and realized the same questions crop up over

00:30:06 ◼   ► time, then you know this.

00:30:08 ◼   ► The most valuable workflows can be hard to capture.

00:30:11 ◼   ► And that means that things often go undocumented and your team, especially new members,

00:30:17 ◼   ► are left guessing.

00:30:19 ◼   ► That's the problem that today's sponsor, Scribe, was built to solve.

00:30:23 ◼   ► Scribe is a specialized intelligence platform trusted by 94% of the Fortune 500.

00:30:30 ◼   ► As you work, Scribe automatically captures workflows that turns them into clear, step-by-step documentation

00:30:37 ◼   ► with screenshots.

00:30:39 ◼   ► So your team can say goodbye to writing everything up by hand, stitching together

00:30:44 ◼   ► screenshots, and recreating the same onboarding documents every time someone new joins the team.

00:30:51 ◼   ► And both your team and your AI can make use of what it captures.

00:30:56 ◼   ► Processes are hard to hand off.

00:30:58 ◼   ► If you're anything like me, you have parts of your workflow that you don't

00:31:01 ◼   ► even really think about and when someone else comes into them, you realize like

00:31:06 ◼   ► that you haven't thought about it, you haven't.

00:31:08 ◼   ► And Scribe is working in the background to resolve those issues.

00:31:12 ◼   ► It's capturing that process as the team works.

00:31:16 ◼   ► And it's critical that know-how doesn't stay stuck with one person in your company.

00:31:23 ◼   ► What should have taken hours of writing, recording, and cleanup is captured as you

00:31:28 ◼   ► work, so it's ready to review and share by the time you're done.

00:31:32 ◼   ► Inscribe automatically redacts sensitive information, names, account, numbers, emails from every screenshot.

00:31:38 ◼   ► As an admin, you can enforce this across your entire team.

00:31:42 ◼   ► And nothing is published about human review, so you can feel confident that nothing

00:31:46 ◼   ► slips through the cracks.

00:31:48 ◼   ► Anyone following the process can launch real-time on screen guidance that shows them where

00:31:53 ◼   ► to click step by step inside the actual tool.

00:31:57 ◼   ► And you can power your AI agents with real context.

00:32:00 ◼   ► Captured workflows give agents a foundation to operate and how on how work actually

00:32:05 ◼   ► happens, not guesses.

00:32:07 ◼   ► To see what scribe could look like for your organization, go to scribe.how slash

00:32:13 ◼   ► connected.

00:32:14 ◼   ► That's S-C-R-I-B-E.

00:32:16 ◼   ► Scribe.

00:32:17 ◼   ► Scribe.how slash connected.

00:32:19 ◼   ► And mention connected.

00:32:20 ◼   ► And you're going to get your first month of scribe capture free on select

00:32:24 ◼   ► plans.

00:32:25 ◼   ► My thanks to Scribe for their support of the show and all of Relay.

00:32:31 ◼   ► So I got an M five Ultra Mac Studio.

00:32:34 ◼   ► Has anybody tried to rub you yet?

00:32:37 ◼   ► No.

00:32:38 ◼   ► Um I also got an M five Pro Mac Mini that I'm using right

00:32:42 ◼   ► now.

00:32:42 ◼   ► Um but I mostly just spent time with the M5 Ultra Mac Studio.

00:32:47 ◼   ► I wrote about it, I put together lots of charts and numbers and things.

00:32:52 ◼   ► Uh Um

00:32:54 ◼   ► and so that I don't just repeat myself, uh you guys can ask me

00:32:59 ◼   ► questions.

00:32:59 ◼   ► Okay.

00:33:00 ◼   ► I'm sh I I think you have some questions.

00:33:02 ◼   ► Yeah, can you give me the full breakdown of the of the machine that

00:33:05 ◼   ► you have?

00:33:05 ◼   ► The ultra so this is the cur the current high end version.

00:33:10 ◼   ► Uh two fifty six gigs of RAM.

00:33:13 ◼   ► Uh uh because the five twelve gig version is not coming out until late

00:33:16 ◼   ► October and it was not sent out uh with review units right now.

00:33:21 ◼   ► Um

00:33:23 ◼   ► four terabytes of storage and uh I also got an M5 Pro Mac mini

00:33:29 ◼   ► and a what's it called?

00:33:30 ◼   ► Studio Display XDR.

00:33:32 ◼   ► Yeah.

00:33:33 ◼   ► Yeah.

00:33:33 ◼   ► Which I'm using right now display, by the way.

00:33:35 ◼   ► Oh, it's stunning.

00:33:37 ◼   ► I really, really like it.

00:33:38 ◼   ► I really like it.

00:33:39 ◼   ► Um it's so nice to use a Mac and high refresh rate on a

00:33:44 ◼   ► big display.

00:33:45 ◼   ► Yes.

00:33:45 ◼   ► Yeah.

00:33:46 ◼   ► So yeah, that's that's the machine.

00:33:48 ◼   ► Um what is the rough price of that machine?

00:33:52 ◼   ► Do you know?

00:33:53 ◼   ► I mean I I'll spec it out if you don't, but I I wondered

00:33:55 ◼   ► if you'd look I'm gonna say like eighteen thousand euros.

00:33:59 ◼   ► Okay.

00:34:00 ◼   ► I'll gotta look in dollars.

00:34:01 ◼   ► So you've got the M five Ultra, the full thirty six and eighty core.

00:34:06 ◼   ► Yeah.

00:34:07 ◼   ► Two fifty six gigabytes of storage.

00:34:09 ◼   ► How much SSD did you say you see?

00:34:11 ◼   ► Eight four terabytes.

00:34:12 ◼   ► Four.

00:34:12 ◼   ► Okay.

00:34:13 ◼   ► So you don't have the maximum SSD.

00:34:14 ◼   ► Yeah that's a that's a plebeian model.

00:34:17 ◼   ► That's twelve thousand three hundred dollars.

00:34:20 ◼   ► Jeez.

00:34:21 ◼   ► It can go up another six grand if you get the maximum.

00:34:25 ◼   ► Yeah, no, it's it's it's fourteen thousand euros.

00:34:28 ◼   ► Yeah.

00:34:29 ◼   ► No.

00:34:29 ◼   ► Yeah.

00:34:30 ◼   ► Okay.

00:34:31 ◼   ► So I mean they obviously gave you

00:34:36 ◼   ► what you needed to do what you did, right?

00:34:38 ◼   ► Which is like 'cause what you need is the ultra with the most amount

00:34:42 ◼   ► of RAM and then you can put local AI for its paces, right?

00:34:46 ◼   ► Like I guess that's what they gave it to you for.

00:34:48 ◼   ► Yes.

00:34:49 ◼   ► So going into this I knew that the reason they were giving me this

00:34:53 ◼   ► Mac Studio was to experiment with local AI models um on this machine.

00:35:00 ◼   ► Uh uh uh uh essentially like the very bleeding edge of local AI at

00:35:06 ◼   ► the moment.

00:35:07 ◼   ► And that's that's what I prepared.

00:35:09 ◼   ► I prepared for it uh by having a system and we can talk about

00:35:14 ◼   ► that uh maybe uh having a system to test the machine.

00:35:17 ◼   ► Uh but my focus was gonna be just on local AI.

00:35:21 ◼   ► Uh I am not an expert in video editing or video rendering.

00:35:26 ◼   ► I did not want to do tests about like just having geekbench numbers.

00:35:31 ◼   ► And there's nothing that could have uh differentiated me about that.

00:35:36 ◼   ► Um and I feel very passionate about local AI and I feel very curious

00:35:41 ◼   ► about local AI.

00:35:42 ◼   ► So that was gonna be my angle.

00:35:44 ◼   ► Well, I mean, I would really recommend that people go to Mac Stories and

00:35:48 ◼   ► look at your charts and your numbers and all this kind of stuff.

00:35:50 ◼   ► So people can I mean one, the the chart The charts are very impressive

00:35:53 ◼   ► and like they do a good job of explaining visually the things that were

00:35:56 ◼   ► going over my head when I was reading them, right?

00:35:58 ◼   ► Where it's it's like a lot of data.

00:36:01 ◼   ► And so I don't particularly want to focus on the numbers because I don't

00:36:05 ◼   ► I don't think that that would translate very well to a to a podcast

00:36:08 ◼   ► anyway, and like do charts do a better job.

00:36:10 ◼   ► I would like to talk about the

00:36:14 ◼   ► the uses of this machine when you're thinking about local AI models, right?

00:36:20 ◼   ► Because

00:36:22 ◼   ► you can just, as you did, put together a bunch of great tests that

00:36:26 ◼   ► can push the thing to the maximum so you can see how fast can

00:36:29 ◼   ► it do something.

00:36:30 ◼   ► Right.

00:36:30 ◼   ► Like that is benchmarking, right?

00:36:32 ◼   ► It's like geekbench.

00:36:33 ◼   ► But ultimately, the geekbench score doesn't really mean anything, right?

00:36:38 ◼   ► Like it's like, oh great, like d here's a number.

00:36:41 ◼   ► But me as a user, like what does that mean?

00:36:43 ◼   ► So what I would like to understand from you is what were you able

00:36:47 ◼   ► to achieve with local AI models on the M5 Ultra that you have otherwise

00:36:54 ◼   ► felt unimpressed by in other scenarios and other computers?

00:37:00 ◼   ► Can you put a pen in this?

00:37:01 ◼   ► I need two minutes to go check on my dog.

00:37:03 ◼   ► Yeah.

00:37:03 ◼   ► And I'll and I'll answer right now.

00:37:05 ◼   ► We have two minutes to think of an answer.

00:37:07 ◼   ► Yeah, okay.

00:37:08 ◼   ► Ah, so this is one of those topics that's not a topic we get

00:37:11 ◼   ► to leave in the show.

00:37:13 ◼   ► You have to say something.

00:37:15 ◼   ► I don't.

00:37:16 ◼   ► You're are you are you giving me the silent treatment so this doesn't stay

00:37:19 ◼   ► in?

00:37:20 ◼   ► I'm not, but I feel like I'm being kept hostage in a in an

00:37:24 ◼   ► environment that I don't like.

00:37:26 ◼   ► You know?

00:37:26 ◼   ► Which is this.

00:37:27 ◼   ► 'Cause now I feel like it's my n I know what's happened, right?

00:37:32 ◼   ► My displeasure for this has now created a thing, right?

00:37:36 ◼   ► That now there's now nothing I can do about it because you're gonna keep

00:37:40 ◼   ► doing it now because you know I don't like it.

00:37:42 ◼   ► He doesn't even have dogs.

00:37:43 ◼   ► He's this is just a bit to keep you here.

00:37:47 ◼   ► Massive con.

00:37:48 ◼   ► Years and years and years.

00:37:51 ◼   ► Those photos, they're all AI.

00:37:53 ◼   ► Yeah.

00:37:54 ◼   ► The dog photos, none of them are real.

00:37:57 ◼   ► They've been fake the whole time.

00:37:58 ◼   ► The whole time.

00:37:59 ◼   ► The whole time.

00:38:01 ◼   ► Which is pretty impressive, really.

00:38:02 ◼   ► Yeah.

00:38:03 ◼   ► Yeah.

00:38:04 ◼   ► If you if you think about it.

00:38:07 ◼   ► Yeah.

00:38:09 ◼   ► Oh

00:38:10 ◼   ► man.

00:38:11 ◼   ► I haven't seen you in a year.

00:38:14 ◼   ► Yeah, 'cause you failed to come to WWE C.

00:38:20 ◼   ► I already said this in the pro show.

00:38:21 ◼   ► Yeah, it's not not what happened.

00:38:24 ◼   ► I mean you did though.

00:38:25 ◼   ► You didn't make it, So Mm hmm Mm hmm.

00:38:30 ◼   ► You have a lot of boxes in my living room.

00:38:33 ◼   ► Yeah, I know.

00:38:34 ◼   ► I'm terrified.

00:38:35 ◼   ► I don't know how it's gonna go.

00:38:36 ◼   ► I do not think y'all will have time to finish them.

00:38:39 ◼   ► Oh how would how dare you?

00:38:40 ◼   ► Well you don't wanna I mean you don't wanna bring that on yourself.

00:38:43 ◼   ► No, I don't.

00:38:44 ◼   ► I don't know 'cause if we can't finish 'em, someone's gonna help.

00:38:47 ◼   ► But I d I do think you should prioritize the signing.

00:38:51 ◼   ► Oh the signing well the signing will be the first thing that happens when

00:38:55 ◼   ► Jason arrives.

00:38:56 ◼   ► Like but before Jason arrives, I'm going to start the assembling of some of

00:39:00 ◼   ► the packages.

00:39:01 ◼   ► Okay.

00:39:02 ◼   ► I have like a whole system.

00:39:03 ◼   ► Like I'm gonna start putting like cardboard dividers inside of envelopes.

00:39:07 ◼   ► Okay.

00:39:08 ◼   ► And then we can just sign them, put them in the plastic envelopes, and

00:39:11 ◼   ► then I then we will then start packaging them over the days.

00:39:16 ◼   ► Like we'll get the signing done in no time.

00:39:18 ◼   ► Sorry, I'm here.

00:39:19 ◼   ► Hello.

00:39:20 ◼   ► Steven did that thing again, so It's all staying in.

00:39:23 ◼   ► It's all staying in.

00:39:25 ◼   ► And I just need to tell you, Federico, that we made a joke about

00:39:27 ◼   ► the fact that you don't have dogs and all the photos of your dogs

00:39:29 ◼   ► have been AI.

00:39:30 ◼   ► I just need you to know that we made that joke.

00:39:33 ◼   ► Okay.

00:39:33 ◼   ► It unfortunately this is just because of like the fact that Steven is is

00:39:38 ◼   ► keeping the show hostage.

00:39:39 ◼   ► Um Okay.

00:39:41 ◼   ► So, you know, please So now I'm back can I can I answer now?

00:39:44 ◼   ► Please.

00:39:45 ◼   ► Yes.

00:39:45 ◼   ► Okay.

00:39:46 ◼   ► I was uh coming from the M3 Ultra My Studio, um Which I still

00:39:52 ◼   ► have with five twelve gigs of RAM.

00:39:55 ◼   ► Um the main the main thing that this computer has allowed is

00:40:02 ◼   ► being able to use uh the best local models that fit on this machine

00:40:09 ◼   ► with the kind of performance that makes them nice to use in apps.

00:40:15 ◼   ► Like these local models were possi could could be used before in terms of

00:40:22 ◼   ► um fitting on the computer, right?

00:40:24 ◼   ► And even on the map on the on the M3 Ultra, I could even

00:40:27 ◼   ► fit bigger models because I have the 512 gig version.

00:40:31 ◼   ► But even if they technically fit in the machine in in memory and in

00:40:36 ◼   ► storage, they were not necessarily nice to use because they were slow.

00:40:42 ◼   ► And they were slow in a in two different metrics: in how long they

00:40:47 ◼   ► took to answer and when they started answering, how slowly they would go in

00:40:55 ◼   ► just placing text on the screen.

00:40:58 ◼   ► And those are two separate metrics.

00:41:00 ◼   ► The time to first token and the generation.

00:41:04 ◼   ► And so technically I could use local models before, which I did the entire

00:41:11 ◼   ► summer.

00:41:11 ◼   ► The thing that I'm writing now for for Club Mac Stories members, the making

00:41:15 ◼   ► of of the review.

00:41:16 ◼   ► I had a deep seek model running all the time for 99 days on

00:41:22 ◼   ► my Mac Studio, and it was slow, right?

00:41:25 ◼   ► It was incredibly slow.

00:41:27 ◼   ► But it ran all the time and it continuously processed my research material for

00:41:32 ◼   ► iOS twenty-seven.

00:41:33 ◼   ► Right.

00:41:34 ◼   ► But it was not the kind of ex it was not the kind of

00:41:36 ◼   ► model that I would look at and have an interaction with because it would

00:41:40 ◼   ► it would generate text incredibly slowly.

00:41:43 ◼   ► It was doing like agent work basically.

00:41:45 ◼   ► It was doing agent work in the background.

00:41:47 ◼   ► Right.

00:41:47 ◼   ► Like headless work in the background.

00:41:49 ◼   ► Yeah.

00:41:50 ◼   ► I was not looking at it.

00:41:51 ◼   ► Um with this Maxudio you can put local models.

00:41:57 ◼   ► Obviously, right now you're limited by the 256 gigs of RAM, and you can

00:42:03 ◼   ► interact with them in real time like a chat bot and they are pretty

00:42:09 ◼   ► nice to use because of the better GPU and the increased memory bandwidth.

00:42:14 ◼   ► So th those two things together they make local models much, much nicer to

00:42:21 ◼   ► use.

00:42:22 ◼   ► So even though your existing Mac Studio have more RAM, that's actually not the

00:42:28 ◼   ► difference maker.

00:42:29 ◼   ► The RAM the RAM helps in terms of uh loading a bigger model in

00:42:35 ◼   ► memory.

00:42:36 ◼   ► Okay.

00:42:36 ◼   ► But the performance of the GPU and the bandwidth of the memory do not

00:42:42 ◼   ► help in making it nice, where by nice I mean it does it go

00:42:48 ◼   ► fast?

00:42:48 ◼   ► Yeah.

00:42:49 ◼   ► Yeah.

00:42:49 ◼   ► Okay.

00:42:50 ◼   ► So it's like the conversations that you can have are enabled by you said

00:42:54 ◼   ► it was the the the memory's speed?

00:42:57 ◼   ► The speed of the memory.

00:42:58 ◼   ► Yeah.

00:42:58 ◼   ► Uh going from eight hundred and nineteen gigabit per second to one point two

00:43:05 ◼   ► terab terabit per second.

00:43:06 ◼   ► Okay.

00:43:07 ◼   ► Yeah that's quite a difference.

00:43:08 ◼   ► So the the the the local A mo I AI models, you are interacting

00:43:13 ◼   ► with them and doing things with them like you would do with codecs or

00:43:18 ◼   ► claw code or whatever.

00:43:19 ◼   ► It's like all the same capabilities essentially.

00:43:21 ◼   ► Yes.

00:43:21 ◼   ► I but it's running locally.

00:43:23 ◼   ► Yes.

00:43:24 ◼   ► I I created um on the Mac Studio a little uh tail scale server

00:43:31 ◼   ► that basically exposes the local model as uh as an API on my network.

00:43:38 ◼   ► And I added that to my Hermes agent and on your phone?

00:43:44 ◼   ► So essentially it became meaningful.

00:43:48 ◼   ► Sorry, like a server for you.

00:43:49 ◼   ► Like the Mac Studio became like a server that you could reach from anywhere?

00:43:53 ◼   ► Absolutely, yes.

00:43:54 ◼   ► Wow, that's pretty cool.

00:43:55 ◼   ► So then you could act on things that were local to your computer

00:44:01 ◼   ► from anywhere.

00:44:02 ◼   ► Yes.

00:44:02 ◼   ► Yes.

00:44:03 ◼   ► I've been doing I've been doing this with the M3 for for since last

00:44:06 ◼   ► year.

00:44:07 ◼   ► But it wasn't good enough.

00:44:08 ◼   ► But it wasn't good enough.

00:44:09 ◼   ► It was slow.

00:44:10 ◼   ► Uh like you will load up a model and ask a question.

00:44:13 ◼   ► Like for example, I'm using open minis um

00:44:17 ◼   ► on the iPhone

00:44:19 ◼   ► and I I would load up a model on the M three and I

00:44:25 ◼   ► would ask an answer and then start waiting at the screen for like a

00:44:30 ◼   ► minute.

00:44:31 ◼   ► Yeah.

00:44:31 ◼   ► Which is just not a good experience.

00:44:33 ◼   ► No, it's it's not it's not good and maybe maybe I could have optimized

00:44:36 ◼   ► it more.

00:44:37 ◼   ► Um, I don't know, but it was not great.

00:44:40 ◼   ► Here I'm asking a question, I wait for like a couple of seconds and

00:44:43 ◼   ► then it starts answering.

00:44:44 ◼   ► Yeah.

00:44:45 ◼   ► And so that that makes it a lot more practical for these types of

00:44:51 ◼   ► uh uh uh foreground, let's say, uh use cases.

00:44:55 ◼   ► Okay.

00:44:56 ◼   ► Yeah.

00:44:56 ◼   ► And I guess uh the thing the reason this is interesting to me is

00:45:00 ◼   ► I I wonder about the future, right?

00:45:02 ◼   ► Because

00:45:04 ◼   ► it is very impressive, like what you're able to do and the power that

00:45:07 ◼   ► you're able to have with a computer that you can buy.

00:45:10 ◼   ► Like it's a lot of money, but you can just go and buy it

00:45:12 ◼   ► from Apple, right?

00:45:13 ◼   ► Like that is impressive.

00:45:15 ◼   ► But Tech you know, the history of technology would suggest to me that in

00:45:22 ◼   ► a not too distant future

00:45:26 ◼   ► this capability could be available on my laptop.

00:45:30 ◼   ► Yes.

00:45:30 ◼   ► Right?

00:45:31 ◼   ► Oh yeah.

00:45:32 ◼   ► And so I find this interesting because I look at some of the things

00:45:37 ◼   ► that I'm able to do in Claude and it's like, well, wouldn't it just

00:45:42 ◼   ► be nicer if I could just do it locally on my computer?

00:45:45 ◼   ► Like because then it's everything is mine.

00:45:47 ◼   ► It doesn't need to go to anyone.

00:45:49 ◼   ► It can't none of my information can be used for any purpose than the

00:45:53 ◼   ► purpose that I have.

00:45:54 ◼   ► And I would assume that the long term effect on the environment is better.

00:46:00 ◼   ► Now like I understand that like the training still has to happen, which is

00:46:04 ◼   ► one thing, but I would expect it's better if every query is not sent

00:46:09 ◼   ► to a data center for tons of reasons.

00:46:11 ◼   ► And this would be another one.

00:46:13 ◼   ► So like I feel like to me it seems like this is the logical

00:46:17 ◼   ► way that this stuff should happen anyway.

00:46:19 ◼   ► Like going into the future, like we shouldn't be doing cloud for everything, always,

00:46:24 ◼   ► forever.

00:46:25 ◼   ► Yeah, I think some power users w will tell you that that that's possible

00:46:30 ◼   ► now, that what you described with Claude, they have that kind of setup in

00:46:35 ◼   ► Hermes Agent, right?

00:46:37 ◼   ► Or or or OpenClaw for that matter.

00:46:40 ◼   ► Um I think obviously that requires a lot of fiddling, a lot of tinkering,

00:46:46 ◼   ► right?

00:46:46 ◼   ► But but to your point, if s down like let's imagine two, three years

00:46:52 ◼   ► from now where you have a laptop that is capable of running a local

00:46:58 ◼   ► model, right?

00:46:59 ◼   ► Um with with frontier adjacent performance and and there's a nice package for for

00:47:09 ◼   ► all of it, right?

00:47:09 ◼   ► Because all of this stuff right now, you gotta go in and and use

00:47:13 ◼   ► the terminal and you know download projects from somebody's GitHub repo, right?

00:47:20 ◼   ► There's not a nice there's not a nice user experience at all.

00:47:25 ◼   ► Yeah.

00:47:26 ◼   ► But imagine if there was an and arguably like something like LM Studio, right?

00:47:31 ◼   ► Yeah.

00:47:31 ◼   ► Is an example.

00:47:32 ◼   ► It's a it's a packaged experience where you still need some technical know-how in

00:47:39 ◼   ► terms of like okay, well what is a model that that you should download?

00:47:43 ◼   ► But the trajectory it seems pretty clear to me.

00:47:45 ◼   ► Like imagine if there was a nice intuitive package where you go in and

00:47:49 ◼   ► you have like a cloud-like environment, but everything happens locally right?

00:47:53 ◼   ► And maybe that's what Apple wants to do, right?

00:47:55 ◼   ► Maybe that's our plan with Siri AI.

00:47:57 ◼   ► If they ever do a Siri agent, maybe that's the idea that they have

00:48:02 ◼   ► a foundation model model that runs on device and you can choose do you

00:48:05 ◼   ► want to go local, do you want to go to the cloud?

00:48:08 ◼   ► I don't I think that is ob that has got to be their plan.

00:48:12 ◼   ► Y yes.

00:48:13 ◼   ► I mean especially now that these things are happening with muse and other things

00:48:18 ◼   ► that we can talk about later.

00:48:20 ◼   ► But um but yeah, I I think right now on this computer you're you're

00:48:25 ◼   ► getting a glimpse at what could be a future of AI, and and even

00:48:32 ◼   ► if it's not.

00:48:34 ◼   ► I think right now it's it's it's just and it's maybe silly to say,

00:48:38 ◼   ► I think it's just cool.

00:48:39 ◼   ► It is very interesting.

00:48:41 ◼   ► It's it's

00:48:43 ◼   ► and and if you're the kind of person who doesn't find it like conceptually

00:48:47 ◼   ► interesting okay.

00:48:50 ◼   ► Uh uh even if this is going nowhere, I still think it's interesting.

00:48:55 ◼   ► Yeah.

00:48:55 ◼   ► Right.

00:48:56 ◼   ► And so that's why I wanted about it.

00:48:58 ◼   ► I don't I mean to me it makes sense.

00:49:00 ◼   ► So here's the thing like the the the current crop of AI companies, right?

00:49:07 ◼   ► The you know, Anthropic, uh, OpenAI and others, they don't want this, right?

00:49:14 ◼   ► Like this doesn't benefit them.

00:49:16 ◼   ► No.

00:49:17 ◼   ► But Apple.

00:49:18 ◼   ► Especially anthropic.

00:49:19 ◼   ► Especially.

00:49:20 ◼   ► But I mean all of them, right?

00:49:21 ◼   ► Like it's just like this is just not what they want.

00:49:25 ◼   ► Because what it suggests is you don't need to continue paying them the money

00:49:29 ◼   ► you're paying them on an ongoing basis if you can run things locally.

00:49:33 ◼   ► But it does I mean, the reason I mention like I doubled down on

00:49:37 ◼   ► your point about Apple was like, well, one, they're building the hardware to do

00:49:40 ◼   ► it, right?

00:49:41 ◼   ► Specifically they're making the hardware to do it.

00:49:44 ◼   ► They're m obviously making changes and are continuing to make changes to their silicon

00:49:49 ◼   ► to enable this, which means that there must be some kind of thought inside

00:49:53 ◼   ► of the company that this is interesting to them.

00:49:55 ◼   ► And also their their economic model aligns where these other companies' economic model does

00:50:02 ◼   ► not align.

00:50:03 ◼   ► It does not matter to Apple if AI goes local because they have no

00:50:08 ◼   ► current cost they have no current incentive to keep it in the cloud.

00:50:13 ◼   ► It it actually would be better for them cost wise if it all went

00:50:17 ◼   ► local because then they don't have to pay for cloud infrastructure because they don't

00:50:21 ◼   ► want to pay for that cloud infrastructure would be my assumption.

00:50:24 ◼   ► Like they they have no desire to do this.

00:50:25 ◼   ► It is not core to their business.

00:50:27 ◼   ► What is core to their business is offering it.

00:50:30 ◼   ► Now, it would be so much nicer for them if they could say, you

00:50:33 ◼   ► could just do all of this on your device now.

00:50:36 ◼   ► And I just don't think that any of the other AI companies it is

00:50:40 ◼   ► in their interest that they're gonna wanna do this.

00:50:42 ◼   ► And so like I can imagine a world where Apple is providing new models

00:50:47 ◼   ► that run completely locally on your device and they do a bunch of stuff

00:50:50 ◼   ► on your computer.

00:50:50 ◼   ► Like it makes sense to me that they would want to do that.

00:50:54 ◼   ► Yeah.

00:50:54 ◼   ► Yeah.

00:50:55 ◼   ► And it and it would and it also makes sense because then they can

00:50:59 ◼   ► say, Well, we make the best computers to to run AI locally.

00:51:03 ◼   ► Which they're already saying, right?

00:51:04 ◼   ► Like they say that already.

00:51:06 ◼   ► Yeah.

00:51:06 ◼   ► Yes, but they're saying it they're they're saying it to to a niche of

00:51:09 ◼   ► a niche of a niche of their audience.

00:51:11 ◼   ► Yeah.

00:51:12 ◼   ► Yeah, that's my question, uh sort of uh about this.

00:51:15 ◼   ► We've talked about it before, but this is a pretty nerdy thing to do

00:51:20 ◼   ► and while I believe the setup has gotten better, I think if Apple wants

00:51:25 ◼   ► this to be more mainstream, like is there a point where Apple needs to

00:51:29 ◼   ► build tooling that more regular people can use to get these things up and

00:51:34 ◼   ► running?

00:51:35 ◼   ► Yeah.

00:51:35 ◼   ► Um

00:51:37 ◼   ► the whole

00:51:39 ◼   ► MLX thing, I mean obviously it was a project started within Apple, but um

00:51:44 ◼   ► to give an example, I I really wasn't able to

00:51:51 ◼   ► I couldn't test models that were specifically targeting the M5 Ultra and its new

00:51:58 ◼   ► metal kernels.

00:52:00 ◼   ► Because nobody had made them.

00:52:02 ◼   ► Like Apple i is relying arguably a little too much on the community right

00:52:08 ◼   ► now to fix their stuff.

00:52:12 ◼   ► Like if they're serious about this, they should they should you know come up

00:52:18 ◼   ► and say, well, here's a new Mac Studio.

00:52:20 ◼   ► It's got an M7 Ultra, and we're already providing you with optimized kernels for

00:52:26 ◼   ► such and such models.

00:52:28 ◼   ► And these back ends, right?

00:52:30 ◼   ► Because I had to go and right now it's all very fragmented, right?

00:52:34 ◼   ► Apple just makes the computer and there's the MLX framework.

00:52:37 ◼   ► But then it's up to the developer community to figure it out.

00:52:40 ◼   ► And so you gotta choose first of all, you gotta choose like what backend

00:52:44 ◼   ► do you wanna use?

00:52:45 ◼   ► Is it OMLX?

00:52:46 ◼   ► Is it um MLX serve?

00:52:49 ◼   ► Is it something else?

00:52:50 ◼   ► And then you gotta choose the models, and then you gotta optimize the models.

00:52:54 ◼   ► And so like it's all uh i it's a developer th thing right now.

00:53:00 ◼   ► And I think if they're serious about this, like w when I when I

00:53:04 ◼   ► when I mentioned like you gotta package it up in a way that makes

00:53:07 ◼   ► sense, um they have a lot of work to do instead of just saying,

00:53:12 ◼   ► well, just wait for, you know, the developer the the open source community to

00:53:16 ◼   ► figure it out.

00:53:17 ◼   ► But that's not a way to build a sustainable product over time, right?

00:53:23 ◼   ► Um so I I think they should maybe take a look at what Nvidia

00:53:27 ◼   ► is doing with the with the with the sparks, right?

00:53:31 ◼   ► And and be a little more invested into uh reaching out to the community

00:53:38 ◼   ► and for example when a new computer comes out having a more prepared package

00:53:43 ◼   ► um to give out to people like me or other uh you know uh

00:53:47 ◼   ► content creators that I saw on YouTube that have the same review unit and

00:53:52 ◼   ► be like,

00:53:54 ◼   ► here's you know, this thing that you can test that we know is already

00:53:58 ◼   ► optimized for the M5 Ultra, and we know is gonna give you the kind

00:54:02 ◼   ► of performance that is way above the M3 Ultra.

00:54:06 ◼   ► Instead, I had to run a generic vanilla model that still gave me a

00:54:12 ◼   ► performance improvement compared to the M3 Ultra, because I mean the GPU is faster

00:54:16 ◼   ► and the memory bandwidth is faster faster.

00:54:18 ◼   ► But the numbers they could have been better, right?

00:54:23 ◼   ► If only Happa had done more work beforehand to optimize some models for this

00:54:31 ◼   ► chipset.

00:54:33 ◼   ► Is there stuff like that out there now?

00:54:35 ◼   ► Like uh are there Well not see not yet, which means I will have

00:54:40 ◼   ► to revisit my story in a few weeks when more developers actually get their

00:54:45 ◼   ► M5 ultras and early experiments that I've seen, I think Max uh Wainbach at

00:54:50 ◼   ► Creative Strategies uh posted some examples.

00:54:53 ◼   ► Uh he's been optimizing.

00:54:55 ◼   ► Um using using uh Claude Opus five five and GPT.

00:55:01 ◼   ► Astra, he's been manually optimizing the metal kernel for the M5 Ultra.

00:55:07 ◼   ► And uh he got like a like a 50% improvement already over the numbers

00:55:12 ◼   ► that I got with my vanilla setup from last week.

00:55:15 ◼   ► Um so I think this is the kind of right now it's the kind

00:55:19 ◼   ► of thing where you need to follow up on it in a month or

00:55:23 ◼   ► so and see how the community has evolved around it.

00:55:27 ◼   ► I would love to see Apple be a little bit uh more involved in

00:55:31 ◼   ► this entire process, honestly.

00:55:33 ◼   ► Yeah.

00:55:34 ◼   ► Okay.

00:55:37 ◼   ► This episode of Connected is brought to you by Sentry.

00:55:41 ◼   ► You get a bug report.

00:55:42 ◼   ► The user says something is broken, but of course you can't reproduce it.

00:55:47 ◼   ► This is how these things go.

00:55:49 ◼   ► So you dig through logs, check the traces, and still have no idea what

00:55:53 ◼   ► they actually saw.

00:55:55 ◼   ► Sentry Century session replay fixes that.

00:55:58 ◼   ► It gives you a video-like reproduction of exactly what the user experienced.

00:56:02 ◼   ► Their clicks, scrolls, what loaded, what didn't it, all tied to the error in

00:56:07 ◼   ► one timeline.

00:56:09 ◼   ► You can see the console output, the network calls, and even rage clicks where

00:56:13 ◼   ► the user hammered a broken button.

00:56:15 ◼   ► And it's built into Sentry, so the replay is already linked to the error,

00:56:19 ◼   ► trace, and logs.

00:56:20 ◼   ► There's no hunting required.

00:56:23 ◼   ► It's really important to have insights into issues when you're building apps.

00:56:29 ◼   ► Apps can be complicated.

00:56:31 ◼   ► You have a lot going on.

00:56:32 ◼   ► You have a bunch of users doing different things.

00:56:34 ◼   ► And it can be really hard to track down an issue.

00:56:37 ◼   ► But century puts all this data together to make it easy.

00:56:41 ◼   ► And it's all private.

00:56:43 ◼   ► Privacy is on by default with Century.

00:56:45 ◼   ► Sensitive text and media is masked, so you're never capturing what you shouldn't be.

00:56:49 ◼   ► And it works on web and mobile, React, Next.js, iOS, Android, Flutter, you name

00:56:55 ◼   ► it.

00:56:56 ◼   ► Millions of developers use Century, including teams at Anthropic and Disney Plus.

00:57:01 ◼   ► So try it for free at Century.io.

00:57:04 ◼   ► That's That's S-E-N-T-R-Y, century.io, and tell them that we sent you.

00:57:10 ◼   ► They have a free dev plan, and listeners of this show can use the

00:57:14 ◼   ► code CONNECTID26 to get $100 in sentry credits when you're a new user.

00:57:21 ◼   ► That's code connected twenty six at century.io for a hundred dollars in credits.

00:57:27 ◼   ► My thanks to Century for their support of the show and all of Relay.

00:57:33 ◼   ► So this week, Meta had their Meta Connect uh conf fr it's like WWC,

00:57:38 ◼   ► right?

00:57:40 ◼   ► But they use it as a way sometimes to announce a bunch of products

00:57:42 ◼   ► and they announced a bunch of products.

00:57:44 ◼   ► Uh I wanna talk about Muse, but I wanna talk about that last.

00:57:48 ◼   ► Uh I want to just run through some of the products that they announced.

00:57:51 ◼   ► And then I need Muse is like I'm aware that this thing exists, but

00:57:56 ◼   ► like I haven't looked into it very much.

00:57:57 ◼   ► But like I've read a couple of stories, but I feel like there's something

00:58:00 ◼   ► going on there.

00:58:01 ◼   ► So they announced a bunch of new Ray bands.

00:58:04 ◼   ► So they have new styles of their like glasses, right?

00:58:09 ◼   ► Like new glasses styles, more collaborations.

00:58:13 ◼   ► They have a better version of the camera glasses.

00:58:15 ◼   ► They spent a ton of time talking about privacy and security upgrades.

00:58:20 ◼   ► Like even Meta can't see the video.

00:58:22 ◼   ► Like uh and I also really recommend Joanna Stern had a great interview with

00:58:26 ◼   ► with Mark Zuckerberg.

00:58:28 ◼   ► Um, in the way in which Mark Zuckerberg can have a great interview, this

00:58:32 ◼   ► is one, but like every time he has an interview, there's like you can

00:58:36 ◼   ► see it in his eyes when he's talking about the things he's been told

00:58:40 ◼   ► to say and then the things he wants to say.

00:58:43 ◼   ► It's like a very f I think a fascinating thing with him where you

00:58:46 ◼   ► can see when he is repeating the lines and then when he is actually

00:58:50 ◼   ► interested in the question.

00:58:52 ◼   ► Um and also Joanna just like she hits him with some stuff which is

00:58:56 ◼   ► like I say hats off to her 'cause she just goes at him in

00:58:59 ◼   ► certain points.

00:59:00 ◼   ► It's it's really good interview.

00:59:01 ◼   ► But anyway, they talk a lot about that.

00:59:02 ◼   ► Like she kinda just asks him, like, how are you gonna beat what people

00:59:06 ◼   ► say about Meta?

00:59:07 ◼   ► And he has he has answers which are interesting to hear, but I don't

00:59:11 ◼   ► know if they're gonna work the way he thinks they will, of like getting

00:59:13 ◼   ► people to trust Meta in the way that he wants.

00:59:15 ◼   ► Anyway, the new Ray Bans have better mics, better battery, slimmer designs.

00:59:19 ◼   ► They also released audio glasses, no camera.

00:59:23 ◼   ► So that's one, right?

00:59:25 ◼   ► That there's no camera on them.

00:59:27 ◼   ► But because there's no camera on them, they can also have some frame styles

00:59:31 ◼   ► which are more stylish and look more like regular glasses because it needs less

00:59:35 ◼   ► technology.

00:59:36 ◼   ► What they can do they're basically they call them audio glasses.

00:59:40 ◼   ► So you can listen to music, you can take calls, you can talk to

00:59:43 ◼   ► your AI.

00:59:44 ◼   ► They're super light.

00:59:45 ◼   ► They're still three forty nine.

00:59:47 ◼   ► They come in a bunch of price uh styles, but

00:59:50 ◼   ► it's interesting to have this as an option.

00:59:52 ◼   ► Also, Meta have uh their all of their glasses, I believe, or at least

00:59:56 ◼   ► all the new ones, are gonna be uh they're gonna work as hearing aids,

01:00:00 ◼   ► which is pretty cool.

01:00:01 ◼   ► Very cool.

01:00:02 ◼   ► Um and and I was thinking like, Oh, AirPods work as hearing aids.

01:00:06 ◼   ► But one of the things that Zuckerberg said, which I thought was a very

01:00:08 ◼   ► good point, is your glasses these glasses just look like glasses.

01:00:13 ◼   ► You never look like you're wearing anything you shouldn't.

01:00:16 ◼   ► Because if you use your AirPods as hearing aids, that's great, but you're still

01:00:19 ◼   ► wearing AirPods, right?

01:00:20 ◼   ► Which makes it look like you're listening to something.

01:00:23 ◼   ► But these are just eyeglasses, essentially, if you wear the audio ones.

01:00:28 ◼   ► And so I thought that was pretty cool.

01:00:30 ◼   ► They also had something called Meta VR glasses, which is their answer to the

01:00:34 ◼   ► Vision Pro, which is hilarious that it's actually not a Quest

01:00:39 ◼   ► it's like it's the it's meta VR glasses, it's not Quest product, but it

01:00:44 ◼   ► has access to all of the Quest apps and games.

01:00:49 ◼   ► So they're clearly getting rid of the Quest brand, right?

01:00:53 ◼   ► Like they are moving away from that.

01:00:55 ◼   ► Um, full VR, color pass through, they're open.

01:00:59 ◼   ► So I don't know how the full VR works.

01:01:01 ◼   ► That is an interesting wrinkle that I don't feel like I have an answer

01:01:04 ◼   ► to yet.

01:01:06 ◼   ► Um 5K OLED displays at 120 Hertz.

01:01:09 ◼   ► They weigh 100 grams.

01:01:10 ◼   ► It offloads the battery and all the compute to a wire-connected puck,

01:01:15 ◼   ► can connect to a Mac or a PC for virtual display, has their own

01:01:19 ◼   ► versions of personas coming in spring twenty twenty-seven for one thousand three hundred dollars.

01:01:25 ◼   ► So

01:01:27 ◼   ► I don't know what this experience is like.

01:01:28 ◼   ► People that have been there say it's very good experience, but it is very

01:01:31 ◼   ► much like this is way lighter than a Vision Pro and way cheaper than

01:01:34 ◼   ► a Vision Pro.

01:01:35 ◼   ► Uh which so it's an interesting product.

01:01:37 ◼   ► I don't know if anyone wants this product though.

01:01:39 ◼   ► Just in general, but it is now another form factor for this kind of

01:01:44 ◼   ► experience.

01:01:45 ◼   ► Um, so that's the hardware.

01:01:48 ◼   ► They also spent a lot of time talking about Muse.

01:01:51 ◼   ► So

01:01:52 ◼   ► Muse is what an AI agent app, right?

01:01:55 ◼   ► I guess is the way to describe it.

01:01:57 ◼   ► Like it is a separate app called Muse, which is number one in the

01:02:01 ◼   ► app store in the US.

01:02:02 ◼   ► I think it's US only at the moment.

01:02:05 ◼   ► Um and it feels like it's gaining in interest and popularity, like people are

01:02:11 ◼   ► using it.

01:02:12 ◼   ► I read a great article from Alison Johnson at The Verge,

01:02:16 ◼   ► and she talks about she basically goes through a series of household and life

01:02:21 ◼   ► tasks that she needed to do and she asked Muse to help her with

01:02:25 ◼   ► and it was either able to complete them or move them along in a

01:02:30 ◼   ► significant way without tak without basically any effort from her.

01:02:34 ◼   ► I've also read some stories about like

01:02:38 ◼   ► it's kind of unsure what happens when you put this thing on your Mac

01:02:42 ◼   ► in some scenarios.

01:02:43 ◼   ► So Jason Ayton of Inc.

01:02:45 ◼   ► wrote an article about how he put the Muse app on his Mac and

01:02:50 ◼   ► Yeah, that's kinda weird.

01:02:51 ◼   ► And it was reading his iMessages and he doesn't know how it was reading

01:02:56 ◼   ► his iMessages.

01:02:59 ◼   ► Because the s the the the the bot was saying that it was reading

01:03:05 ◼   ► notifications, but he was able to find evidence somehow I don't remember off the

01:03:10 ◼   ► top of my head, I read the article, but I don't remember exactly how

01:03:12 ◼   ► he said it.

01:03:13 ◼   ► He was able to see that it had processed his iMessage back up.

01:03:18 ◼   ► But he did not grant the app full disk access.

01:03:21 ◼   ► Now, my feeling on this is this thing is just doing something.

01:03:26 ◼   ► How it works, I don't know.

01:03:27 ◼   ► A lot of these tools are just doing things, but this is a story

01:03:31 ◼   ► to me of why I think it's interesting that Meta have what seems to

01:03:37 ◼   ► be like a a buzzy maybe hit product on its hand hands against the

01:03:42 ◼   ► general mistrust of this company, right?

01:03:47 ◼   ► That like something like this can happen, like with what Jason found, and you're

01:03:51 ◼   ► like, Well of course Meta's doing that.

01:03:53 ◼   ► Right?

01:03:53 ◼   ► Like it's just like the thing that everybody jumps to.

01:03:56 ◼   ► And so I think it is it's an interesting is interesting.

01:04:00 ◼   ► Now in the interview between Zuck and uh Joanna Stern, she kind of presses

01:04:06 ◼   ► him on this exact issue of like I've been intro I've been using Muse

01:04:11 ◼   ► and I find it very impressive, but I'm new access to my Gmail because

01:04:14 ◼   ► I don't trust Meta.

01:04:16 ◼   ► And he gives an answer about how it's like you're in a virtual machine,

01:04:19 ◼   ► they have a bunch of security protocols.

01:04:21 ◼   ► And I kind of like conceptually believe

01:04:24 ◼   ► that like they probably they don't want to steal information from you because it

01:04:30 ◼   ► will come out and they'll look bad, right?

01:04:32 ◼   ► Like if you think about it just like conceptually like what are they gonna

01:04:34 ◼   ► do with your Gmail information?

01:04:35 ◼   ► I mean they could probably do a lot, but if they do steal all

01:04:38 ◼   ► of your messages,

01:04:40 ◼   ► people are gonna find out and it's only gonna look bad for them, but

01:04:44 ◼   ► also I also don't trust them.

01:04:46 ◼   ► Like I also wouldn't do this.

01:04:48 ◼   ► So it's a v they're in a very weird spot.

01:04:52 ◼   ► So I'm sure Federico, you have a lot more information kind of about what's

01:04:56 ◼   ► going on with Muse.

01:04:59 ◼   ► Yeah, so I have not been able to use Muse myself, uh because it's

01:05:04 ◼   ► not available in my country yet.

01:05:07 ◼   ► Yeah.

01:05:07 ◼   ► Um I will say up front

01:05:11 ◼   ► that uh uh this whole scenario with meta and uh what's it called?

01:05:16 ◼   ► MSL, like meta super intelligence labs.

01:05:18 ◼   ► Yeah.

01:05:19 ◼   ► It's like the meme of like the worst guy you know just said something

01:05:24 ◼   ► that is accurate.

01:05:26 ◼   ► Um in the sense that I I fundamentally dislike this company,

01:05:32 ◼   ► but objectively speaking, and and I'm tired of having you know n not saying

01:05:39 ◼   ► what seems to me t to be a pretty objective thing.

01:05:42 ◼   ► But just really speaking, they have turned things around in their AI department.

01:05:47 ◼   ► Okay.

01:05:47 ◼   ► Um

01:05:49 ◼   ► They are now the third American lab for AI in, you know, the community

01:05:56 ◼   ► agrees on all the benchmarks.

01:05:57 ◼   ► Uh uh Muse Spark is the name of their model.

01:06:02 ◼   ► Um his companies love the word spark.

01:06:05 ◼   ► It's everywhere.

01:06:06 ◼   ► Yeah.

01:06:07 ◼   ► Um I mean it they have whatever, you know, we can talk about the

01:06:12 ◼   ► layoffs and the crazy salaries that they promise people.

01:06:16 ◼   ► But But whatever, you know, they trained their model by collecting data from the

01:06:22 ◼   ► computer screens of meta employees.

01:06:25 ◼   ► But the resulting model

01:06:28 ◼   ► is great.

01:06:29 ◼   ► And I I have used uh Muse Spark.

01:06:33 ◼   ► It's a good model.

01:06:35 ◼   ► It's it's really good.

01:06:37 ◼   ► What I think is fascinating here is that the one company that I dislike

01:06:44 ◼   ► Uh they have what I think right now the after Chad GPT, or maybe

01:06:50 ◼   ► me in some cases even better than Chad GPT, they have the one

01:06:55 ◼   ► AI product that is understandable enough for regular people.

01:06:59 ◼   ► Yeah.

01:07:00 ◼   ► I don't I don't think it's a coincidence that Muse, the app, it's very

01:07:06 ◼   ► confusing.

01:07:06 ◼   ► Muse Spark is the model, Muse is the app.

01:07:10 ◼   ► Muse, the app, has been the number one app on the US app store,

01:07:16 ◼   ► a spot that's been basically occupied by Chat GPT for like the past couple

01:07:21 ◼   ► of years.

01:07:23 ◼   ► For uh we're going into week two, I believe.

01:07:26 ◼   ► Uh Um, is it gonna be a fad?

01:07:29 ◼   ► I don't know.

01:07:30 ◼   ► But I have a strong feeling that after many years they have had their

01:07:38 ◼   ► most successful product launch.

01:07:42 ◼   ► In a while.

01:07:43 ◼   ► I mean you can tell because it's all they're talking about.

01:07:47 ◼   ► Yes.

01:07:47 ◼   ► Like it's all Meto is talking about is is Muse right now.

01:07:52 ◼   ► They won't stop talking about it.

01:07:53 ◼   ► Like they're treating it like it's operating system for everything now.

01:07:57 ◼   ► Like it they they they know they've hit something.

01:07:59 ◼   ► And I kind of understand it because it's like just from a simple level,

01:08:02 ◼   ► it's oh, it does stuff for you.

01:08:05 ◼   ► Yes.

01:08:06 ◼   ► And I think it's so if you zoom out a little.

01:08:10 ◼   ► I think this is what all the AI companies, except maybe Anthropic, Anthropic doesn't

01:08:16 ◼   ► care about consumers.

01:08:17 ◼   ► Um, but all these AI companies are doing this now, which is

01:08:23 ◼   ► We have focused on the models.

01:08:26 ◼   ► And then we have done business things.

01:08:30 ◼   ► And then we have done um the coding agents stuff.

01:08:35 ◼   ► And now we want to do personal agents.

01:08:38 ◼   ► But in a simplified way.

01:08:41 ◼   ► I think it's interesting that if you put uh uh the rumor has it

01:08:46 ◼   ► that maybe today or maybe next week, OpenAI will come out with uh their

01:08:51 ◼   ► own take on on this, their own answer to Grokbot and Muse.

01:08:57 ◼   ► I think if you lay it out, you know, all things, you know, evaluating

01:09:01 ◼   ► the products, I think from from a feature perspective, from a technical perspective, Grokbot

01:09:08 ◼   ► should be the more popular product.

01:09:11 ◼   ► Yeah, but like if we're talking about brands people don't like.

01:09:15 ◼   ► Exactly.

01:09:16 ◼   ► Exactly.

01:09:16 ◼   ► And it's the number two hundredth app on the app store.

01:09:20 ◼   ► Yeah.

01:09:20 ◼   ► Right.

01:09:22 ◼   ► Muse is not this is a perfect example of like has fewer features and

01:09:28 ◼   ► people love it.

01:09:29 ◼   ► Um honestly, I think one of the reasons is the characters are cute.

01:09:34 ◼   ► Exactly.

01:09:35 ◼   ► Yeah.

01:09:36 ◼   ► And maybe the AI bros don't wanna hear this.

01:09:40 ◼   ► But and I keep saying this, if you want to package AI in a

01:09:44 ◼   ► way that regular people understand it, you gotta do things like silly characters.

01:09:51 ◼   ► And and to prove it, they're bringing out essentially a Tamagotchi for this thing

01:09:55 ◼   ► somehow before the holidays, which if they do this like incredible work, right?

01:10:01 ◼   ► Like by everyone.

01:10:02 ◼   ► It does undermine something.

01:10:03 ◼   ► I'm sorry, To jump a bit.

01:10:05 ◼   ► It under it made me think of something.

01:10:06 ◼   ► It undermines something that Mark says in the video.

01:10:08 ◼   ► So Joanna Stern says, are the audio glasses a response to the pervert glasses

01:10:14 ◼   ► thing?

01:10:14 ◼   ► And he's like, No way, we have to we have to spend years developing

01:10:17 ◼   ► these things.

01:10:18 ◼   ► You just don't do a camera.

01:10:21 ◼   ► He said it you know, it takes years.

01:10:23 ◼   ► So he's w I'm just saying he what he said is it takes years

01:10:25 ◼   ► to develop these products, but they weren't this little Tamagotchi, were they?

01:10:29 ◼   ► Now I assume this is a much more simple thing, like But you know,

01:10:33 ◼   ► it's just like a fun th they have they have done this in response

01:10:35 ◼   ► to what they have seen.

01:10:36 ◼   ► There was no way they were making this little Tamagotchi thing before Muse took

01:10:40 ◼   ► off.

01:10:40 ◼   ► Like you wouldn't do this, I think.

01:10:42 ◼   ► Um but maybe maybe I'm missing something.

01:10:45 ◼   ► But yeah.

01:10:46 ◼   ► And so I think they have tapped into a portion of the market that

01:10:51 ◼   ► has heard of AI and has used AI, you know, to and to to

01:10:58 ◼   ► to have answers to questions.

01:11:01 ◼   ► And instead, they have created a product that is accessible enough for people to

01:11:08 ◼   ► think about it in terms of, oh, this is AI that does things.

01:11:12 ◼   ► If anything,

01:11:14 ◼   ► This successful and maybe unexpected launch, again the rumor has it that OpenAI was

01:11:20 ◼   ► really caught by surprise by both Meta and SpaceX, and which is why they're

01:11:26 ◼   ► rushing out their answer before or at their dev day conference next week.

01:11:33 ◼   ► But they should have didn't they buy OpenClaw?

01:11:36 ◼   ► Well they didn't buy OpenClaw.

01:11:37 ◼   ► They hired the guy they bought OpenClaw.

01:11:40 ◼   ► This is the way acquisitions in AI work.

01:11:42 ◼   ► Open claw is OpenClaw is still an open source foundation, actually.

01:11:46 ◼   ► Sure.

01:11:47 ◼   ► They bought the guy and all the technology like you know what I mean?

01:11:49 ◼   ► Like they actually have like employees and and stuff.

01:11:53 ◼   ► I believe it.

01:11:54 ◼   ► I believe it.

01:11:55 ◼   ► But Anyway.

01:11:57 ◼   ► Anyway, I think

01:12:00 ◼   ► maybe they thought it was gonna be more of a geeky thing and not

01:12:04 ◼   ► for the masses.

01:12:05 ◼   ► But if anything, I think it shows a failure on on OpenAI's part to

01:12:10 ◼   ► understand

01:12:12 ◼   ► because they have done all these things, right?

01:12:14 ◼   ► They had Chad GPT work, but I think their failure was to think of

01:12:18 ◼   ► these things all in the context of the office.

01:12:21 ◼   ► Chat GPT work only came after co-work.

01:12:25 ◼   ► Opening article is not a good product company.

01:12:29 ◼   ► What do you mean?

01:12:29 ◼   ► They did that video with Johnny Ive in San Francisco.

01:12:33 ◼   ► They are not but like they are not a good product company.

01:12:35 ◼   ► They they had one good product idea that was a that was a mistake

01:12:38 ◼   ► in the first place, right?

01:12:40 ◼   ► Like they didn't mean for Chat GPT to be what ChatGPT is.

01:12:44 ◼   ► Like th they didn't think it was going to be this huge consumer product,

01:12:47 ◼   ► but people were interested in it.

01:12:49 ◼   ► I think they don't have their finger on the consumer pulse as much as

01:12:54 ◼   ► meta does because they make Instagram.

01:12:57 ◼   ► Sure.

01:12:58 ◼   ► And they know how to approach brands and celebrities and regular people.

01:13:05 ◼   ► Instead open AI, and I mean this in the best possible lovable way, they're

01:13:10 ◼   ► a bunch of nerds, right?

01:13:12 ◼   ► And

01:13:14 ◼   ► I think

01:13:16 ◼   ► they will need to s switch up their strategy a little bit more if

01:13:20 ◼   ► they want to have a massive consumer success.

01:13:25 ◼   ► And I'm really curious to see what happens with the international rollout of Muse.

01:13:32 ◼   ► Um I am seeing Like I I'm seeing regular people talk about it.

01:13:39 ◼   ► I mean, the number one app on the App Store, people notice it.

01:13:43 ◼   ► People go in there and download it.

01:13:46 ◼   ► And Meta obviously has the advantage of being able to push this, you know,

01:13:49 ◼   ► across multiple surfaces.

01:13:51 ◼   ► Yeah.

01:13:51 ◼   ► WhatsApp, uh Instagram, Facebook for that matter.

01:13:55 ◼   ► They're doing the thing I saw on Instagram where they put Muse in a

01:13:59 ◼   ► button, like in a placement of a button that people used all the time,

01:14:03 ◼   ► which is like the darkest of dark patterns.

01:14:06 ◼   ► But uh you know, some product manager at Meta's gonna raise for it.

01:14:10 ◼   ► So yeah, it works.

01:14:11 ◼   ► Yeah.

01:14:11 ◼   ► It's like what they do with threads too, right?

01:14:12 ◼   ► It's just like we're putting half of a thread inside inside of his so

01:14:17 ◼   ► you tap it to read the rest of it?

01:14:19 ◼   ► I mean, well this is what you do when you have a big product,

01:14:21 ◼   ► right?

01:14:21 ◼   ► So you leverage it.

01:14:22 ◼   ► Like

01:14:23 ◼   ► it's a thing.

01:14:25 ◼   ► To to summarize

01:14:27 ◼   ► I think we all

01:14:30 ◼   ► Collectively made fun of how meta fell behind in AI, their previous llama models

01:14:39 ◼   ► didn't really work out, and then they lost a bunch of people, and then

01:14:42 ◼   ► they started hiring people with.

01:14:44 ◼   ► Then they faked the benchmark.

01:14:46 ◼   ► They faked a benchmark for Lama Four and they started hiring people promising like,

01:14:51 ◼   ► Oh, we're gonna give you like a hundred million dollars.

01:14:54 ◼   ► And they hired Alexander Wang and and people thought what's this twenty-nine year old

01:14:59 ◼   ► guy gonna do for you?

01:15:01 ◼   ► Look, I don't know how these things worked internally, but now they got the

01:15:05 ◼   ► product to show for it.

01:15:06 ◼   ► They got the product, yeah.

01:15:07 ◼   ► And they got the models.

01:15:09 ◼   ► And you know.

01:15:11 ◼   ► I think it if anything, I think it's gonna be interesting to observe.

01:15:15 ◼   ► Uh Anthropic is out of this consumer game, I think.

01:15:20 ◼   ► Um They've found a I think a pretty nice business for themselves.

01:15:24 ◼   ► I agree.

01:15:25 ◼   ► I agree.

01:15:26 ◼   ► And I think uh, you know, uh obviously Apple continues to be the dark

01:15:30 ◼   ► horse in in this entire story with Siri.

01:15:33 ◼   ► Uh, but right now uh the fight between number one and number two app

01:15:39 ◼   ► on the app store, I think is gonna be an interesting one to watch.

01:15:44 ◼   ► Yeah.

01:15:45 ◼   ► I am very keen to try and use 'cause I just want to see

01:15:48 ◼   ► what it's about.

01:15:48 ◼   ► Like after I read, I really recommend people read Allison Johnson's article because I

01:15:53 ◼   ► was like that sounds pretty nice actually.

01:15:56 ◼   ► Like it's just doing a bunch of stuff for her.

01:15:58 ◼   ► It's like you have to give up a lot, like potentially, but it's doing

01:16:02 ◼   ► things.

01:16:02 ◼   ► It's like I kind of like the sound of that.

01:16:04 ◼   ► So Yeah.

01:16:05 ◼   ► Uh it was very funny to hear Mark Zuckerberg talk about all the things

01:16:09 ◼   ► he has Muse do and I'm like, I'm sure you could do these things,

01:16:12 ◼   ► Mark, but you probably also have like a hundred and fifty assistants.

01:16:15 ◼   ► Yeah.

01:16:16 ◼   ► Like actual human assistants that are definitely the ones doing things for you.

01:16:20 ◼   ► Like

01:16:22 ◼   ► Like you don't need to speak to the AI.

01:16:24 ◼   ► You could just have the human do it in like the world.

01:16:27 ◼   ► Um, but nevertheless, yeah.

01:16:30 ◼   ► Um nevertheless.

01:16:33 ◼   ► All right.

01:16:34 ◼   ► I think that's a show.

01:16:35 ◼   ► Yeah.

01:16:36 ◼   ► I feel like we did it.

01:16:37 ◼   ► If you uh are sad next week, it's Thursday and you don't hear us,

01:16:43 ◼   ► uh, that is because we will be off next week for the podcastathon.

01:16:48 ◼   ► But you can see Mike and I for twelve hours next Friday.

01:16:52 ◼   ► There's a link in the show notes to the YouTube video.

01:16:54 ◼   ► You can get notified when we go live.

01:16:56 ◼   ► It's gonna be a real fun time.

01:16:58 ◼   ► If you want to support St.

01:16:59 ◼   ► Jude, then why wouldn't it?

01:17:01 ◼   ► You you want to go to St.

01:17:02 ◼   ► Jude.org slash relay and donate today.

01:17:04 ◼   ► Sign up to be a fundraiser.

01:17:05 ◼   ► Help us get to our goal of $700,000.

01:17:09 ◼   ► If you want to find stuff, uh links to things we spoke about, they're

01:17:13 ◼   ► in your podcast player and on the web at relay.fm slash connected slash six

01:17:18 ◼   ► twenty-two.

01:17:19 ◼   ► There you will find a link to joining at Connected Pro, the longer and

01:17:22 ◼   ► ad-free version of the show, and to leave us feedback, leave us a note.

01:17:26 ◼   ► You can find Federico's work at maxstories dot net.

01:17:30 ◼   ► You can find Mike's writing at the enthusiast dot net.

01:17:33 ◼   ► You've been blogging, Mike.

01:17:34 ◼   ► It's like two posts recently.

01:17:35 ◼   ► It's you keep calling me out, so you know.

01:17:37 ◼   ► I I do.

01:17:38 ◼   ► Yes.

01:17:39 ◼   ► Mm-hmm.

01:17:39 ◼   ► Blogging is the future of all media.

01:17:41 ◼   ► Mm-hmm.

01:17:42 ◼   ► And you can find my writing at five twelvepixels.net.

01:17:45 ◼   ► Did I hear a disagreement about blogging being the future of all media?

01:17:48 ◼   ► Yes.

01:17:49 ◼   ► No, not from me.

01:17:50 ◼   ► From me.

01:17:51 ◼   ► Absolutely yes, from me.

01:17:52 ◼   ► I believe that it is not the future of all media.

01:17:55 ◼   ► Wow.

01:17:55 ◼   ► I'm sorry.

01:17:56 ◼   ► I I I know it means a lot to you to you, but I

01:18:00 ◼   ► I just don't think that it is.

01:18:01 ◼   ► You know, I love blogging.

01:18:03 ◼   ► Blogging's great, but I don't think it's the future.

01:18:06 ◼   ► Rough.

01:18:07 ◼   ► I like to thank our sponsors this week, Scribe and sentry for the support

01:18:12 ◼   ► of the show.

01:18:12 ◼   ► And until uh two weeks from now, uh till two weeks from now, guys,

01:18:18 ◼   ► catch.

01:18:18 ◼   ► Say goodbye.

01:18:20 ◼   ► Arrivederci.

01:18:21 ◼   ► Cheerio.

01:18:21 ◼   ► Bye, y'all.