Today on the Newcomer Podcast, we’re joined by Tony Fadell, one of the most influential figures in modern hardware design. Fadell helped bring some of the most important consumer electronics to life and has shaped how people interact with technology.We talk about where the next major tech device might come from, whether it’s a pin, a pen, headphones, or the device already in your pocket, and how Apple and other major tech companies are approaching the future of hardware.We also discuss the rumors surrounding Fadell as a potential contender for the next CEO of Apple, what he would do if he were in that role, and how leadership decisions at that level actually get made. Fadell shares his view on why OpenAI is pursuing a strategy of becoming too big to fail, and what that signals about the next phase of the industry.This conversation goes beyond product launches and press releases, focusing instead on how power, scale, and design choices shape the tech ecosystem.This is the Newcomer Podcast.🎙️ New episodes every week! Subscribe and turn on notifications to stay ahead of the next big story.👇 Watch more from The Newcomer Podcast:https://www.youtube.com/playlist?list=PL0Yg1id5olJMOyUnYF5AZnIo2ZXdyYd-L 🔗 Read more at newcomer.co🐦 Follow us: @NewcomerMediaIf you want honest, insider analysis from the heart of tech and venture capital…you’re in the right place.
00:00:00
Apple has never done marketing bullshit before and I saw that
00:00:03
was complete bullshit. I mean, do you think Opening Eye
00:00:05
goes bankrupt? Look, if they're asking for $50
00:00:08
billion now, like how fast you're going to burn through
00:00:10
that. Is that what you would focus on
00:00:12
if you're Apple CEO? I have many things that I would
00:00:17
do if I. Was all right, give us some of
00:00:19
them. Is the next big tech device a
00:00:21
pin, a pen, your headphones or the device that you've already
00:00:25
got in your pocket? Today's guest is Tony Fadell,
00:00:28
godfather of hardware design, who brought some of the most
00:00:31
important consumer electronics to life at Apple and then at
00:00:35
Google with Nest, I sat with him and got his thoughts on how
00:00:38
Apple and the rest of the big technology companies are
00:00:40
looking, even as he's rumored to be a contender for Apple's next
00:00:44
CEO. We talked about what he would do
00:00:47
if he ever got that top job. We talked about why he thinks
00:00:50
open AI is building a strategy of being too big to fail and all
00:00:53
those unfiltered thoughts without the constraints of a
00:00:56
corporate PR department. This is the Newcomer podcast.
00:01:07
Tony, Fidel, great to have you on the podcast.
00:01:09
Thanks for joining me. It's great to be here, great to
00:01:11
be. Eric, we're in an insane moment
00:01:14
for the future of devices. So I'm super excited to talk to
00:01:18
you. Pins pens, like what do we what
00:01:21
do we want? So I really want to get direct
00:01:23
into it. I mean, there are rumors I think
00:01:25
this week that Apple is thinking about a pin in the wake of
00:01:29
humane working on a pin. There's I have, you know,
00:01:33
friends and I think there's been some reporting now that maybe
00:01:36
Johnny, I have an open AI are thinking about a pen.
00:01:39
I'll I'll try and press you on what you think about the
00:01:41
particulars. But like from first principles,
00:01:44
what do you think is most likely to be the AI native device
00:01:50
format? AI can be applied to many
00:01:53
things, and AI is great for certain applications with
00:02:00
certain hardware components around it.
00:02:03
So there's AI with a screen, and there's AI without a screen.
00:02:06
In all cases, it needs to have audio, right?
00:02:10
But what AI needs mostly is it needs context.
00:02:15
OK, so there you have to look at the input side and you have to
00:02:18
look at the output side. So first let's talk about the
00:02:21
input side. The more context you give any
00:02:28
kind of LLM, and I'm not saying they're the panacea, there's
00:02:31
going to be lots of models beyond LLM.
00:02:33
So I want to make sure I'm very clear and I'm not the LLM.
00:02:37
Yeah, yeah, yeah. It's, we're going to get AGI
00:02:39
with it. It's all fucking bullshit.
00:02:41
OK, Mike, I'm concerned. So you first start with the
00:02:46
input. And if we know everything and
00:02:48
you've been hearing about over, over the year, 2 years, so fast
00:02:53
here now it's like we've been hearing about context windows
00:02:56
and context windows getting longer and longer and longer.
00:02:59
And then now we're ragging all your data and, and, and gem.
00:03:02
And I just announced yesterday that you can bring in all of
00:03:05
your Gmail and, and all your Google Photos and and G Drive.
00:03:09
And I feel like I'm being reckless.
00:03:10
I've connected everything to Anthropic already.
00:03:12
It's like, yeah, search through my e-mail.
00:03:14
Right, exactly. And so the more context, the
00:03:17
better. So there's one which is context
00:03:20
about what you've done, OK, and that's you know, and that makes
00:03:24
it better. But then there's the context of
00:03:26
what you're doing and knowing that really, really clearly.
00:03:30
And So what you want and you know, what your, your smartphone
00:03:35
has is you want to know your and just like advertising and all
00:03:38
the other, you know, websites and social wanted before, it's
00:03:42
no different. They wanted your location.
00:03:45
They wanted, hey, they would love your audio all the time.
00:03:47
They want your camera all the time.
00:03:49
They, they want what you're typing, who you're nearby, what
00:03:53
Wi-Fi networks and Bluetooth things are there and what other
00:03:56
devices around you and, and, and what even what time of day,
00:04:00
what's the, what's the lighting, you know, what's the weather?
00:04:03
Are you indoors, outdoors? Why is that?
00:04:06
The more context you have in the in the in the moment combined
00:04:11
with what it knows about you and what's around you in the moment,
00:04:15
you don't have to ask it When you make a query, whether that's
00:04:19
a voice query or a text query or whatever the query might be, it
00:04:24
knows. It's like, oh, you are at this
00:04:26
place at this time around these people with these resources,
00:04:29
with this kind of weather. But, and you, when you ask a
00:04:32
short question, it's like thing as opposed to back in the Siri
00:04:37
days and the Alexa days, you had to tell it everything because it
00:04:41
had no context. So in this case, the better the
00:04:45
interface becomes, the more historical context as well as
00:04:51
real time context. It has to be able to.
00:04:54
So it's like, we know what we're doing right now.
00:04:56
I can, I can briefly say 2 words to you and you probably
00:04:59
understand like, oh, look at that plan over there.
00:05:01
You know what I mean? Something like that.
00:05:02
That's what it needs to know to be able to give you much quicker
00:05:05
answers. Like it's in the room.
00:05:07
But do you buy Like do you think most humans will ultimately want
00:05:11
some AI D have access to their current audio?
00:05:14
Like, do you think that is something that needs to be?
00:05:16
I think, I think there's obviously use cases for that and
00:05:19
also for video. I I don't see and all the other,
00:05:22
all the other sensor suites for sensor fusion out there to make
00:05:25
it work. So in certain contexts and
00:05:27
certain applications, you'll want to do that.
00:05:29
Certain people be no way, you know, people put the tape over
00:05:33
their laptops, cameras and everything else.
00:05:35
So it's just it's going to be an individual preference on what it
00:05:37
is. You know, I remember the glass
00:05:39
hole days of, you know, Google Glass.
00:05:42
We're like, you're not recording me and now we're running with
00:05:44
Meta glasses. So it's all a social context
00:05:48
awareness thing where you know, it it, it's not just about when
00:05:51
the technology's ready, it's about when a particular segment
00:05:54
of society is ready to adopt it. I remember this from my general
00:05:58
magic days when we were making the iPhone 15 years too early.
00:06:01
You know, nobody knew the problems we were solving because
00:06:03
they didn't have those problems, right?
00:06:04
But the idea of like a pin type device is that mostly would you
00:06:09
buy? Is there a pin type device?
00:06:11
Sure, there's a pin type device. There's also a pin type device
00:06:13
now. It's called plod.
00:06:14
But is that trying to solve the social problem of saying I'm
00:06:17
recording? Like why can't the phone just do
00:06:19
the recording? We already have it with us.
00:06:21
OK, so, so first we were talking about the input.
00:06:24
I want to talk about the output, right?
00:06:25
And then there's the form. OK, OK, so let's talk about
00:06:28
input. Now we'll talk about output,
00:06:29
then we'll talk about forms. OK, I know you want to talk
00:06:31
about pins, so let's talk about output.
00:06:34
Yeah. So there's really great times
00:06:37
when you have, and this is what, you know, I really experienced
00:06:40
when I was working with Google Glass and these kinds of things.
00:06:44
Output is really important in very contextual ways.
00:06:49
So if you're on the move, you're going.
00:06:52
I see people out here walking in the, you know, in New York,
00:06:55
walking in the polls and stuff because they're sitting here
00:06:57
looking at their thing or they're on their bike, you know,
00:07:00
doing it or driving, worst case. So, and they're all distracted
00:07:04
from that. And, and a lot of times you just
00:07:05
want audio and you want to have a much better audio interface
00:07:08
versus a, a display. But the best way to represent
00:07:13
visual information is visually, you know, we've heard people
00:07:17
trying to communicate, you know, I'm changing this, this
00:07:20
schematic or whatever, and they say, I'm going to change this
00:07:22
and this. And they're like, huh, And you
00:07:24
just send them a drawing and they go, Oh, now I get it,
00:07:26
right. So you have to understand the
00:07:28
context you're in and what's what's the best way to represent
00:07:32
present that output. Is it visually, is it oddity?
00:07:36
It could be like vibrations, it could be many different ways.
00:07:39
And so when you start to break down the applications, you start
00:07:44
to end the and the environment you're in, like are you in your
00:07:48
office, are you in your home, are you walking, are you in a
00:07:50
bus, are you in a train, are you in a plane, whatever.
00:07:52
And then you start to understand the applications and the ways of
00:07:58
representing that output that makes sense in that given
00:08:01
situation. For output I understand like
00:08:05
Airpods or earphones as an output device for AII don't.
00:08:10
I guess part of why I was limited in The thing is, like
00:08:12
the pin and the pen format, I don't see what the output is
00:08:17
like. To me they mostly exist as input
00:08:19
as. You hit a nail on that.
00:08:20
You absolutely. And that's my that was my next
00:08:23
point, which is the form. OK.
00:08:24
And so what's going on with a pen with a pin with whatever in
00:08:30
the case of let's say it's a pen at open AI, because that's I've
00:08:33
heard the same thing. And I used to know some people
00:08:35
working on things like that there and they used to work.
00:08:38
For me, do you think it's a pen? So, and I also worked with
00:08:42
Johnny a long time for a decade, right?
00:08:44
So I think I understand and I probably think it's something
00:08:47
like a pen, but it can also be glasses and other things.
00:08:49
But given Johnny's predisposition for how he has to
00:08:52
do something very different. So he's got to do that because
00:08:55
he'll have some story around. It's got to be a pen.
00:08:58
And maybe it'll be like the Noto pen.
00:09:00
Do you remember the Noto pen? The Noto pen actually had a pad
00:09:03
of paper and you could just, it was like a digital pen.
00:09:06
You just write on paper and the Noto would keep all of the the
00:09:10
things and it would then create APDF of that thing that you
00:09:15
could then, you know, do text to speech and.
00:09:18
Stuff like that, how AI is bringing us back to our
00:09:20
humanity. Of course there's going to be
00:09:22
this whole big flowery, Oh my God, we're getting away from the
00:09:25
phone and blah, blah blah, blah and and all that diarrhea.
00:09:28
So what really happens though, is that the reason why it has
00:09:33
it's it's a pen for a story, OK, but it doesn't have output.
00:09:38
Because at the end of the day, the best form of output is the
00:09:42
thing you already have with you, which is your smartphone.
00:09:45
OK, It does the best visual output.
00:09:48
You're not going to carry yet another screen.
00:09:49
You might have a couple of LEDs or a very, very tiny thing on
00:09:53
whatever this other thing is for status or whatever, but it's not
00:09:57
going to replace the phone. It can't replace.
00:10:00
That's the the bullshit thing of humane.
00:10:03
We're going to kill the phone. It was like, that was the
00:10:05
stupidest marketing message I've ever heard because you're not
00:10:08
going to kill it. So you have to be a companion to
00:10:11
it. Just like Airpods, just like
00:10:13
Apple Watches. It's a companion to a phone.
00:10:17
So the reason why it's in this form factor is because they need
00:10:23
all the sensor suites. Why do they need the sensor
00:10:27
suites? Because Apple's not going to
00:10:28
give them access to that. On the phone.
00:10:30
So they just don't want to be disintermediated by Apple, no?
00:10:33
No, they're saying you have to get the device because they need
00:10:36
to get all the context to get all the real time input so they
00:10:40
can feed it back to the servers to give you the because all the
00:10:43
security of Bluetooth security, video security, the dots, all
00:10:47
the GPS coordinates when they when you have to ask Apple for
00:10:53
permission and and the user permission for LS things, you
00:10:56
go, Oh my God, they're hyper tracking us as opposed to it's
00:10:59
one network link, whether it's a Bluetooth or Wi-Fi thing, you
00:11:04
know, it's most likely Bluetooth to your device that then does
00:11:07
goes back to the network. It's one thing.
00:11:09
Yeah, I'm just connecting like my headphones, but it's got GPS,
00:11:13
everything else in it. So it's just sucking up all your
00:11:16
contacts and throwing away and getting rid of all those privacy
00:11:18
concerns. You see.
00:11:20
So it makes it real simple and people go, Oh yeah, it just does
00:11:23
it and you don't understand. It's sucking everything it
00:11:26
possibly can suck to get back to open AI to get around all the
00:11:30
restrictions if. You're an AI true believer.
00:11:32
You want all the you too sort of want it to have all the contacts
00:11:35
you have. If you're, if we start to have a
00:11:37
relationship with our own chat bot where it's like, I really
00:11:41
depend on you, I want you to know what I'm doing.
00:11:43
You don't feel this impulse that it's sort of the deficit is in
00:11:47
its lack of context on what I'm doing and that it would be you
00:11:50
don't feel. That you're an early adopter,
00:11:51
right? But do you most of.
00:11:52
The world, most of the world isn't most of the world is like,
00:11:55
what the fuck are you doing with my data?
00:11:57
Right? So am I I'm, I am a I'm a late
00:12:03
early adopter. I have learned how how these
00:12:07
products are made. I've watched this stuff.
00:12:10
I did not adopt IQ Maine. I knew how bad it was because
00:12:13
when I had the first meeting when they were four people, I
00:12:15
was like, this thing's going nowhere.
00:12:17
The same thing happened with various other devices.
00:12:19
I'm like, I don't have time for this crap.
00:12:21
Same thing happened with the rabbit.
00:12:22
I was like, what is this thing? I met with Jesse.
00:12:25
I'm like, Nah. So the, the, the issue is, is
00:12:28
you, you want to be a late enough early adopter, not so
00:12:32
crazy that you're just wasting your time.
00:12:34
So I'm on that side. So, and then I also look at all
00:12:37
the data stuff because I'm also pretty wise and I know how
00:12:40
these, this, this, this, this data stuff is used.
00:12:44
And I'm, I'm in between, you know, you, you just said you
00:12:49
have a, a, a a your first child, a daughter, right?
00:12:51
This is your first. Child, yes, 3 1/2 months.
00:12:54
Congratulations. Thank.
00:12:56
You, I remember a market way difference in the way I thought
00:12:59
about the world before I had kids and after I had kids.
00:13:03
And most of these founders, that was including Larry and Sergey
00:13:07
and and Mark Zuckerberg, all those guys think about the world
00:13:11
before they had kids, before they were married in a very
00:13:13
different way. I want everything all the time.
00:13:16
I'll give away all my privacy. I don't, I don't care what
00:13:18
happens with all this data. As soon as you have kids and you
00:13:22
start hearing about defects and you start hearing about social
00:13:25
engineering and you start hearing about all this other
00:13:26
stuff, you start changing your view on how much data you want
00:13:31
sucked up, OK, and what how you're being protected.
00:13:34
And there's a very, and I know for a fact, because I worked
00:13:38
with these guys, some of these founders that they think about
00:13:41
the world differently and they wish they would have made us
00:13:43
some different moves and different decisions early on and
00:13:49
would like to go back, but they can't.
00:13:50
So I'm just saying going back on your data thing, it's very
00:13:53
different, depends on your. Journey for my daughter we have
00:13:57
tiny Beans, which is like a private app just for our family.
00:14:00
So I can see just you start using products differently in
00:14:03
terms, okay, so if the pen is really an end run around the
00:14:08
control of the iPhone, why would Apple need a PIN like they they
00:14:13
have the iPhone? How does that story?
00:14:15
Because you might be on the device using it, it might be
00:14:18
also a lot of times you don't want to have the device there in
00:14:22
the middle because you're saying like it's, it can also be
00:14:24
ambient all the time because your thing's going to be in your
00:14:27
pocket. So there is, there are some like
00:14:29
the form of having this other thing that if you choose to want
00:14:32
it to watch it, because we've watched these life recorders
00:14:35
before, you know, 10 years ago, there's these life recorders.
00:14:38
This is just another form of. That that's how it gets
00:14:40
positioned. It's like, don't forget any
00:14:42
moment. Yeah, you know, Plod's doing
00:14:44
that now with their their voice 1 and this one I.
00:14:47
Actually don't know that one. Plod's out of the UK and it's,
00:14:49
it's, you know, granola. Yeah, yeah, of course.
00:14:51
It's basically granola, but as a.
00:14:53
Wearable. Are you an investor or in plod?
00:14:56
No, I'm not OK. You just brought it up, so I
00:14:57
didn't. No, no, no, no, no, no, no.
00:14:59
I'm not investor. Do you buy that Apple's working
00:15:02
on a pin? I'm sure I hope to God they're
00:15:05
working on something right, You know, right, they need to be
00:15:08
working on a little bit, but knowing their chip, their chip
00:15:12
expertise, Johnny Saruji and the team, they're incredibly awesome
00:15:15
knowing the mechanicals, knowing that they've made the watch
00:15:19
right and they're making you know, the air, the tracker, all
00:15:24
the other stuff they have the tracker, sorry, you know, the,
00:15:26
the, the Apple, the air tag. They have air tag.
00:15:29
They have all this really, really miniature custom silicon
00:15:32
that and all the Airpod custom silicon.
00:15:37
They can make one of these things really easy compared to
00:15:40
anybody else because they have all the expertise to make an
00:15:43
incredible piece of hardware. They have to get the software
00:15:45
right, but the hardware, the connectivity, they have their
00:15:48
own network stack. They could make a an awesome one
00:15:51
because they have all the camera technology too.
00:15:53
So I it, it would either it's going to be something you sit on
00:15:56
the desk or you wear and you sit on the desk or whatever.
00:15:58
It's like, duh, it's like, or you build it into your, you
00:16:01
know, into your Airpods Pro. Like imagine cameras that are
00:16:05
coming out of your you're using. Could or it's like where, how,
00:16:09
how much do you expect Apple to deliver?
00:16:11
I guess you know, you're talking to somebody.
00:16:12
I'm deep in the Apple ecosystem, my laptop Airpods watch phone so
00:16:17
but then they I think I bought the phone that was sold on AI
00:16:20
and it's one of the most egregious misleads perhaps and
00:16:24
ever marketing a product. When I saw AI first laptop, AI
00:16:27
first phone, I was ready to tweet something that was like
00:16:31
really, really nasty. And I pulled back because I'm
00:16:33
like, you know, it's, it's, it's Apple has never done marketing
00:16:38
bullshit before. And I saw that was complete
00:16:41
bullshit, you know, and I was like, I was holding back.
00:16:44
So, you know, hopefully they're finding new religion and going
00:16:49
back to the old real smart days, which is under promise and over
00:16:52
deliver. What do you think they can do
00:16:53
well right now, and what do you think is holding them back?
00:16:56
What can they do? Well, right now, chips, low
00:16:59
level software devices. It's really, really, you know,
00:17:06
Apple, Apple Vision Pro is an abject failure, but it's an
00:17:09
incredible technical marvel from a technology point of view.
00:17:13
So they can make great stuff. It's do they have the product
00:17:17
chops like they used to, to be able to create the full stack,
00:17:21
create the incredible user experience and deliver all those
00:17:25
pieces of the puzzle in the middle necessary to pull it all
00:17:28
together. And that's the thing that, you
00:17:30
know, has been really important for the reason why Apple is
00:17:35
where it is today. To what extent do you think just
00:17:37
the underlying capability of AI is holding Apple back, right?
00:17:41
Like for a great user experience you want sort of the 100% case,
00:17:45
right? A lot of people talk about some
00:17:47
of the, you know, Chachi BT capabilities.
00:17:49
It's almost like the self driving 90%, like, you know, it
00:17:52
can do most of what you want, but then the more you drill into
00:17:56
a problem, the more you start hitting problems.
00:17:58
And so if you're trying to set up these consumer loops where
00:18:01
it's like every time I step into my house, you do this and the AI
00:18:05
gets it right 90% of the time, Can Apple be expected to deliver
00:18:09
that consumer experience? That they want.
00:18:11
Well, the very first thing to understand is LLMS have
00:18:14
fundamental problems. They have fundamental problems
00:18:17
and that's and in in the way they're created and that's why
00:18:20
they hallucinate. And unless they're de scoped and
00:18:24
they're really trained on very, very contextually specific
00:18:28
information like medical or legal or, you know, whatever
00:18:31
general consumer stuff instead of these incredibly huge models
00:18:36
that can hallucinate because they don't know whether you're
00:18:38
talking medical, legal, whatever it's it tries to be a know it
00:18:40
all. You'd never hire a know it all
00:18:42
in the real. World, we can't stand those,
00:18:44
right? I admit what I don't know right?
00:18:47
Chacha PT. Never it always.
00:18:48
Gives you an answer and it doesn't know whether it's right
00:18:51
or wrong. And so, you know, we're invested
00:18:53
in various companies in the medical domain like nabla and
00:18:56
other things that have very, very specific context and very,
00:19:01
very specific guard rails for what they will do and what they
00:19:05
will answer and they will not answer stuff.
00:19:07
It's when you're trying to make these general purpose ones,
00:19:09
that's when it's a problem. So let's so getting back to the
00:19:14
the LLM specifically, if Apple has very, very specific models
00:19:19
for consumers in certain contexts, it can work really
00:19:22
well. It's when you apply these very,
00:19:24
very broad models and you don't know what's going to come out,
00:19:28
then that is where the problem is.
00:19:30
There's still models and other things beyond LLMS, even highly
00:19:34
contextually constrained LLM trained LLMS that are coming
00:19:38
that are going to get a get around a lot of those things.
00:19:41
But right now everybody's blah, blah, blah, LLMS, AGI, bigger
00:19:45
and bigger models. What do you?
00:19:46
Think is coming? Shit, what's coming?
00:19:49
Well, there's world models coming, right?
00:19:51
So world models combined with LLMS.
00:19:53
So that's understanding physics, understanding chemistry, all
00:19:56
this stuff that we know because there are real rules,
00:19:59
mathematics around those things. You know, today, Chet GPT can't
00:20:03
even draw a face of a clock, right?
00:20:06
Or understand what a clock says because it doesn't graphically
00:20:08
think it's easy for us for a graphical clock and not a
00:20:11
digital 1. And so it's like though, that a
00:20:14
world model would go, oh, that's a graphic.
00:20:16
I'm going to use this model. It you know this is how it
00:20:19
works, you would build that stuff.
00:20:21
Into those models, tools where it says for this question you,
00:20:24
you go off to this software or this technique.
00:20:27
To solve some, yeah, absolutely. And then the language model is
00:20:30
good for what the language model is good for, as opposed to
00:20:33
trying to think it reasons about chemistry because it doesn't
00:20:37
have a real basic chemistry model.
00:20:39
When we started this conversation, I think you were
00:20:41
saying, you know, audio and voice is sort of at the core of
00:20:44
what AI wants to consume. Where or maybe is that fair?
00:20:48
Or I mean, clearly listening to everything we're doing is part
00:20:52
of what these devices we've talked about want.
00:20:54
But where, yeah, where do you think about sort of the Airpods
00:20:57
and earbuds in terms of the AI device story?
00:21:02
Well, you know, I, I, I, there's, there's edge devices
00:21:06
which are Airpods and, and, and you know, and phones and these
00:21:11
kinds of things. I think of the edge of the edge
00:21:13
first. OK, so the edge of the edge is
00:21:16
the very, very fringe of like, are you speaking a word?
00:21:20
How do I do noise cancellation? How do I do speaker focusing
00:21:24
those kinds of things? There is AI that can do that
00:21:27
stuff really well. Your ear does that.
00:21:29
You do that. You turn your head, you focus
00:21:31
in. So we can apply IAI into each of
00:21:34
these at the edge of the edge devices to do much better
00:21:38
intelligence to help us be able to filter out noise from.
00:21:42
To do the devices core task. The the very basic task of
00:21:48
separating signal from noise, whether that's visually or or or
00:21:53
auditorily or other things of that nature.
00:21:55
So we need to get there. So that's the first level of AI
00:21:59
applied. Then there's the AI above that,
00:22:02
which is labeling, interpretation, those kinds of
00:22:04
things. And that can be done on device
00:22:06
in some cases or it'd be done with the companion.
00:22:09
It's which would most likely be your smartphone.
00:22:11
So that's another set of AI. Then there's another one on top
00:22:15
of that, which is like, OK, what's the reasoning?
00:22:17
Let's make, let's get the context and then figure out
00:22:20
what's going on. So there's all these levels and
00:22:24
we need to break down the AI and levels.
00:22:25
Just like your brain has many different, different, you know,
00:22:31
centers in your brain to do this, this, this, this, this,
00:22:34
you're going to start seeing AI becoming much more diverse and
00:22:38
not just one big thing. And we're seeing that with self
00:22:40
driving cars already. I know we're all over the map.
00:22:44
No, I know. I love it.
00:22:45
To some degree. The AI wave creates this broken
00:22:50
thinking because you have companies that have that want to
00:22:53
justify valuations, so they need to sell something and they have
00:22:56
the story, which is AI. And so in some ways, I'm
00:22:59
complicit in what they're doing, which is like think from the
00:23:02
need for an AI device out rather than like some human problem and
00:23:07
then sort of build the device correct up.
00:23:09
And that's why this conversation ends up being chaotic, because
00:23:12
we're shifting between. From the technology to the use
00:23:15
technology, it's like we're. In search of a But don't you
00:23:18
think there is this moment to sell the consumer on a new
00:23:21
device? They've heard about this
00:23:22
technology. They're they're sort of open to
00:23:24
adopt a new sort of way of behaving.
00:23:27
And so people are searching for what, what, what device can we
00:23:30
could deliver them in this scene is.
00:23:32
To me, these devices, a pin, a pen or whatever, are enablers,
00:23:35
right? They're not a solution and where
00:23:38
we were going was I was saying that we need more context
00:23:41
specific LLMS drug so that consumers can get a a reliable
00:23:47
experience without hallucinations.
00:23:49
What that means is you have to come up with the applications
00:23:52
and constrain the application to better understand what it is you
00:23:55
want. It's not just a know it all
00:23:57
that's going to go and try everything.
00:23:59
So when we when you switch into certain applications, you want
00:24:02
that because why? At the end of the day, we're not
00:24:05
where we want to be. It's not, we're not at her yet.
00:24:07
Everybody wants her the thing, but we don't have the technology
00:24:13
yet. So we have to creep into it,
00:24:15
right, which means application, more application specific
00:24:18
context specific LLMS, forgiven workflows and not just it's a
00:24:24
general anything device, OK. And so, so we're going to see
00:24:29
much it's like kind of web .1 point O to two point O to
00:24:33
three-point O. We've had to make these leaps
00:24:35
because we have to have the technology to move along.
00:24:37
With it. But we, we had Siri many years
00:24:40
ago. We've had Alexa.
00:24:41
Like, do you think we're still far away from the sort of voice
00:24:45
communicated personal assistant that people talk to to engage
00:24:49
with a computer? Yes, because why?
00:24:52
What I see is you're going to have your federation of your
00:24:55
devices. That's going to be everything
00:24:57
for your laptop, your TV, your game machine, all the things you
00:25:00
carry on, you watches, whatever it is, you're going to have
00:25:03
that. Then you're going to have a
00:25:05
smart agent in the background really understanding which
00:25:08
device you're on, which users and all the data and all the
00:25:11
context and then be able to switch how it's working based on
00:25:15
those things. Today what we're saying is put
00:25:17
everything on a device and it's going to work or put all the
00:25:20
sensors here and put all of the the the smarts in the cloud and
00:25:25
it's and everything else is just dumb.
00:25:27
In between. We're going to build into this
00:25:30
much more multiple AI world where edge of the edge, the edge
00:25:35
devices, the federation of edge devices and the cloud all work
00:25:38
together to produce the her application.
00:25:42
We're hearing that we would love to have, but we're just not
00:25:45
going to get there today from what everyone's telling you, you
00:25:48
know? Just because of the capability
00:25:50
of LLM. Yeah, because we're just not.
00:25:51
We don't have the world models, we don't have the sensors.
00:25:54
When you look at these humanoid robots, everyone's like, oh,
00:25:57
they do flips and tricks and everything.
00:26:00
Great, we can do that. But we don't have the necessary
00:26:03
sensing and actuation necessary to be able to get the data to be
00:26:07
able to write. And they need world models to to
00:26:10
do what anything like a humanoid stuff that people are hoping
00:26:15
it'll do. Remember self driving cars came
00:26:18
out in the late 22. We're just there now.
00:26:22
Right. No, it's great.
00:26:23
Right. And we've spent billions and
00:26:25
billions and we're now getting, and it's just getting rolled out
00:26:28
and there's still 300 $1000 cars with massive sensors and
00:26:32
compute and those are massive edge devices plus a whole set of
00:26:37
coordination in the cloud. So that's what we need to do.
00:26:41
And we need to do it at a very cheap level and battery
00:26:44
efficient level and right. And it's gonna take time to get
00:26:48
there, right? What is the state of the smart
00:26:52
home today or what? OK, take, I know you were
00:26:56
talking the big device questions here like and it fits into this
00:26:59
AI conversation, but like what is your read on just the success
00:27:04
or failure of building, you know, smart home devices?
00:27:07
OK, I'm going to build a remodel of a big new smart home thing
00:27:12
for myself. So I'm, you know, I'm tracking
00:27:14
the very latest and now I'm not in it like I used to be so.
00:27:18
Does that include custom devices or you're just sourcing the best
00:27:21
stuff that exists? Well, I'd say that there there's
00:27:24
the best of what exists is from the device side, but then
00:27:27
there's custom ways of putting it all together.
00:27:29
It's like what we just talked about the self driving car thing
00:27:33
you have and when I'm designing this thing, I have to think
00:27:35
about the devices, the low level network, the application level
00:27:39
network, what's in the server, what's in the cloud like and
00:27:42
it's still very, very difficult if you want to pull off what
00:27:45
again the smart home that the Jetson smart home that we all
00:27:48
think about. Is that what you're trying to
00:27:50
get? Yeah, and I just know better.
00:27:52
Like I'm like OK I'm going to wire everything and make sure I
00:27:54
have thread border routers in the necessary places.
00:27:56
And what are the main things you want the house to be able to do?
00:27:59
Oh, right now, just the simplest things like I have a you know,
00:28:03
I'm the the in in in this case, I have a self playing piano,
00:28:09
right. So I'm I'm I'm pumping MIDI into
00:28:12
it. So I pump MIDI and I put songs
00:28:14
from the net or from locally MIDI into it to them.
00:28:18
It self plays and then it has it's miked and the mics are then
00:28:24
A to deed into a Dante audio network, which then distributes
00:28:28
all the way through the entire house or whatever zones I want
00:28:32
through Dante into all of these Dante amplified speakers around
00:28:38
the house and the Sonos can pump in and so I can have a player
00:28:41
piano. And the keys are pressing.
00:28:44
The keys are pressing, so it's a analog piano getting digitally
00:28:47
recorded and streamed all through the house.
00:28:51
So it's. Pressing the key, is it like an?
00:28:52
It's not an old school. No, no, no, it's a modern, it's
00:28:56
a modern regular grand grand piano, but it has actuators on
00:29:00
it and and it plays, it plays the songs just as it 'cause they
00:29:06
record it from a professional. So it's like you have a concert
00:29:09
pianist with you and then, and then you're and you're miking it
00:29:14
and then you're amplifying it and playing it throughout the
00:29:16
house. So you, it feels like you have a
00:29:17
concert pianist in the house. Have you?
00:29:19
Executed on that yet? Or is this still a dream?
00:29:21
Oh no, we're it's in, it's in the mix and it it can work.
00:29:24
It absolutely can work. But then there's all the
00:29:26
switching to make the Sonos move from one thing to another.
00:29:31
And how do you and which zones did the this the player piano go
00:29:34
to to the? And are you like a speaker head?
00:29:38
Like, do you have this sort of fancy old school speakers or
00:29:41
you're using like Sonos? Well, no, I'm using Sonos ports
00:29:45
to to get the house to amplified speakers for certain either in
00:29:50
ceiling speakers without you don't see the grills.
00:29:53
You've seen some cases. I don't like them but there's
00:29:55
some cases in certain wet locations you need the grills
00:29:58
and other things where it's L acoustics like Dolby.
00:30:01
Atmos Right, OK, you. Know whole thing and you have to
00:30:05
worry about all those levels of networks to make all that stuff
00:30:08
work. You know, got AVC, you got AEC,
00:30:11
There's so many different things like I have, plus I also have
00:30:15
full lighting. So there's DMX lighting control
00:30:18
that you do on stages. There's also the MIDI over
00:30:23
Ethernet and everything runs over Ethernet.
00:30:25
So it's MIDI over Ethernet, lighting control over Ethernet,
00:30:30
Dante over Ethernet, lighting control over Ethernet, right?
00:30:34
So it's a full on, you know, next generation system.
00:30:40
Anything else, you're a smart home.
00:30:42
Well, it's going to have thread of course, because we did we
00:30:44
what is? Thread, sorry.
00:30:45
So thread, we, we had basically invented it and now Thread is
00:30:49
the Bluetooth equipment equivalent for mesh networking
00:30:54
for entire house. So it mesh networks an entire
00:30:56
house for low speed data and control very, very low latency.
00:31:01
So you can do light switches, light bulbs, sensors, everything
00:31:04
else that can be battery operated for 10 years, that kind
00:31:08
of thing. And it's very resilient.
00:31:10
You do anything like? Mechanical like I don't know,
00:31:12
are there meals being made in automated?
00:31:15
There's mechanical lighting and TV's that pop out things, the
00:31:19
lights drop down, robo lights that move.
00:31:24
Yeah, that's the most of the mechanical except mechanical
00:31:29
locks and you know and you know and NFC based door door locks
00:31:37
that kind. Of stuff And what do you use for
00:31:38
the actual like voice communication with your house?
00:31:41
That is to be determined. It's a Bake Off or.
00:31:44
What do you mean nothing's ready yet, right, right.
00:31:47
So you have to next generation Alexa is not there yet right.
00:31:50
So it's it's set up that's the next layer on top of all of this
00:31:53
infrastructure. The next layer is on top of that
00:31:55
and that's TBD. What are?
00:31:57
You typing it? What's what's what's gonna
00:31:58
happen? You have to make a decision,
00:32:00
right? No, I don't.
00:32:01
Why? Well, you're.
00:32:03
Gonna have to talk to it. What are you gonna use?
00:32:06
When is this house supposed to be ready to go?
00:32:09
What are you gonna use? You use buttons just like cars
00:32:12
are taking coming from screens, and you're not gonna use
00:32:14
buttons. Right now you're not ready to
00:32:16
use voice. No it's cuz it's fucking
00:32:18
bullshit. So you just have like modules
00:32:20
like. I don't feel like having
00:32:21
everyone scream at me and you know, and.
00:32:23
Turn on the lights. You don't have any of that.
00:32:26
You know, look, this is the other thing I learned from smart
00:32:28
home, right? OK.
00:32:30
Is that most people who design the smart home design it for
00:32:35
themselves. It's like for single guys I want
00:32:38
1 button that does this and one button and then when you have.
00:32:41
People over there. When you have kids or you have
00:32:43
friends over or you have, or you have a wife or partner with you
00:32:48
and they're like, I don't want that button.
00:32:50
I want this button. Everyone has their own 1 button
00:32:53
interface. So you have to design and
00:32:56
understand your, your just like anything.
00:33:00
It's not just you. You have to design this whole
00:33:02
world when it's a, a shared space versus your own AI, you
00:33:06
know her space. It's a very different thing.
00:33:09
What I mean, one of the things that's amazing about the devices
00:33:12
you guys produced at Apple, obviously they were expensive,
00:33:15
but they're fundamentally mass market.
00:33:17
It feels like the billionaire phone is the phone that sort of
00:33:21
the middle class American is able to afford.
00:33:24
I mean, the portrait you're painting is not that where it's
00:33:27
like with the home and. And that's sort of like old
00:33:29
school. Oh.
00:33:30
My stuff's really crazy, but you can make a bunch, right?
00:33:33
But easier. If you think about like sort of
00:33:35
getting the home to that that piece because that on some level
00:33:39
it's what you want too, which is a device that's made that sort
00:33:42
of works that solves some of these problems without the sort
00:33:45
of. I want less screens in my home.
00:33:47
I don't want more screens. I hate screens, right?
00:33:49
Frankly, I hate screens. I don't want them in the home.
00:33:51
So. Why hasn't the home had that
00:33:53
sort of like mass market amazing electronic work yet?
00:33:57
Oh, because we, we're still in the days of like, you know,
00:34:02
remember Wi-Fi didn't even exist until 2001 too, right.
00:34:06
So we thread just finally got, even though it started in 2011,
00:34:10
it's just now there. We're still building all the
00:34:13
base layers and linking everything and then and we and
00:34:17
we're now going to get the LLMS and voice recognition, all that
00:34:20
stuff to then put it all together.
00:34:23
But it's still on the verge just like the self driving car was.
00:34:27
When that smart home like that is, you got to give it another I
00:34:32
think 5 to 6 years for what we think of when you think smart
00:34:37
home you want it to be, I think it's going to be there.
00:34:39
Just set up your home to make sure it.
00:34:41
What? What company do you think is
00:34:43
best positioned? No idea it's 5.
00:34:45
Six years, No, because the thing that I learned at Nest was all
00:34:48
these sensors and smart plugs and all that other stuff.
00:34:52
There's no money in it. There's absolutely no money in
00:34:55
that stuff. That's why you go online and
00:34:57
that's why there's. People are only willing to buy
00:35:01
20% more expensive. Yeah, cuz I have a switch on my
00:35:04
wall. Why do I need anything else?
00:35:05
And any crazy guy like me doing this other thing, it's just.
00:35:09
Not a big enough. Market, it's not a big enough
00:35:10
market, so nobody wants to play in that space.
00:35:14
So IKEA is finally, you know, to tell you the truth, because
00:35:17
they're using thread and they're using matter and that kind of
00:35:19
stuff. So IKEA is starting interesting,
00:35:21
but it's just starting. It's all nascent.
00:35:24
So I think that the ones that are going to be successful are
00:35:28
the ones who are going to finally deliver the Nest vision
00:35:31
that we really had. But it's still, you know, if I
00:35:35
was I would be rebuilding Nest right now, you know, and and and
00:35:40
because in three to four years you can be there to get what
00:35:43
we've been talking about that we had the vision for in 2010
00:35:47
eleven. You're obviously investing right
00:35:49
now, but do you, do you still have the entrepreneur bug?
00:35:52
Like do you think you have another great company in you?
00:35:54
Oh, I have over 100 great companies in me right now.
00:35:57
I, I, we, you know, I've invested directly in many, many
00:36:00
things and I invest indirectly in many things.
00:36:02
So I have over 100 companies now, 100 and something companies
00:36:06
where I'm working with them every day.
00:36:09
Like today I was working with a company and helping them with
00:36:11
Apple. I'm a shadow CEO at a couple of
00:36:13
companies. I'm doing product design at
00:36:15
companies, so I get to do this in many different domains.
00:36:19
To be a shadow CEO. So there's this chat, there's
00:36:22
the CEO and then he's like, please help me.
00:36:23
So I get to be behind. I'm like, would you think about
00:36:26
it this way? Operations head count, how
00:36:28
should we should do funding, product design?
00:36:31
And so I have various roles like that where they please help.
00:36:35
And so I do, and I go in and I get to a.
00:36:37
Couple of companies are spending the most time on right now.
00:36:42
I'm spending a lot of time with Orionis Biosciences.
00:36:45
They are doing AI drug discovery and we have drugs in FDA phase
00:36:51
two clinical where they're actually working and helping
00:36:55
with the with that. So I'm spending a lot of time,
00:36:58
I'm not a shadow CEO there, but I'm, I'm on the board and
00:37:01
tapping out. I'm, I'm at Menlo Micro.
00:37:05
Menlo micro is a, a MEMS switch. So we have transistors and we
00:37:10
have relays MEMS, which sits in the middle.
00:37:13
And it's an incredible new technology that we've been
00:37:15
working on for 10 years. They've been working on it for
00:37:18
almost 40 years to power AI data centers and robots and all these
00:37:23
other things. It's absolutely phenomenal thing
00:37:27
and helping with the business development, the product plans,
00:37:30
the the strategy, the financing, that kind of stuff.
00:37:34
Let's see, what else am I working with?
00:37:37
I'm doing. I was just doing some VR work
00:37:40
this morning. Oh, really?
00:37:41
Yeah. Are you a believer in VRI?
00:37:43
Believe in I believe I fucked the metaverse.
00:37:47
I believe in VR for episodic B to B and some gaming, but
00:37:55
episodic. So you're not living in it.
00:37:57
Like, you know, if some people are saying you're going to live
00:37:59
in it, it's like, no. But where it's collaborative,
00:38:02
it's high value and mostly B to B Sure, there's gaming and
00:38:06
that's a fringe, you know, it's big, but it's still a fringe in
00:38:09
from my perspective. But where you're doing design
00:38:12
online. So we're in a company called
00:38:13
Gravity Sketch, OK, And Gravity Sketch is the number one.
00:38:17
Think of it as Figma in the in the VR world where you're doing
00:38:21
collaborative design in a 3D space and you can do markups,
00:38:26
you can create in it. And we're at all the biggest
00:38:28
companies where it's the Lannis, Nike, the Ford.
00:38:31
Everybody uses us to do car design, shoe design, furniture
00:38:35
design. So, so I believe in it for that.
00:38:37
So they have teams of people wearing headsets, doing this for
00:38:40
an hour or something, getting in it, coming out, you know, or
00:38:43
reviewing cars or whatever. And I believe in VR for that for
00:38:46
sure. But for most of the stuff that
00:38:48
was pitched by, you know, Meta and Apple, I don't, I don't
00:38:52
believe in it at all because I've been working on VR since
00:38:54
1988 and I've seen the issues with it.
00:38:58
This is circling back to some of the the big themes, excuse me,
00:39:01
some of the big themes of this conversation, like what company
00:39:05
do you think is best positioned to produce like a new iPad scale
00:39:11
device? Like someone like what if you
00:39:13
had to make a prediction about a company producing an advice that
00:39:17
reaches our culture at that scale?
00:39:19
Who are you betting on? Hopefully Apple makes the iPad
00:39:24
phone or the iPhone pad which is basically a foldable iPhone.
00:39:29
Okay, okay, yeah, yeah, bring the benefits of the iPad to the.
00:39:32
IPad just bring the foldable stuff and it's been bandied
00:39:35
about and stuff like that, but the operating system is that OS
00:39:40
new. Device or.
00:39:42
I, I think it's a meaningful segment for Apple to continue to
00:39:45
grow and, and to make sure that they don't lose market share to
00:39:48
people with foldables. But I do think there's a lot of
00:39:51
use cases where if you can carry with you everywhere and you can
00:39:54
open it, especially in some cases with AI, that it could be
00:39:59
pretty powerful, right, right. Because an iPad, you just can't
00:40:02
slip in your pocket. You want a pocketable.
00:40:05
You know, kind of iPad experience where you can use
00:40:08
voice so you don't need a keyboard, all that stuff to
00:40:11
interact with. It could be pretty pretty damn
00:40:13
cool. What is that what you would
00:40:16
focus on if you were Apple CEO? I have many things that I would
00:40:22
do if I was well. All right, give us some of them.
00:40:25
Give you some of them. Yeah.
00:40:29
Well, the first thing I think they really flubbed on was
00:40:31
mobility. You know, there was the Apple
00:40:33
Car and we all know that that was a real thing and they killed
00:40:35
it. But there's other ways of.
00:40:36
Doing that, you think they should have kept going?
00:40:38
No, no, no. No, Apple shouldn't.
00:40:40
Apple is a company that redefines certain it's certain
00:40:46
aspects of life. It redefined what it was to be
00:40:50
desktop publishing, right? Or publishing in general.
00:40:52
It redefined music. It redefined various things.
00:40:55
So they should redefine mobility.
00:40:58
Don't make a four wheel car that competes with everybody else and
00:41:01
the Chinese. How would you change mobility in
00:41:05
two in three and lightweight 4 wheel vehicles?
00:41:08
How do you do that that the kids are like at 14 can use and go.
00:41:13
I want to keep using it right. Don't not a a lucid blah, blah,
00:41:18
blah for just you know, you got to think, you know, when when
00:41:22
Steve Jobs and I would walk around the Apple campus back in
00:41:25
2000, 1008 nine, we talked about the Apple car.
00:41:28
What would it be? And he was really like we need
00:41:31
to do. He thought it was revolutionary
00:41:34
was the Volkswagen, right? The the people's wagon, the
00:41:37
people's car. And he was like, what's the
00:41:39
people's car? What's the next generation
00:41:41
people's car? What's going to be used in the
00:41:43
cities? And now you're seeing the Fiat
00:41:45
Topolino, you see the the Twizzy, you see all these
00:41:48
different things. I live in Europe, right?
00:41:49
So you see all this stuff running around and they're
00:41:51
selling out like crazy. And you're like that Apple or
00:41:54
two wheels Apple or three wheels Apple?
00:41:58
And so you would say go keep working on that vehicle, work on
00:42:01
this vehicle vision. I think that's only one thing.
00:42:06
I think that's one. I think mobility is something
00:42:08
that should be addressed and Apple could address it in a very
00:42:11
different way than it and has all kinds of brands that has the
00:42:14
technology, all that stuff. This other thing is I think it's
00:42:17
a red herring and it's more or less a me too kind of thing.
00:42:21
Frankly, I think that the their accessories line up could go
00:42:24
much, much, you know, they could lean in much more into that.
00:42:27
Lots of things to be done around that.
00:42:29
Obviously glasses like more, you know.
00:42:30
Rainer, or the aesthetic piece of it, or.
00:42:33
The no, no, just usability, no, no aesthetics, No, no, no, no,
00:42:35
no. We're talking about new markets,
00:42:38
new things that they could. Go after Aura directly.
00:42:41
Sure. You know, like duh.
00:42:44
Are you bullish on Aura right now?
00:42:46
I got one on, you know, to do truth be told, No, no, to be
00:42:50
truth be told, I was invested in the company before Aura.
00:42:54
It was called Motive. We made the first smart ring.
00:42:57
We had all of the patents, all the fundamentals and everything.
00:43:02
And the team who was from Apple just wanted to do something
00:43:05
other than a ring. After they got the ring, I'm
00:43:07
like, guys focus on the business and they're like, no, we want to
00:43:10
make, you know, some other wearable.
00:43:11
I'm like, no, focus on the business.
00:43:13
So they lost it and then what happened?
00:43:15
Aura ended up buying all of our patents.
00:43:18
I. Didn't know that so so to me,
00:43:21
like we were doing the we were doing it four years before Aura
00:43:24
did. And now my team from Nest is the
00:43:26
marketing team at Aura, which I love those guys and they're
00:43:29
doing great. I just wish we would have motive
00:43:31
would have been it instead of. Is Apple supposed to be thinking
00:43:33
about a ring right now or? If they're not, they'd be crazy
00:43:36
they if they weren't, and it's so easy for them to do.
00:43:38
They have all the chip like the house aura doesn't have chip
00:43:41
guys. They don't have all that.
00:43:42
So they have everything they need to make an incredible ring
00:43:46
like duh, and they have all the all the the composites for, you
00:43:49
know, the materials, everything like it is a no brainer.
00:43:52
It's just like the pin no brainers to do.
00:43:54
They have everything they need to.
00:43:56
They just need to have the gumption to get it out and do
00:43:59
it. It's really simple.
00:44:00
What are other accessories? I think there's smarter ways of,
00:44:05
you know, making a Airpod that has an iPod in it.
00:44:11
So I think they need to bring back the iPod.
00:44:13
OK, I like this because people don't want to have their phone
00:44:16
with them all the time. Well.
00:44:17
There's two reasons. One is there's the nostalgic
00:44:19
value of everybody who got into Apple, a lot of them got in
00:44:22
because of the iPod, right, Right.
00:44:24
So people want don't want the they would like the nostalgic.
00:44:28
For my first Apple device. Yeah, most people is because at
00:44:31
the time of the iPod 1% market share in the US only, not around
00:44:36
the world. So most people and there was no
00:44:38
Apple retail, so most people's device because they didn't have
00:44:41
a Mac and they didn't buy a Mac, right, because they had APC.
00:44:45
So they had to get an Apple device and was most likely an
00:44:47
iPod. So iPod from the systalgic point
00:44:50
of view for all those people who's like, oh, I remember the
00:44:52
iPod, but they don't want distractions anymore, right?
00:44:56
They want they don't want distractions.
00:44:58
The third thing? Sorry, sorry Apple, who doesn't
00:45:00
want distraction? Well, a lot of people want that
00:45:03
pure missile. So the music player.
00:45:05
Who was? I just bought a record player so
00:45:07
I'm deep in this, right? Did you see what Sony?
00:45:09
Did this week, No. Sony killed their TV's and sold
00:45:13
it to TCL and they went back to the turntables.
00:45:17
They just released 2 new turntables.
00:45:20
They kill the TV's and they bring back turntables.
00:45:23
You tell me. So is there a nostalgia element
00:45:26
for the iPod? I think so.
00:45:28
I'm. Sorry, is the iPod in this
00:45:29
vision software or it's a hardware device?
00:45:32
You're saying? Where does it?
00:45:34
What format? It's I don't want to give away
00:45:36
all my stuff. But you're saving this.
00:45:39
Has anybody talked to you about it?
00:45:41
You know, there was obviously reporting about, you know, the
00:45:43
next CEO and your name was, you know, bandied about.
00:45:45
Has anyone done any feelers or you're like I got to hold back?
00:45:49
I've got a lot of incoming I, I, I, let me, I want to just say
00:45:52
thank you to those, those ex Apple employees and those
00:45:56
current Apple employees who have reached out and have tried to
00:45:58
canvas me and try to get me involved.
00:46:01
You know, Apple will make the right decision when they need to
00:46:03
about who the next CEO is. I love the company.
00:46:07
It's been in my blood since 198081 when I my first Apple
00:46:11
two. You know, it's it's been the
00:46:13
thing since since I was 11 to 11.
00:46:16
If anyone calls, you know, from the board or Tim calls, right,
00:46:20
I'll be happy to pick up the phone right, and that's it.
00:46:23
I'll, I'll help them any way they think they they, they would
00:46:25
like to be helped, right? The accessory, that's a great,
00:46:31
I'm fascinated by that and then. There's all the things they
00:46:33
should do. There's a whole another product
00:46:35
category there, so, but we'll leave that for.
00:46:37
Later, OK. We'll leave that for later.
00:46:39
What I mean the the found the the the labs, the foundation
00:46:42
model companies. What's your what's your read on
00:46:45
Open AI Anthropic Xai I? Know you want to go there, Yeah.
00:46:49
Yeah, of course, this is a big technology question of the day.
00:46:52
Are they great businesses? I guess that's the first grade.
00:46:54
Are they good businesses? They have to really target
00:46:57
applications to be great business because platforms,
00:46:59
platforms are not businesses. Applications are great business
00:47:03
and that was proven. The last great platform business
00:47:06
was Windows 95, you know, or Windows.
00:47:09
There's been no platform business other than that.
00:47:13
OK, so you could call maybe AWS or you know, Google or whatever,
00:47:18
but that's AB to B thing. It's not a, it's really about
00:47:21
applications, not platforms. So, so when you look, you stack
00:47:26
up, open AI, Google Anthropic, you know, Mistral, you can pick,
00:47:31
you know, whoever is. And then let's talk about the
00:47:32
Chinese and even the Koreans and stuff like that.
00:47:36
Look, there is a point and Dario's doing a great job at
00:47:43
Anthropic turning it into a real business and coding, because
00:47:49
it's so language focused is an incredible application for an
00:47:54
LLM, right? And we're seeing that, right.
00:47:56
It's just crazy with now they're, you know, they're non
00:48:00
coding version of it of Claude as well as the coding version.
00:48:04
I'm using it now I'm playing, I'm like, Oh my God, I'm so I'm
00:48:08
getting back into coding. I haven't done coding in years.
00:48:10
I'm not a coder but I yeah I got the desktop version.
00:48:13
It's a little. I feel like I don't have files
00:48:15
so I'm a little confused as a user that I have stuff on Google
00:48:18
that I want them to use which I think I need to use in their web
00:48:20
browser stuff. I mean, these are I.
00:48:23
Must say, look, I must say that they have a whole set of
00:48:28
application stuff and training because it still feels like it's
00:48:31
a GitHub model for Claude, like it's very giddy.
00:48:35
And that's, and if you look at if you look at that, it's a very
00:48:41
small number of people. They have to come from a
00:48:43
different way to get people involved.
00:48:45
And it's still not approachable. Like I'm like still like
00:48:47
confused when I'm using it or when I'm even installing it and
00:48:50
trying to run it like what? OK, terminal this like how many
00:48:53
people do that? There's other ways of getting
00:48:55
around that and they're gonna, I'm sure they'll get there.
00:48:57
There's just no way they're not. But again, they're doing a great
00:49:01
job of really focusing and figuring out that stuff around,
00:49:04
coding around that's guard rails, all those things.
00:49:08
And they're trying to get to profitability so that they are a
00:49:12
real standing business. So I commend them for all that
00:49:15
they're doing. They're, you know, they were EXA
00:49:17
open AI people, right? And everything And they really
00:49:19
are targeting. So that's really great.
00:49:22
Then we go to Open AI. Open AI right now is spray and
00:49:26
pray. Right.
00:49:27
Spray and pray. Look, this week it was rumored
00:49:32
and I'm sure it was a very strong rumor they're trying to
00:49:34
reach $50 billion, right, 50 billion.
00:49:38
Just a few months ago it was 5 billion and just before that it
00:49:41
was 500 million. I don't think unless you've been
00:49:44
in the market and understand this is massive amounts of
00:49:47
capital that even large fortune 5 hundreds can't get access to
00:49:51
like and So what I the game I believe Sam is playing, and it's
00:49:56
a really dangerous game, is too big to fail.
00:49:59
He's getting every single person to give that money, including
00:50:03
large governments and everything else.
00:50:05
So they're too big to fail, right?
00:50:08
Well, this was the whole Sarah Fryer backstop.
00:50:11
Yeah, yeah, the back, that backstop, which he was like, no,
00:50:13
no, and they're backpedaling it, but.
00:50:14
They they've written stuff that's.
00:50:16
Just you don't think they're not having those discussions at open
00:50:18
AI? Like, yeah, these governments
00:50:20
will backstop us at some point because we're going to crash
00:50:22
everything, the semiconductors that they're, you know,
00:50:25
everything. So so I think it's a incredibly
00:50:29
dangerous game they're playing and with especially with the
00:50:32
circular deals and everything. Is there a there there?
00:50:36
I believe there's a there there, but Google's gonna win it,
00:50:39
right? Google has everything from the
00:50:41
chips to the data centers to the all the way to the applications
00:50:45
and tons of data for all this stuff.
00:50:47
I wish they had a better devices team.
00:50:49
It sucks there. I'll go whatever, but at the end
00:50:54
of the day, they have all the things necessary and plus an ad
00:50:59
business that can fund billions of dollars 1/4 and distribution.
00:51:03
You know, it's just it's everywhere.
00:51:05
It's like it's it's. Funny, I feel like there's so
00:51:07
many AI haters that, you know, they obviously Google things and
00:51:10
so then they'll get the AI results from Google.
00:51:13
And so I feel like there's or, you know, Gemini's fundamentally
00:51:15
delivering the results. So there's this weird world
00:51:18
where people hate AI and still use it all the time via Google,
00:51:22
which is just I. I I hear what you're saying.
00:51:25
You know, there, there's this, it's more like I I hate what
00:51:28
it's going to do to the world. I mean, I hate, but I it works
00:51:31
for me, you know, so I think there's a little dissonance
00:51:34
there. But but you know, I have to
00:51:37
credit Sundar, you know, and the team that they've had an
00:51:40
existential crisis and I think they cleaned up the culture
00:51:44
enough. I'm sure it's not fully clean,
00:51:45
but enough to say we have an existential crisis.
00:51:48
We're going to get to work. We got to go back clean up was.
00:51:52
Too casual? And sort of it was too casual to
00:51:55
you know, we, you remember the first things of Gemini that came
00:51:58
out. We can't blah, blah, blah.
00:52:00
And we saved, you know, in the in the MAGA words woke, right,
00:52:04
too safe and too like, you know, right, right, right.
00:52:08
Well, you could, but you could you're like, come on, right.
00:52:12
So I so do you. Have a read on whether how much
00:52:16
Sergei really gets credit or what the internal dynamics are
00:52:19
there in terms of who's really drawing.
00:52:21
Well, I hear a little bit. I don't really want to comment.
00:52:23
It's, it's, it's, it's look at the results.
00:52:26
The results are sure they fumble a few times, but they're,
00:52:29
they're locked in and engaged and they're, and so I, I have to
00:52:33
give them every credit for turning the ship.
00:52:35
I you know, it was a big, it's a huge ship at Google, but it
00:52:39
feels like it's really going in the direction.
00:52:41
And sometimes it does take a near death experience to get the
00:52:46
best out of people. Right.
00:52:47
I say that's the only way you change cultures is through a
00:52:49
near death experience. Do you think they'll innovate or
00:52:53
using? I mean they, you know, they
00:52:54
wrote the attention. Genesis is amazing.
00:52:57
Demis is incredibly smart and and he was like, they were the
00:53:01
ones who came up with this. I know, of course.
00:53:03
So now it's just it's playing a game of ketchup, right.
00:53:06
But it's a it's it's really David and Goliath open AM might
00:53:10
have a lot of cash, but they don't have anywhere near the
00:53:13
resources. And now today, like here's
00:53:15
here's the thing that's really interesting and people don't, a
00:53:19
lot of people don't realize this is that if you look at where
00:53:23
open air is trying to make its money and where Google is making
00:53:29
its money today, and you say, oh, and Anthropic's making it
00:53:35
Anthropic's very applications, right?
00:53:39
Or API? You're saying the API business
00:53:40
and coding? Yeah.
00:53:42
Well, API, good coding, that kind of thing.
00:53:44
Then there's, and they're doing that and they're going to get
00:53:47
there. Open AI is doing everything
00:53:49
like, you know, they're, they're picking out, OK, we're going to
00:53:52
worry about, you know, when they just announced health and then
00:53:56
they announced this. They're doing whatever it takes
00:53:58
to get more application specific.
00:54:00
And now they're announcing ads, right?
00:54:02
They're doing like again, spray and pray to hope to go because
00:54:06
people they can get only so many people to pay a $20.00 or $200
00:54:11
or whatever it is. And so they have to get to this
00:54:13
ad model to get more revenue to come in.
00:54:17
The Chinese are doing this so smart.
00:54:20
Well, maybe because they don't have the NVIDIA chips, but I
00:54:23
always like resource constrained environments because you're more
00:54:26
creative. So if you look at DeepSeek, if
00:54:28
you look at these other things and they're open source, the
00:54:33
profits of open AI today or not profits, I should say, just
00:54:37
revenues of open AI today. If those open source models or a
00:54:41
meta model, it becomes reality and they're close enough.
00:54:46
And I know a lot of developers using the Chinese models or
00:54:49
other open source models. And if Open AI doesn't get there
00:54:55
at the application side, they're going.
00:54:58
Everything in the middle's gone. They don't have data centers.
00:55:00
They don't depend. On a persistent technological
00:55:03
advantage in such a correct. And what you're gonna, what I
00:55:05
believe you're gonna see is application and to
00:55:08
infrastructure. That's where it's gonna set.
00:55:10
And in the middle, it's gonna be just a, it's gonna be like
00:55:13
everyone's got, you know, openness and open that and
00:55:16
there's gonna be no real money in it.
00:55:18
And that's why people are saying there's no money in SAS anymore.
00:55:20
So you're gonna see, oddly enough, infrastructure and scale
00:55:24
and application is where the capturing the customer or
00:55:28
capturing the platform is gonna be.
00:55:30
And that's where I believe where we're headed.
00:55:33
And opening is stuck here because they don't have any
00:55:35
data. They've been very clear they
00:55:37
don't want it, but they're not here.
00:55:39
And you already have Google up here at the apps, you have Clot
00:55:42
or a Tropic there and maybe one day Apple up there.
00:55:45
So it's pretty don't. You think like general purpose
00:55:47
assistant is an application, I mean to defend open AI?
00:55:51
They have all this context on me, you know, the longer I use
00:55:54
it. Google has much more.
00:55:56
Apple has. Much more drifting, but yeah,
00:55:59
they have much more. And that's why they want a
00:56:00
device. That's why they need all, you
00:56:01
know, these things to sort of have this whole picture.
00:56:04
They need the context. They, you know, people are
00:56:08
saying they should make a phone, but then they have, then there's
00:56:10
an Android phone because they don't have enough.
00:56:12
Well, maybe they can find more money, but they're bleeding
00:56:15
money. The other thing that people
00:56:17
don't realize is as they add more capacity, as they bring on
00:56:20
more data centers, they're all each of those users is more
00:56:25
negative money that they have to spend to keep those users,
00:56:30
right. So as they scale up and they
00:56:33
don't have a revenue out, they're losing more money per
00:56:36
user because they're getting more.
00:56:37
Users, it's the old ride sharing days.
00:56:39
So we'll make it up in volume. We're net, we're unit economics
00:56:42
are negative, but we'll make it up in volume.
00:56:44
It's like, no you won't. I mean, do you think Open AI
00:56:46
goes bankrupt? I look, if they're asking for
00:56:49
$50 billion now, like how fast you gonna burn through that?
00:56:55
We haven't talked about XAI. Do you have a view on that?
00:56:57
Look, Elon plays a very, very different game.
00:57:01
You know, he started with he he started with Tesla.
00:57:05
Then it was Twitter that turned into X and now Twitter or X is
00:57:10
part of Tesla. And then he was doing data
00:57:13
center stuff and now that's going to be part of Tesla.
00:57:15
And now he wants to make. And what I heard latest is, is
00:57:19
SpaceX is not going to go public.
00:57:21
SpaceX is going to get bought by Tesla.
00:57:23
Yeah, I think I made a prediction maybe on this
00:57:26
podcast. Yeah, I think, I mean clearly
00:57:28
so. He's just rolling up and just
00:57:30
becoming a conglomerate. Are you saying no?
00:57:32
It's sort of House of Cards, always the next thing.
00:57:35
Yeah, yeah, yeah. Oh no, I'm missing his House of
00:57:37
Cards. SpaceX is an incredible.
00:57:38
Business, OK. Incredible business.
00:57:40
Tesla would be much better off if Elon didn't do what Elon did
00:57:45
a year ago, right, Right. So that was.
00:57:47
Which is like lose focus and go do well no.
00:57:50
Be in politics and then piss off the world.
00:57:52
And everyone said I'm not buying that product because I tainted
00:57:54
the brand, right? Right.
00:57:56
So the The thing is, it's like, OK, I need money to keep Tesla
00:58:00
alive, right? OK, I'm going to put SpaceX in,
00:58:03
raise a bunch of money off SpaceX and make because the
00:58:06
Tesla stock's going to come up. And so I'll be able to continue
00:58:09
to float Tesla. You know what I mean?
00:58:11
With money, right? So that's how that works.
00:58:15
He ultimately needs to protect his investors because this whole
00:58:17
thing only works if he's like, well, he's going to look out for
00:58:20
me somehow, you know, he'll acquire Solar City some way.
00:58:23
He'll do some deal. That next thing works out and
00:58:25
like whatever good idea he has will save the rest.
00:58:30
But but you don't but like Xai you don't see as sort of a big.
00:58:33
Part of that I, I think Xai look, he's, he's going to, he's
00:58:37
doing what he does with Xai. It can be something you know,
00:58:41
again, now we're doing, you know, you know, what is it de
00:58:47
robing people on? And then it's like, I didn't
00:58:50
know. And it's like now we have
00:58:51
3 images of illegal images running around.
00:58:54
So, you know, there's this kind of skirting back and forth like
00:58:58
we just play within the Gray area of the law and then we pull
00:59:02
back just like self driving is gonna be up and running 10 years
00:59:06
ago. And he makes everyone pay.
00:59:07
But yeah, no one gets delivered what they were, right?
00:59:11
So he gets somehow gets away with everything.
00:59:14
I'm playing just the line. As a leader, you know, we sort
00:59:17
of touched on how woke potentially killed Google.
00:59:20
Now we're in this sort of like nihilist period.
00:59:23
Where do you think tech should aspire to be in terms of having
00:59:27
some meaning underpinning what what they're building right now?
00:59:31
Well, look, I'm, I'm an optimist and I'm also a guy who, you
00:59:35
know, maybe I couldn't play in this game because I'm much more
00:59:38
of the high moral standards, high ethics, high.
00:59:41
You know, when I say you're going to be my partner, you're
00:59:44
my partner and we work together on stuff, not open AI and I'm
00:59:48
going to partner with everyone and we're exclusive, but we're
00:59:50
not, you know, that's, you know, that's kind of today's social
00:59:53
media dating scheme, you know, so, so I think that, you know, I
00:59:58
don't know if I I always ask myself all the time, I'm like.
01:00:02
How would Steve Jobs fare in this environment if he was here?
01:00:06
Well, what do you think? I don't.
01:00:07
I don't. Have an answer?
01:00:08
Do you think you would be in the White House?
01:00:10
You know, you sort of have to. I have no idea I think.
01:00:12
Gavin Newsom said something fairly charitable about Tim
01:00:15
Cook, which is sort of, you know, he has to, he has to play
01:00:18
the game, but it's hard to know if Steve Jobs would have.
01:00:24
You know, I, I knowing what I know and working alongside him
01:00:30
for a decade, I don't think he would have played, but I don't
01:00:33
know this I I played along, but I don't know.
01:00:36
But so no, really, I think that, you know, I think Dario's doing
01:00:42
a great job. He got in front of Davos saying,
01:00:45
you know, this is BS, this is not right.
01:00:47
This is not Redeemis got up and saying LLMS aren't the thing,
01:00:51
you know, and there's more to come.
01:00:53
So I think we need more leaders like that because we have too
01:00:56
many, you know, people with questionable value systems
01:01:01
running these things and, and, you know, I wish they were there
01:01:05
or they were more vocal because Tim's a great guy.
01:01:09
Tim's got high high values, you know I.
01:01:12
Mean, I agree with you. I mean Dario, they just
01:01:14
anthropic just released a constitution for saw that
01:01:17
thought where it's like they're being extremely thoughtful.
01:01:19
I tweeted, It's like in a moment of anti intellectualism, here
01:01:21
they are. It's like, let's have some
01:01:23
philosophy PHD's academics, you know, think through within a
01:01:27
real principled way. Like, I really admire what
01:01:29
they're doing. I think when a lot of investors
01:01:31
and a lot of people get burned by this thing that's we're in
01:01:35
right now, there's gonna be like a lot of naval gazing going.
01:01:39
What did we do right? What do we wrong?
01:01:41
You're always gonna have that Wall Street Wolf of Wall Street
01:01:44
mentality and everybody's playing.
01:01:45
But I remember that in 9899 too, when it all fell out and then
01:01:50
great companies like Google really came into play, Right?
01:01:53
Right. So we'll see what happens.
01:01:55
I don't know. This has been a great
01:01:56
conversation. Last one last question.
01:01:59
We've touched on a lot of devices.
01:02:01
I mean, if you were, you know, I don't know, the smart engineer
01:02:04
kid out of college, like, I want to make my name in devices.
01:02:07
Like where would you point him or her?
01:02:10
Like what is the sort of device area that said OK, make a bet on
01:02:13
this for the next two decades? Understand the application
01:02:18
really, really well and understand if you even need
01:02:20
hardware. See, we have more than enough
01:02:23
hardware and all kinds of different things.
01:02:25
It's going to be a federation of things with a really smart
01:02:29
software thing. So don't think of just
01:02:32
mechanical and devices. You got to think system and you
01:02:36
got to think Federated of devices to really get to the
01:02:39
next level of what is going to happen.
01:02:42
All this other stuff is all kind of features, not a product.
01:02:46
I know that I said that was the last question.
01:02:48
I have to like, I love my iPhone.
01:02:50
I use it all the time. But don't you get we get sick of
01:02:52
looking at it or it's like you're looking at the phone too
01:02:54
much. Do you think in in two decades,
01:02:56
like we're still going to be using like a?
01:02:59
Here, which is proceeds in our ears and then and when we put
01:03:02
the phone in our pocket, it'll know the context and it'll just
01:03:04
start talking to us and then. More listening, less screen
01:03:07
time. Yeah.
01:03:08
You know, are there going to be different forms?
01:03:10
These are open ear, over the ear, in the ear.
01:03:12
The nose cancellation is great. OK, great.
01:03:14
You're going to have the glasses.
01:03:16
So we have enough of the forms they need to be pulled together
01:03:20
with smarter software and better things to get them to do what
01:03:23
they need. What?
01:03:25
Headphones do. You use these are nothings.
01:03:26
Oh, OK, these are nothing 3 as I got the over the ear, the over
01:03:30
the headphone ones. Yeah, and and also the the
01:03:33
opens. I love them, absolutely love
01:03:35
them. Tony Fidel, this has been
01:03:36
awesome. Thanks so much for doing this
01:03:37
with me. Yeah, great.
01:03:38
Thanks. Thanks a lot, Eric.
01:03:39
Thank you for tuning in to this week's episode of the podcast.
01:03:42
If you're new here, please like and subscribe.
01:03:44
It really helps out the channel. Listen in for new episodes every
01:03:46
week, wherever you get your podcasts.