00:00:03
We just had a very successful all-in
00:00:06
spirited full contact debate in front of
00:00:08
1,700 people. A packed house. There were
00:00:11
no open seats. Is the power of the
00:00:12
all-in brand is really to have the
00:00:14
conversations with a little bit of fun,
00:00:16
a little bit of spiciness and no
00:00:18
question, no topic can be banned. There
00:00:20
was no censorship.
00:00:22
No censorship. We're going to just go
00:00:23
right after the hardest topics. But we
00:00:25
had a really fun time. Discussed so many
00:00:27
important topics. What a great panel.
00:00:29
These guys are tip of the spear in terms
00:00:31
of doing really exciting things in
00:00:32
business.
00:00:40
>> Thanks for coming out everybody. Uh
00:00:41
we're going to have a great super
00:00:43
hardcore discussion about the future
00:00:46
specifically around AI which I think is
00:00:48
the most important theme not only of uh
00:00:52
CES 2026 as we've seen with all the
00:00:54
incredible gadgets chips being launched
00:00:57
self-driving but it's going to be the
00:00:59
most important transformation of our
00:01:00
lifetimes. I think everything we've seen
00:01:03
over the last 30 years of technology
00:01:05
from the PC revolution to cloud
00:01:07
computing to the internet, mobile, all
00:01:10
of that is going to be dwarfed in
00:01:12
comparison to the impact that AI is
00:01:14
going to have on society. If you're here
00:01:16
at CES, you know that you're here for
00:01:18
that reason. And we've got two amazing
00:01:21
guests who are going to join us to have
00:01:23
this debate. And additionally, I've
00:01:26
brought my box, a box filled with all
00:01:29
the ghosts and gadgets of Christmas
00:01:31
past. And we're going to go through
00:01:33
those uh at the end of our discussion.
00:01:35
But here's a quick video of our guests
00:01:37
who will be joining me today.
00:01:39
>> From boardrooms to the White House and
00:01:41
beyond, Mackenzie's influence in
00:01:43
business is virtually unparalleled.
00:01:44
>> It's one of the largest and most
00:01:46
influential consulting firms in the
00:01:48
world. enterprise can move faster than
00:01:50
any of us expected, which is good news
00:01:53
because these problems aren't going to
00:01:55
be the problems of the next generation.
00:01:56
They're going to be the problems of our
00:01:58
leadership generation. Making this
00:02:00
system of government better on both
00:02:02
efficiency and effectiveness is key for
00:02:04
economic growth and for national
00:02:06
defense. We also think that some of the
00:02:09
private sector insights that we have
00:02:11
brought to the public sector can drive
00:02:13
innovation. Our next guest leads venture
00:02:15
capital firm General Catalyst
00:02:16
>> with 40 billion in assets under
00:02:18
management as of midyear.
00:02:19
>> Our aspirations in venture capital is to
00:02:21
be the best seed firm in the world. The
00:02:23
decisions we're making, the companies
00:02:24
we're building are going to impact the
00:02:26
world for centuries to come.
00:02:29
>> Ladies and gentlemen, please welcome Bob
00:02:33
Sternfelds and Hamont Tasia.
00:02:39
>> All right,
00:02:42
gentlemen. Welcome. Yeah. All right
00:02:44
team. Let's do it. Let's do it.
00:02:48
>> How do you look at the pace of
00:02:52
innovation and change in this past two
00:02:55
years since Chat GPT was launched
00:02:57
compared to the first 30 years of our
00:03:00
careers. We're all of a certain Gen X
00:03:02
age. Compare the last two or three years
00:03:04
to the 30 before it.
00:03:06
>> Yeah. Yeah. Well, first Jason, thanks.
00:03:08
And it's great to be up here with you
00:03:09
guys. And uh yeah, I would just say this
00:03:13
week is amazing. I I um think there's
00:03:15
over 150,000 folks here this week. And
00:03:18
you talk about CES being back. I think
00:03:20
CES is back. And uh and that's great,
00:03:23
right? And uh
00:03:24
>> and with all of these things happening,
00:03:27
I think there is such a premium on folks
00:03:30
from different perspectives getting
00:03:32
together because that's where new ideas
00:03:33
are created. And and my big hope and why
00:03:36
we're here is when you mix and mingle
00:03:38
with different folks, you come up with
00:03:40
new things and the world needs new
00:03:42
things and you and what I love is you
00:03:44
mentioned a lot of the the tech leaders.
00:03:47
What's exciting for this is I think
00:03:48
everybody sees tech as part of the
00:03:50
equation. And and so when I look at the
00:03:52
folks here in CES, you see not only the
00:03:54
technology leaders, the investors, but
00:03:57
you see folks from almost every industry
00:03:59
vertical that are here now because they
00:04:01
know that technology doesn't sit on the
00:04:03
side, it's central to everything we do.
00:04:05
And I get to your question, look, I
00:04:07
think we're moving at at literally warp
00:04:09
speed now. It's just night and day
00:04:11
different. It's almost a, you know, BC A
00:04:15
type of thing when you can see the
00:04:17
change of pace. And I haven't met a CEO
00:04:20
yet that isn't talking about how do I
00:04:22
get my organization moving faster. It's
00:04:24
quite frankly less about strategy. It's
00:04:26
more about organizational speed. Hey
00:04:28
Mont, how does this feel compared to our
00:04:31
first, you know, couple of decades where
00:04:35
companies would take two or three years
00:04:36
to release a product and now companies
00:04:39
are releasing products in two or three
00:04:41
weeks, two or three months.
00:04:43
>> Yeah. So look, the world has completely
00:04:46
changed, right? We we've often said this
00:04:48
is peak ambiguity. You have massive
00:04:50
geopolitical change. You have an
00:04:52
incredible amount of change around every
00:04:53
country trying to drive strategic
00:04:55
autonomy in different industries. And
00:04:56
all those dynamics keep changing.
00:04:58
Alliances, the new world order,
00:05:00
everything. And then underneath that,
00:05:02
our tool of implementation is technology
00:05:04
that keeps changing, right? So what the
00:05:06
the technologies can build today versus
00:05:08
what the LLM could do, let's say, two
00:05:11
years ago or November 22, let's say,
00:05:12
when Chad GBD came about is
00:05:14
fundamentally different. So what are you
00:05:16
building towards as to what the world's
00:05:18
going to look like so you can have
00:05:20
enduring value? And then what are you
00:05:22
building with where the technologies
00:05:24
you're using aren't going to obsoles and
00:05:26
and destroy your value proposition over
00:05:28
time. It's just all kinds of change. And
00:05:30
so it's really dynamic time. And the
00:05:33
other thing you will see is you know we
00:05:34
invested in Stripe in 2010. It became a
00:05:38
you know a hundred billion dollar
00:05:40
company let's say 12 13 years later. You
00:05:42
look at Antropic,
00:05:44
which we're also investors in, that goes
00:05:46
from $60 billion last year to, you know,
00:05:50
a couple hundred billion. So like, and
00:05:52
by the way, with good economic uh
00:05:54
progress, these are not pie in the sky
00:05:56
valuations. They're based on actual
00:05:57
growth of the business. Well, that goes
00:05:59
back to your point, which is the
00:06:00
compression of how fast value can create
00:06:03
when code self-writes and access to
00:06:05
distributions change. So fundamentally,
00:06:07
it's just it's just really exciting and
00:06:08
I think it's going to accelerate from
00:06:10
here. This was one of the statistics we
00:06:12
would look at in venture capital. Hey,
00:06:14
how long does it take this company to
00:06:15
get to a hundred million in revenue? How
00:06:17
long does it take to get to a billion in
00:06:19
revenue? Unpack anthropic and that
00:06:22
journey because this company's revenue
00:06:25
and you have open AI obviously they're
00:06:27
contemporary trending towards $20
00:06:29
billion in revenue a year. Where's
00:06:31
anthropic at and what's the revenue mix?
00:06:35
Where does the revenue come from? Well,
00:06:36
well, look, so Enthropic builds u
00:06:39
language models. It's got one of some of
00:06:41
the best models out there. There's a
00:06:42
couple of companies that are doing a
00:06:43
good job at that. And then they've got
00:06:45
cloud on top, which is to me the essence
00:06:47
of transforming the engineering
00:06:48
department of enterprise, right? And
00:06:50
that's a killer application where
00:06:52
everybody is now using these tools. So
00:06:53
that business when we invested was doing
00:06:56
about $880 million, which was a 10x
00:06:58
growth from the year before,
00:07:00
>> 10x year,
00:07:01
>> 10x growth the year before. and and then
00:07:03
and this last year they've announced
00:07:05
this they're growing another 10x or more
00:07:07
and so when you look at that and we
00:07:08
invested uh at the $60 billion valuation
00:07:11
assuming it's going to be like 3x growth
00:07:13
from there because those are staggering
00:07:14
numbers and it does 10x can't predict it
00:07:17
uh but to see adoption is so fast and so
00:07:19
you look at that and say so we ended up
00:07:21
investing at you know um a 8910 billion
00:07:26
kind of run rate business at 60 billion
00:07:29
that's the cheapest deal that got done
00:07:30
last year in venture capital on
00:07:32
financial spaces. So, so we just have to
00:07:34
get our head around what does scale
00:07:36
really mean and what you know are we in
00:07:39
the business of creating what we used to
00:07:40
think was like can we create decoorns
00:07:42
now we're talking about can we create
00:07:43
trillion dollar companies right I mean
00:07:45
that's not a pie in the sky idea with
00:07:47
anthropic and uh open AI and a couple
00:07:49
others game changed scale of technology
00:07:52
is uh you know fundamentally different
00:07:54
in what it can do
00:07:55
>> Bob what's behind this massive revenue
00:07:58
ramp because you get to see all the
00:08:01
incumbent businesses you get see the
00:08:04
elite businesses that are growing, you
00:08:06
know, two or three times X each year.
00:08:08
You also get to see the ones that are
00:08:10
struggling and then you see these large
00:08:13
numbers and 10x your growth. What what's
00:08:16
driving this in your mind and is it
00:08:18
sustainable? I mean, that's the other
00:08:20
question. I hate to give you the the
00:08:22
classic consultant answer, but I you
00:08:24
know, I I do think it depends. And I I
00:08:26
think we're at a we're at a we're at a
00:08:27
tipping point this year. And I'll tell
00:08:29
you why. I think what's underpinning
00:08:31
this 10x and 10x you know we work with
00:08:34
most of the large enterprise in the
00:08:36
world um across all industry verticals
00:08:38
and what we have seen is a huge uptake
00:08:41
of leveraging these technologies like
00:08:44
anthropic we're we're we're leveraging
00:08:45
anthropic and so large enterprise is
00:08:48
using technology at a scale and rate
00:08:50
that they haven't before and if you look
00:08:51
at IT spend as a percent of revenue etc
00:08:54
all this stuff has gone up and I think
00:08:55
that is propelling the 10x to 10x um The
00:08:59
conundrum is, and it's, you know, been
00:09:02
widely written about, um,
00:09:05
realizing enterprise at scale value in
00:09:08
non-technology companies is proving
00:09:10
harder than people think.
00:09:11
>> Got it. So, in plain English, that
00:09:13
means, hey, you've got a travel company.
00:09:16
There's somebody deploying AI and you're
00:09:19
watching what's happening at Tesla or
00:09:21
Google, and they're getting these
00:09:23
phenomenal results, but maybe that
00:09:24
legacy business is having a harder time
00:09:27
achieving those results. I'll make it
00:09:28
even simpler. Uh typical non- tech CEO
00:09:32
might say, "Hey, Bob, do I listen to my
00:09:34
CFO or my CIO right now,
00:09:36
>> right?" CFO is saying, "We've spent all
00:09:38
this money. Why do we need to be the
00:09:40
fast adopter? I'm not seeing the ROI
00:09:43
yet. Can we pause?" CIO is saying, "Are
00:09:45
you freaking crazy? This is the moment
00:09:47
that if we don't, we'll be disrupted."
00:09:50
We think, now I will say the the shining
00:09:52
part is I think there's a path where you
00:09:54
bring those two together as allies. and
00:09:55
you say, "Yeah, but let's rethink this,
00:09:58
get out of pilot purgatory, really think
00:10:00
the reorganization, all this stuff."
00:10:01
There is a path, but I think right now
00:10:03
most CEOs are getting torn a bit
00:10:05
between, "Do I listen to my CFO or do I
00:10:07
listen to my CIO?" And I think this is a
00:10:10
really good jump off point, Hamont, of
00:10:13
your strategy at General Catalyst. You
00:10:15
and I have known each other for a long
00:10:16
time. You really a long time, decades,
00:10:19
and you always prided yourself on being
00:10:23
the great seed fund. We're gonna we're
00:10:25
going to get to these companies when
00:10:27
they're $10 million and 10 people and
00:10:29
put that first check in. But then I saw
00:10:32
this news item go by a month ago that
00:10:34
you raised N billion and then I see
00:10:36
you're buying companies. So are you out
00:10:39
of the seed business and now doing
00:10:41
random acts of private equity? What's
00:10:43
going on here?
00:10:44
>> Explain to me the strategy and general
00:10:45
catalyst.
00:10:46
>> How much time do we have? It's going to
00:10:47
take some time. So, so look, I would say
00:10:50
um uh we very much view ourselves as a
00:10:53
venture capital for Americ
00:10:58
for the 25 years we've been around has
00:10:59
been meeting founders where they are.
00:11:01
And what that means is essentially help
00:11:04
them navigate ambiguity in the past when
00:11:06
they're sort of beginning and the
00:11:07
business isn't clear uh all the way to
00:11:09
figuring out how to scale in the complex
00:11:11
markets that they go into. So everything
00:11:13
we've done has been in that context of
00:11:16
creating these catalysts the flexible
00:11:18
capital they need the policy
00:11:20
capabilities they need the market access
00:11:22
they need uh and those sort of
00:11:25
relationships globally to actually build
00:11:26
an enduring company. So that hasn't
00:11:28
changed. So So why did we go um you know
00:11:32
acquire a health system in u uh Ohio was
00:11:36
a nonprofit. We worked with the attorney
00:11:38
general converted it. And I say this by
00:11:39
the way with a great sense of
00:11:40
responsibility because that's a
00:11:41
community of uh in Akran Ohio that we
00:11:44
take care of. So that hospital has to
00:11:47
continue operations in all the
00:11:48
dimensions. It takes care of um with
00:11:50
people. We bought it to actually have a
00:11:53
place where we can work with our
00:11:55
founders and transform with AI create
00:11:58
abundance and resilience for this health
00:12:00
system so we can take care of the people
00:12:01
a lot better. Okay. And if we if we did
00:12:03
that then we can go do that for the
00:12:05
other hundreds of systems across the
00:12:06
country and be able to do that. So some
00:12:08
of it is that's market access. It's very
00:12:10
hard for healthcare startups to go
00:12:12
deploy successfully at scale in these
00:12:14
systems. We are going to go show how.
00:12:17
we're gonna actually go on the ground
00:12:18
with do with them and show what the
00:12:19
world how so it can transform the health
00:12:21
system. The you know your other point
00:12:24
about uh sort of buying uh companies we
00:12:28
look at that as there's a lot of
00:12:29
workforce transformation happening. Bob
00:12:31
and I talk a lot about this. I think
00:12:32
work fundamentally in these companies is
00:12:34
going to change. So if you're a call
00:12:35
center in uh an emerging country today
00:12:39
the a declining asset value because you
00:12:41
know it's going to be place displaced
00:12:42
with AI. So we look at that and say well
00:12:44
but those are customers on the other
00:12:46
side. If we bought that as a piece of
00:12:48
the puzzle to work with an early stage
00:12:51
founder to learn how to quickly
00:12:53
accelerate adoption of AI into uh the
00:12:56
call center space and serve these
00:12:57
customers and scale a lot faster. The
00:12:59
compressed value creation we're talking
00:13:00
about that is uh that is a new playbook.
00:13:03
So this is not about trying to be PE.
00:13:06
This is about acquiring uh uh businesses
00:13:09
in PE that actually have declining value
00:13:11
but have important customers that need
00:13:13
to be served and help them get to that
00:13:15
AI transformation that Bob's talking
00:13:17
about faster by getting our founders in
00:13:19
there. This is extraordinary Bob when
00:13:22
you think about it just so the audience
00:13:24
can put their head around this venture
00:13:26
capitalist used to back founders to then
00:13:28
be the barbarians at the gate to try to
00:13:31
take on these big industries. Now these
00:13:33
big industries in some cases are in
00:13:35
significant decline struggling and the
00:13:38
venture capitalists are coming in and
00:13:39
saying we'll just buy the castle, open
00:13:42
the draw bridge. We're going to buy it
00:13:44
so that we can take our startups and
00:13:46
accelerate whether it's healthcare,
00:13:48
financial services or customer support
00:13:52
uh and outsourcing, business process
00:13:54
outsourcing. And essentially we don't we
00:13:57
don't care about that business
00:13:58
economically necessarily as much as we
00:14:01
care about it for access to that
00:14:02
customer base
00:14:04
running McKenzie. This is a playbook
00:14:07
that is like coming out of the future in
00:14:10
a time capsule and saying we're going to
00:14:12
just upend the entire ecosystem. Yeah.
00:14:16
>> Yeah. I mean, it's I I just gave a talk
00:14:18
at a um a university and uh was was
00:14:22
talking to some potential folks to to
00:14:24
join us and I said, "Look, um I'm
00:14:26
jealous. I'm jealous of all of you
00:14:29
because you have a lot more time to do
00:14:32
what we do than I do. And you're doing
00:14:35
it at a time where it's going to be a
00:14:36
lot more exciting because if you just
00:14:38
link this and hey, what I love is um it
00:14:41
effectively you're creating a new asset
00:14:43
class, right? This is not private
00:14:44
equity. This is about how do you
00:14:47
transform incumbent entities into
00:14:50
something different, right? Private
00:14:52
equity typically optimizes an existing
00:14:55
asset class at a certain scale. This is
00:14:57
about transformation. So you think of
00:14:59
large existing enterprise and I think
00:15:02
you have a choice. You have a choice of
00:15:05
transform or die.
00:15:07
>> And so there's this wonderful moment and
00:15:09
but because of some of the incumbent
00:15:11
advantages, I wouldn't say that it's
00:15:12
predetermined which way you're going to
00:15:14
go. Right.
00:15:14
>> Right. you can actually do this quite
00:15:16
quite quickly and I think you're showing
00:15:17
the power of private capital can
00:15:20
actually do this.
00:15:20
>> So we've chatted a lot about this right?
00:15:22
So one of the things when you think
00:15:23
about transforming a large enterprise
00:15:25
what do you really need? You need a few
00:15:27
pieces. One is you need data
00:15:29
infrastructure that can ready you for
00:15:31
the enterprise. You need the models
00:15:32
adapted to you and then you actually
00:15:35
need a new model for how the workforce
00:15:37
is going to function because you have
00:15:38
agents and humans and there's a massive
00:15:40
change management exercise. So a lot of
00:15:43
our partnership has been about sort of
00:15:45
figuring out okay what is that new model
00:15:46
going to be to transform these
00:15:47
businesses and what does that mean when
00:15:49
you get on the ground you look at
00:15:50
department by department take HR how do
00:15:53
you drive uh transformation of
00:15:56
healthcare and how you take care of your
00:15:57
people that process is horrible today
00:15:59
and we have a business called
00:16:00
transparent that goes and essentially
00:16:03
creates abundance in that regard which
00:16:05
uses AI so you have direct access
00:16:07
pentically to all kinds of healthcare
00:16:10
services take yourself and then be able
00:16:12
ble to be routed whether you need a
00:16:13
surgery or you need to have cancer
00:16:15
therapy or mental health and do it in a
00:16:18
way that is seamless cost effective and
00:16:19
helps uh enterprises take control of
00:16:21
your cost structure you need coding to
00:16:24
fundamentally transform that's what
00:16:25
anthropic does there are companies
00:16:27
working on transforming the call centers
00:16:29
there companies working on transforming
00:16:31
your sales and marketing but then when
00:16:33
you have these technologies in there how
00:16:35
are the actual people going to do their
00:16:37
work in concert with these agentic
00:16:39
capabilities that is a whole new model
00:16:42
and you know you guys are inventing a
00:16:43
talk about that because there's a lot of
00:16:44
in innovation that needs to happen in
00:16:46
that that whole sort of workforce
00:16:48
transformation that I think is
00:16:49
ultimately where rubber is going to meet
00:16:50
the road on how quickly teams embrace it
00:16:53
customers embrace it and we can actually
00:16:55
diffuse AI into these businesses
00:16:57
>> and Bob you've had to deal with this
00:16:59
internally at your organization what's
00:17:01
the right size and what happens when a
00:17:04
piece of technology takes a career and
00:17:08
takes out the first five years
00:17:11
>> what happens to an organization when you
00:17:13
just basically gut the first five years
00:17:15
of development. And this is why some
00:17:18
people in the economy are looking at AI
00:17:21
and they're scared
00:17:23
>> and they're looking at AI and saying is
00:17:25
this going to benefit me, my family, my
00:17:27
kids who are graduating from school,
00:17:29
this technology, and I think management
00:17:31
consulting is the perfect place to look
00:17:33
at it. Tell me honestly, the first
00:17:36
couple of years you're training up one
00:17:37
of these really smart kids to write up,
00:17:39
you know, reports and do analysis that
00:17:42
can be done with AI today perfectly,
00:17:47
close to perfectly.
00:17:49
>> Yeah. Um, so I'll give you a a couple
00:17:52
stats first on us, then more general.
00:17:54
Um, 25 squared and 40,000 and 25,000.
00:17:59
What do I mean by that? So, let's look
00:18:01
at at McKenzie as a bit of a of an
00:18:03
incubator. The 25 squared is we're
00:18:05
simultaneously doing two things at the
00:18:07
same time. So we have client-f facing
00:18:10
folks which most of you in the audience
00:18:12
would know and think about when you
00:18:14
think about a Mackenzie consultant.
00:18:15
We're growing that body at 25% next
00:18:18
year. 25%. Unprecedented number of uh of
00:18:22
of of new hires because the work is
00:18:26
changing. They're not doing the stuff
00:18:28
that you talked about. We saved we
00:18:30
looked at it. We saved 1.5 million hours
00:18:33
in search and synthesis last year, but
00:18:36
we're dividending that to solve more
00:18:38
complicated problems and do different
00:18:39
things. You guys are probably sick of uh
00:18:41
of McKenzie charts out there. You know,
00:18:44
we have agents that do this. They just
00:18:45
gave you 2.5 million of them in the last
00:18:47
6 months. I want to get rid of charts.
00:18:49
But the the consultants are doing
00:18:51
different things. We're adding 25% to
00:18:52
that body.
00:18:53
>> So, they're moving up the stack.
00:18:55
>> They're moving up the stack and doing
00:18:56
these more complicated problem. At the
00:18:58
same time though about half our firm uh
00:19:01
are nonclientf facing folks. We're down
00:19:03
25% in that group um with 10% increase
00:19:07
in output.
00:19:08
>> And so simultaneously and I know this is
00:19:11
hard also particularly for you know for
00:19:13
folks to get we're going to be adding
00:19:15
and shrinking simultaneously with the
00:19:17
two halves.
00:19:18
>> And this has never happened in the
00:19:19
history of the firm.
00:19:20
>> I've our model has always been
00:19:22
synonymous that growth only occurs with
00:19:25
total headcount growth. Now it's
00:19:27
actually splitting. We can grow in this
00:19:29
part, the client facing side, and we can
00:19:31
shrink in this part and have aggregate
00:19:33
growth in total. And that's, you know,
00:19:35
that's a new paradigm and a new new
00:19:37
dynamic.
00:19:38
>> We're seeing this in venture. You and I
00:19:41
were around for the days where you'd
00:19:43
give a team $3 million and they would
00:19:46
come back in 18 months having spent it
00:19:48
on data centers and building a team of
00:19:51
20 people, 30 people. Then we'd see the
00:19:52
first version of the product 18 months
00:19:54
later. And then the three billion now.
00:19:57
>> Yeah, it's but it's it's insane just uh
00:20:00
how much more is getting done with less.
00:20:02
And so do we worry about society and our
00:20:09
industry's ability to communicate to
00:20:11
society this change? Because if you were
00:20:15
to tell an average business executive 10
00:20:18
years ago, prepare to hire 25% more
00:20:20
people on this side of the business and
00:20:22
cut 25% on this side of the business. In
00:20:24
the same 18month period, their head
00:20:25
would explode.
00:20:26
>> Exactly.
00:20:26
>> What? Why? How come? It doesn't make any
00:20:29
sense. And then young people are
00:20:30
graduating and they're sending out a 100
00:20:34
200 resumes getting no job offers. We
00:20:37
were sitting here 10 years ago. every
00:20:39
graduate from a decent school was like,
00:20:41
I have an Uber, a Coinbase, and a Google
00:20:43
offer. Which one should I take for 150K?
00:20:46
And those offers just aren't there. So,
00:20:49
how do we communicate better as an
00:20:51
industry? And then what's the advice to
00:20:53
young people coming into the workforce?
00:20:55
>> Yeah, look, um, every company in I was
00:20:59
just say Silicon Valley, but broadly in
00:21:00
the tech industry, in the startup
00:21:01
industry, what was it? Essentially, it
00:21:02
looks like a C corp with a bunch of
00:21:04
engineers. But in a world where code
00:21:06
self-writes,
00:21:08
what what is that next level innovation?
00:21:10
What are these companies actually going
00:21:11
to do? I think that's ultimately the
00:21:12
transition we're going through uh in
00:21:14
what does innovation actually mean. It's
00:21:16
going to be less about being able to
00:21:18
write code fat. It's much more going to
00:21:19
be about systemically how do we adopt
00:21:21
this uh into the world and capabilities
00:21:24
to your point about ambiguity in the
00:21:25
opening because we don't know about the
00:21:27
capabilities of these technologies or
00:21:29
the how the world's shaping up. There's
00:21:30
a lot of ambiguity. So to me the
00:21:33
companies that do really well and the
00:21:35
the guide the way we guide the founders
00:21:37
a lot of it is become iterative
00:21:40
constantly change constantly move
00:21:41
forward as opposed to what used to be
00:21:43
before was become precise find this
00:21:45
narrow edge create your growth loop and
00:21:48
go build a company. Now it's like
00:21:49
constantly iterate and in order to have
00:21:51
the customers give you the license to
00:21:53
iterate comes down to trust and
00:21:55
relationships. So founders that are very
00:21:58
good at engaging with customers,
00:22:00
building trusting relationships and say,
00:22:01
"Hey, we're going to go figure this out
00:22:03
together. We know how to leverage this
00:22:05
technology, but we don't really know how
00:22:07
where it leads and what's what the
00:22:08
possibilities are." Will we co-create?
00:22:10
And so it the advice I always give is
00:22:13
it's all about radical collaboration. In
00:22:14
this next phase, we got to figure this
00:22:16
out together where different
00:22:17
stakeholders that all touch a system are
00:22:20
figuring out what this means to them and
00:22:22
then what it means in terms of an
00:22:24
overall optimization and the
00:22:25
transformation that we can do with it.
00:22:27
you know, one maybe um exciting part to
00:22:30
this because I I think you know, you
00:22:32
framed it as you're a graduate and how
00:22:34
do you get into the workforce and is it
00:22:36
getting tougher and um we we did a
00:22:39
little bit of work that said um what
00:22:41
kind of skills are folks going to need
00:22:43
in an AI infused world, right? From an
00:22:46
employer's point of view, less the the
00:22:47
startup, but you're more an atscale
00:22:49
enterprise. What can the models not do?
00:22:52
And so therefore, what skills will
00:22:53
humans play? And the work isn't done but
00:22:56
but came back with with kind of three
00:22:59
key ideas. Um what can the models not
00:23:01
do? Aspire, set the right aspiration. Do
00:23:05
you go to low earth orbit? Do you go to
00:23:07
the moon? Do you go to Mars? That's a
00:23:09
uniquely human capability. So how do you
00:23:12
look for the skills about aspiring and
00:23:14
getting others to believe in the
00:23:16
aspiration?
00:23:17
>> So leadership and goal setting
00:23:19
>> human
00:23:20
>> human
00:23:20
>> judgment, right? And we've seen a lot
00:23:22
around in this room eval. But there's no
00:23:26
right and wrong in these models. And so
00:23:28
how do you set the right parameters? The
00:23:30
architecture um based on firm values,
00:23:33
based on societal norms, whatever. How
00:23:35
do you build the skills to set what the
00:23:37
right parameters are? And then finally,
00:23:41
true creativity, right? The models are
00:23:43
inference models. The next most likely
00:23:45
step, how do you think about orthogonal
00:23:47
stuff? And so some of the work we've
00:23:49
been doing with large enterprises, if
00:23:51
you believe in some of that, it can take
00:23:53
you back to challenging some of your
00:23:55
assumptions on where you look for
00:23:56
talent. It actually means that where you
00:23:59
went to school matters a lot less. And
00:24:01
so, do you start looking for raw
00:24:03
intrinsics? Can you widen the base? Can
00:24:06
you actually look at, let's take a tech
00:24:08
background, not which university you
00:24:10
graduated from, but what does your
00:24:12
GitHub profile look like? Let's actually
00:24:14
get to the content. And could that
00:24:16
actually start meaning that a wider set
00:24:18
of people can enter the workforce with
00:24:20
different pathways?
00:24:21
>> One of the things you said which really
00:24:23
resonates is around creativity because
00:24:25
what when you know we were going to
00:24:28
college it was all about learn how to
00:24:29
solve problems really well right and and
00:24:32
and now in a world where we have this
00:24:34
technology that can solve problems for
00:24:35
us. It really is about asking the right
00:24:37
questions. It's like going back to that
00:24:38
socratic dialogue. It is about
00:24:40
creativity and who can imagine best what
00:24:43
the world's going to look like and then
00:24:44
leverage these technologies to go shape
00:24:46
the world towards that. To your point
00:24:47
about vision and and teaching our kids,
00:24:49
you know, I get this question a lot
00:24:50
about what do you want how do you want
00:24:52
your kids growing up? It's like learning
00:24:54
how to ask the right questions versus
00:24:55
solving how to you know work on hard
00:24:58
problems is a it's a very different
00:25:00
mindset and it is about curiosity and
00:25:02
kind of back to being kids when you're
00:25:03
growing up. It is about challenging your
00:25:05
curiosity. Can we actually rethink our
00:25:07
pedagogy in a way that we can develop
00:25:10
this next generation to be more that
00:25:12
than it's 8:00 on Wednesday morning and
00:25:14
I'm going to factor polinomials cuz I'm
00:25:16
in you know 7th grade which is what our
00:25:18
system looks like today.
00:25:20
>> Yeah. The the advice I've been giving to
00:25:22
young people is there's nobody coming
00:25:25
for you. There's no training program.
00:25:27
You have to make that for yourself
00:25:30
>> and do not go in through the front door
00:25:32
with a resume. just email the CEO of the
00:25:34
company
00:25:35
>> and redesign their landing page and say,
00:25:38
"These are the three things that I think
00:25:39
could be better and I saw you speak on
00:25:41
this podcast. I think your company's
00:25:43
incredible. I would love to come work
00:25:44
there and I did this spec work." Now,
00:25:46
people are like, "Why should I do free
00:25:48
work to get a job to prove you actually
00:25:51
have a skill that is meaningful?" You're
00:25:53
not going to be able to get into a
00:25:55
training program. So many folks now in
00:25:58
corporate America, especially the the
00:26:00
people who are onboarding people are
00:26:01
just like hiring somebody and training
00:26:03
them is going to take longer than
00:26:05
building an agent. I can build an agent.
00:26:08
Young people coming into the workforce I
00:26:10
have to train are annoying.
00:26:13
Setting up an agent who just does the
00:26:15
work is easy. That's the game on the
00:26:17
field right now that people don't want
00:26:18
to talk about. Which means to stand out,
00:26:20
you're going to have to show hootsp.
00:26:22
You're going to have to show drive.
00:26:23
You're going to have to show passion.
00:26:25
And what college is doing that? What
00:26:27
college is teaching that? What course is
00:26:28
that?
00:26:29
>> We look, I think there's a massive gap
00:26:31
in resilience. Yes. You know,
00:26:34
resilience. Because what you've got
00:26:35
under that is you're going to get
00:26:36
knocked down.
00:26:37
>> Yeah. Right. The question is, do you get
00:26:39
back up?
00:26:40
>> And how do you get back up? And I think
00:26:42
the the educational system today doesn't
00:26:45
necessarily build institutional or
00:26:47
individual capability in resilience. If
00:26:49
we could wave a magic wand, just to go
00:26:51
off on a complete tangent here, what
00:26:53
should the education system look like in
00:26:55
2026? Because you're buying businesses,
00:26:58
you have one in healthcare. That's one
00:27:01
of the three hardest businesses to make
00:27:03
change in historically. The other two
00:27:06
here in America that have the most
00:27:08
regulation, are the most expensive, and
00:27:10
are the hardest, and that Americans are
00:27:12
suffering under the most are housing and
00:27:13
education. Those are the three big ones.
00:27:15
When I run for president, that's going
00:27:17
to be my platform.
00:27:18
is those three. I'm going to solve those
00:27:20
three. But go ahead and solve education
00:27:22
for us right now. And are you going to
00:27:23
buy a college next
00:27:26
>> transform?
00:27:27
Basically buying business all the
00:27:28
businesses that make no money. Is that
00:27:30
where we're going?
00:27:31
>> Um I I would say
00:27:32
>> oh the businesses that are the most
00:27:34
>> Yeah. But the ones that need to endure
00:27:36
for the longest actually. That's the way
00:27:37
I look at it. So here's the thing about
00:27:38
education. This idea that we spend 22
00:27:41
years learning and then we spend 40
00:27:43
years working is a broken idea. if if
00:27:45
the the learning of technology and the
00:27:48
development of technology is going to be
00:27:49
so dynamic. So what about going from a
00:27:52
four-year college to a lifelong college
00:27:54
which is actually your relationship with
00:27:56
learning is that it's a lifelong
00:27:58
skilling and reskilling outcome
00:27:59
experience. We've talked about this
00:28:00
before as well. I mean there's there's
00:28:02
there's some innovative college
00:28:03
presidents that are thinking about that
00:28:04
which is first of all better business
00:28:07
better lifetime value if you're a
00:28:09
college and you actually have a a client
00:28:11
or a student you know for perpetuity
00:28:13
versus paying you for four years. and
00:28:15
much more um useful for us to be able to
00:28:19
go and have that capability and
00:28:21
constantly learn what these cap
00:28:22
technologies are doing and how the
00:28:23
workforce is evolving and how to stay
00:28:25
ahead in terms of where the opportunity
00:28:27
is. This so learning has to become much
00:28:30
more fluid and we need to become a
00:28:31
community of lifelong learners uh as as
00:28:34
we adapt to a world where AI is being
00:28:36
diffusing you know through us over the
00:28:38
years. And I would just add uh I'm I'm
00:28:42
with you on this and you know that the
00:28:44
system built close to 700 years ago was
00:28:48
designed around a high fixed cost
00:28:51
libraries and professors to then take
00:28:55
you out for a a finite period of time to
00:28:57
learn and then effectively you're set
00:28:58
off into the workforce. I if you start
00:29:01
to think about the half-life of skills
00:29:04
getting shorter and shorter and we've
00:29:06
done some work at our global institute
00:29:07
that said for an employer the return on
00:29:10
investment that you give an employee in
00:29:12
terms of skills has shrunk by about half
00:29:14
over the last 30 years used to be about
00:29:15
seven years return it's less than four
00:29:17
years so about 3.6 six years now return
00:29:20
and that's only getting shorter and
00:29:21
shorter as things change. So if you if
00:29:23
you believe that I think you start to
00:29:26
pivot to are we teaching people to
00:29:28
continue to learn new things as opposed
00:29:31
to master a particular subject. Do you
00:29:33
have that ability? Um, one of the things
00:29:35
that um, we've now indexed on and I
00:29:38
mentioned this 40,000 and 25,000 that is
00:29:41
the number of humans we have and the
00:29:43
number of personalized agents we have as
00:29:45
of last week in McKenzie and I think
00:29:47
we'll be at parody before the end of uh,
00:29:49
by the end of this year. So you're
00:29:50
literally deploying agents that can do a
00:29:53
full 360 degree trusted job function.
00:29:57
>> Absolutely.
00:29:58
>> Where is it working really well and
00:29:59
where is it not working well?
00:30:02
It works when you have a specific um
00:30:05
domain area that you know ultimately
00:30:08
where value can be created. So for us
00:30:11
that's in structured problem solving.
00:30:13
It's in around search and synthesis.
00:30:15
It's around commu more effective
00:30:17
communication these types of of domain
00:30:20
areas. But where I was going with this
00:30:22
so the skill is are you skilling people
00:30:25
to actually become superhuman by
00:30:27
leveraging agents.
00:30:28
>> Right. right? That becomes a skill. And
00:30:31
I don't think we're actually equipping
00:30:32
that right now. It's a bit more random
00:30:35
or sometimes actually excluded in the
00:30:38
classroom as opposed to embracing it and
00:30:40
figuring out how do you actually take
00:30:41
advantage of
00:30:41
>> it. It's almost like we're we need to
00:30:43
train people to go from being part of
00:30:45
the orchestra to everybody being the
00:30:47
conductor
00:30:48
>> and everybody having their own orchestra
00:30:50
of agents working for them. And you
00:30:52
know, I always look to startups because
00:30:55
they're resource constrained. And I was
00:30:57
at a dinner in Singapore and I had a
00:31:00
dozen founders there and I said, "Has
00:31:02
anybody um you know hired anybody in the
00:31:05
last, you know, 60 days?" They all
00:31:07
raised their hands and then I said,
00:31:09
"Okay, um how many of you have an HR
00:31:12
person who wrote the job description?"
00:31:14
Nobody raises hand. I said, "How many of
00:31:16
you typed into an LLM, write a job
00:31:20
description for this?" All 12 hands go
00:31:21
up. So now you think just HR the entire
00:31:26
blocking and tackling of it has been
00:31:28
writing the job description and sorting
00:31:29
through the resume. So then I asked the
00:31:31
next question which was how did you sort
00:31:32
through the resumes coming in and they
00:31:34
said half of them had built agents
00:31:37
>> Mhm.
00:31:37
>> to sort through the resumes and stack
00:31:39
rank them using AI and I said whoa holy
00:31:42
cow like this is like the typing pool
00:31:45
the mail room the photo for those of you
00:31:48
who are under 40 years old. We had a
00:31:50
room which was called the typing pool.
00:31:52
Then we had one called the mail room
00:31:53
where packages came in messengers all
00:31:55
those went away. that floor of the
00:31:57
building got redeployed and I think
00:32:00
that's what we're going to see like the
00:32:01
HR department the legal department all
00:32:03
getting compressed really interesting
00:32:05
>> it's already happening and I think as we
00:32:07
think about our own transformation for
00:32:08
our own business we basically say every
00:32:10
department needs to have AI teammates
00:32:12
now are those AI teammates like the
00:32:14
co-pilot or pilot can you fully empower
00:32:16
them to do stuff or are they giving you
00:32:18
efficiency that depends on how well the
00:32:21
technology works how complex the problem
00:32:22
is how severe the problem is so like in
00:32:25
healthcare for example if it's life and
00:32:26
death decisions, you want humans making
00:32:28
those uh today because that technology
00:32:31
isn't uh as reliable. So, so I think
00:32:34
sort of having a framework but saying
00:32:36
every one of your departments is going
00:32:37
to have these AI agents. Um, if you're
00:32:40
not doing that then you're not preparing
00:32:41
yourself for this next phase and that's
00:32:43
that's a lot of what you're see you're
00:32:44
already going to be one to one that's an
00:32:46
enormous uh ratio.
00:32:48
>> Well, and but the problem I think Jason
00:32:49
that you alluded to earlier in this is
00:32:51
there's all this potential um but folks
00:32:54
aren't thinking through the dynamic
00:32:56
implications in their enterprise model
00:32:59
versus the static. So the static might
00:33:01
be hey there's all these departments I
00:33:04
can apply this I'll radically shrink it
00:33:06
I'll reduce the number of layers in an
00:33:07
organization I may slow hiring on the
00:33:09
inbound to your point the dynamic is
00:33:12
okay but what does your company look
00:33:14
like in five years time
00:33:16
>> and what I also often ask a CEO is okay
00:33:19
you're doing all this stuff what's the
00:33:20
pathway to your job
00:33:22
>> how does somebody get to your job in the
00:33:24
org of the future you had a pathway it's
00:33:26
not going to be that same pathway but if
00:33:28
you don't hire inbound folks You can't
00:33:30
continually laterally bring in a CEO.
00:33:33
It's literally like taking the the
00:33:34
bottom four rungs off the ladder to save
00:33:37
money today.
00:33:38
>> And then everybody's jumping up trying
00:33:39
to get in the organization. It's like,
00:33:41
well, we don't have a path there. You're
00:33:42
going to have to be really thoughtful
00:33:43
about making that investment. And it
00:33:46
feels like the first two years of AI
00:33:47
were about cutting jobs and we really
00:33:50
need to think about, hey, it's not just
00:33:51
about efficiency, it's about
00:33:52
opportunity.
00:33:53
>> Exact. What's that other 25%. Right.
00:33:55
That's what I think we got to lean into.
00:33:57
Let's take a little diversion here
00:33:58
before I open my box.
00:34:00
>> The black box,
00:34:01
>> my box here of all the great CES
00:34:04
innovations over the last 20 years. Um,
00:34:08
physical AI. We've been talking here
00:34:10
very cerebrally about what's happening
00:34:12
in enterprises, what's happening in
00:34:14
software. But, um, you have self-driving
00:34:17
is probably the theme. I would I would
00:34:19
dub 2026 CES as self-driving CES. I will
00:34:24
dub 2027 as robotics, humanoid robotics
00:34:28
specifically. We're starting, obviously
00:34:30
people are showing off all these
00:34:31
incredible robots here, but I think
00:34:33
consumers will be experiencing them in
00:34:35
27, but consumers are experiencing this
00:34:37
year self-driving. Neuro and Lucid have
00:34:41
an incredible product. Zuks has been
00:34:42
here. Obviously,
00:34:45
Elon's doing great things with Roboaxi.
00:34:47
Feels like he's closing in on a solution
00:34:49
and getting very close. Whimo obviously
00:34:51
is leading the pack but then you also
00:34:53
have BU Libaba we write Pony AI this is
00:34:57
a global race what will the world look
00:34:59
like in 2026 in terms of self-driving
00:35:02
and then any second and third order
00:35:04
impacts of those and then do the same
00:35:06
for robotics
00:35:07
>> if you go around the world today right
00:35:09
you you go to Europe you go to the
00:35:11
Middle East where um you know there's a
00:35:14
there is focus on interesting luxury
00:35:17
products there's a market for it BYD and
00:35:19
a lot of these Chinese companies are
00:35:20
actually penetrating deeply everywhere
00:35:23
because these these cap these companies
00:35:25
have all the features and functionality
00:35:26
and they're really low cost and so one
00:35:29
thing is that the the the dynamic of the
00:35:31
auto industry and European auto makers
00:35:34
are all very dejected because they don't
00:35:35
know how they're going to compete with
00:35:36
Chinese industry. US has innovation
00:35:40
self-driving be innovation which allows
00:35:41
you to say the next generation of
00:35:43
winning automotive companies um will
00:35:45
take advantage of this platform shift.
00:35:47
US has a technology but it doesn't have
00:35:49
the manufacturing uh capabilities to
00:35:52
actually say can you actually make it as
00:35:54
cost- effectively as uh a Chinese maker
00:35:56
is going to be be. So it's not it's not
00:35:58
as easy to figure out how the the world
00:36:00
order around automotive is going to re
00:36:02
shift around the world. uh and and so
00:36:05
part of the physical AI and the use of
00:36:07
AI in manufacturing is to figure out how
00:36:10
do you design and manufacture um
00:36:13
products next generation products right
00:36:14
here uh in in the US in a way that
00:36:18
mimics the cost advantages of of China
00:36:21
so that then our innovation can then
00:36:23
carry the data for us to be the global
00:36:24
leaders yet again in this next phase so
00:36:26
we have a company rebuild manufacturing
00:36:28
that's focusing on this for example
00:36:30
there's a lot of focus that needs to get
00:36:31
on on that because if it's self-driving
00:36:33
and it's not cost effective. Yeah, some
00:36:36
of us will buy it, but it's never going
00:36:37
to be a mainstream product because cost
00:36:39
has I mean there's a reserve price that
00:36:42
really shifts the demand patterns around
00:36:44
automotive and you probably have good
00:36:46
data on this as well. We should talk
00:36:47
about that. But we we got to get the AI
00:36:50
right and we got to get the
00:36:51
manufacturing cost right as well.
00:36:53
>> No, I think that there's a massive
00:36:55
coming down the cost curve on this. I'm
00:36:57
with you, Jason. I think we're going to
00:36:58
see literally over the next 12 to 24
00:37:01
months a massive transformation. I think
00:37:04
the race is a foot, right? The race is a
00:37:06
foot between a let's say a western stack
00:37:08
and a and a Chinese stack on this and
00:37:11
then in rest of world it'll be
00:37:13
interesting as a battleground to see
00:37:14
where that plays out. But and you and I
00:37:17
were were talking a little bit about
00:37:18
this. I think that is a massive trend.
00:37:21
I think a larger trend will be the trend
00:37:24
um to robotics and and not just for
00:37:27
human interaction but in manufacturing
00:37:29
and and when you think about the
00:37:32
challenges that um the western world
00:37:35
faces. So take the US. I was talking to
00:37:37
the CEO of one of the large contract
00:37:39
manufacturers and she has 50,000 job
00:37:42
openings right now for US manufacturing
00:37:44
jobs in America that she can't fill.
00:37:46
>> Right? And our demographics aren't
00:37:47
getting better on this front. Germany is
00:37:50
even worse situation. Korea,
00:37:52
>> Germany,
00:37:53
>> like another level.
00:37:54
>> Yeah.
00:37:54
>> And I think the only way that you build
00:37:57
resilient supply chains at the cost
00:37:59
point that you're talking about is it's
00:38:01
going to be robotics at the heart. And
00:38:03
and this race, I think, is wide open.
00:38:05
Korea leads the way in robots per per
00:38:08
worker, right? They're about 1 to 10
00:38:10
right now. Germany and China are tied at
00:38:13
second and the US then is a distant
00:38:15
third. And so there's a real race. You
00:38:18
talked about the autonomy thing. I would
00:38:19
actually jump to the robotics thing and
00:38:21
wonder how do I
00:38:22
>> one of the issues in robotics is so so
00:38:24
when you build the LLMs you could dump
00:38:26
them in the cloud experiment with
00:38:28
something called chat GPT and becomes
00:38:29
pervasive. If you have good robotics
00:38:31
models what's next? You don't have a
00:38:34
hardware um capability uh that's like an
00:38:38
API infrastructure that diffuses those
00:38:39
models fast. So like there's a lot that
00:38:41
needs to get built. So I actually think
00:38:42
robotics will be slower than people
00:38:44
think in terms of really taking hold.
00:38:46
But it's essential to go lead in that if
00:38:49
you're going to lead in manufacturing
00:38:50
and therefore have that core advantage
00:38:53
to play up the stack in industries like
00:38:54
automotive. There's no other way to do
00:38:56
it.
00:38:56
>> Yeah. I don't want to I don't want to
00:38:58
name drop but I went two weeks two
00:39:00
Sundays ago I went to Tesla with Elon
00:39:02
and I went and visited the Optimus lab.
00:39:06
There were a large number of people
00:39:09
working on a Sunday at 10 a.m.
00:39:10
>> Yeah.
00:39:11
>> And I saw Optimus 3. I can tell you now,
00:39:14
nobody will remember that Tesla ever
00:39:16
made a car. They will only remember the
00:39:18
Optimus and that he is going to make a
00:39:20
billion of those. And it is going to be
00:39:22
the most transformative technology
00:39:25
product ever made in the history of
00:39:27
humanity because what LLMs are going to
00:39:29
enable those products to do is
00:39:31
understand the world
00:39:34
and then do things in the world that we
00:39:36
don't want to do.
00:39:37
>> Yeah.
00:39:37
>> I I believe it'll be a onetoone ratio of
00:39:40
humans to optimists. And I think he's
00:39:41
already won, but I don't want to speak
00:39:43
out of school. But I do have a box. We
00:39:46
go to the box.
00:39:47
>> I have a box.
00:39:49
And these are all really interesting
00:39:51
technologies that we all got to see. How
00:39:53
many people owned one of these?
00:39:57
I mean,
00:40:00
Michael Douglas made this famous.
00:40:01
Remember Wall Street on the beach making
00:40:03
trades? And there was an amazing You
00:40:05
Will commercial. Remember the AT&T You
00:40:07
Will commercial? And this was one of
00:40:11
them. You'll be able to work remote from
00:40:12
the beach.
00:40:14
What is the equivalent of this today?
00:40:16
What is the equivalent of this today?
00:40:18
What do you think, you know, we're going
00:40:20
to look back on this year and laugh at
00:40:23
in 30 years? This is something from the
00:40:25
80s, so I guess this is 30 years ago.
00:40:28
What are we going to look at that we're
00:40:30
all enamored with today that we'll kind
00:40:33
of get a little gau out of?
00:40:34
>> Well, you know what? I'll I'll tell you,
00:40:36
by the way, I love it. It says
00:40:37
California mobile phone on this. That
00:40:39
was like the brand associated. And two
00:40:42
im two memories come to mind for me on
00:40:44
this. One was Envy
00:40:45
>> because when I started only the most
00:40:47
senior people could get one of these and
00:40:49
I couldn't, right? Like I want they're
00:40:52
just like when you grabbed like I want
00:40:53
one of those. $4 a minute. What was
00:40:55
three or$4 dollars a minute?
00:40:57
>> And some battery lasts about 30 minutes.
00:40:59
>> Exactly. Some new associate doesn't get
00:41:00
one of these.
00:41:01
>> Did you when did you have your first
00:41:02
mobile phone?
00:41:02
>> I'm too young for this.
00:41:04
>> Too young for that. You're such a liar.
00:41:06
You had the star tag like me. Had star
00:41:08
tag like me.
00:41:09
>> But the uh but the second and this was
00:41:10
made infamous was one of the great
00:41:12
things unfortunately great failures that
00:41:15
we had was we did a project and it was
00:41:17
it was published a while ago for AT&T in
00:41:19
the mid80s that said these things are
00:41:21
never going to take off.
00:41:22
>> Cell phones we're going to really get
00:41:23
going on the
00:41:24
>> I don't know why you burned me with this
00:41:25
one.
00:41:26
>> By the way, let me remind like something
00:41:27
today just to answer your question.
00:41:29
>> Think about like the a lot of the
00:41:30
eyeglass innovation that's happening.
00:41:32
This was with your ears. the innovation
00:41:33
we're trying to with the eyes on how to
00:41:35
intelligently navigate I think that
00:41:37
there's so many attempts that have not
00:41:38
worked in the last year or something.
00:41:39
There we go.
00:41:40
>> All right.
00:41:40
>> It's a really good segue because here's
00:41:42
the Google.
00:41:44
>> Now, as ridiculous as I look right now
00:41:46
and I can hear the cameras taking my
00:41:48
picture
00:41:49
>> and you will not be spared because
00:41:50
you'll be wearing them as well. Um I
00:41:52
remember when Larry and Sergey started
00:41:54
walking around with these.
00:41:56
>> In fact, Larry, uh I was at a party and
00:41:58
he came on the dance floor with these
00:41:59
and I said, "Larry, take those off. All
00:42:01
the girls are gonna stop dancing if you
00:42:03
keep walking around with them. He goes,
00:42:04
"Really?" I was like, "Yeah, that's not
00:42:06
how dancing works." You don't. But if
00:42:08
you think about this product, why did
00:42:09
they stop making this? They should have
00:42:11
kept iterating.
00:42:13
>> And this was AR before AR. You you see
00:42:16
right through it.
00:42:17
>> Yeah. Ahead of its time, right? And
00:42:20
>> go ahead and try it out, Bob. There you
00:42:21
go.
00:42:21
>> And now forever you will also be in
00:42:25
infamy.
00:42:26
>> There you go.
00:42:27
>> Your turn.
00:42:29
smart to do it. But
00:42:30
>> all right, I'll do it. I'll But but by
00:42:31
the way, the new ones aren't much
00:42:32
better. The form factor is better, but
00:42:34
the utility isn't there.
00:42:36
>> So when you look at it,
00:42:37
>> this this the today's version of this is
00:42:40
what this was, I
00:42:43
>> now here's one. This is a miniature
00:42:44
version. I tried to get this and if
00:42:46
anybody can get me this, I I'll pay
00:42:48
$10,000 for it. Uh maybe $25,000.
00:42:52
The Theronos one drop blood machine.
00:42:54
This was like one of their chachkis.
00:42:56
Ooh.
00:42:56
>> But in in truth, you're now in
00:42:59
healthcare.
00:43:01
This may have been a fraud allegedly.
00:43:04
In reality, she's in jail, I guess. So,
00:43:06
I don't want to I mean, maybe there's a
00:43:08
chance it was all she's innocent. Um,
00:43:11
who knows? Um, I'll leave that
00:43:14
possibility out there. Um, allegedly.
00:43:17
Um, but this the promise of this
00:43:20
captured people's imagination. A small
00:43:22
amount of blood to get a lot of data
00:43:24
back. And in fact, in fairness to
00:43:26
Elizabeth, she was able to do a couple
00:43:28
of interesting tests with a small
00:43:30
amount. This was a great product idea.
00:43:33
Correct.
00:43:33
>> Yes. Yes.
00:43:34
>> Will somebody create that with AI in the
00:43:37
next 10 years?
00:43:38
>> Uh I think it's very likely because the
00:43:40
the the challenge with this is how can
00:43:42
you actually manufacture those nano
00:43:44
devices where you can take really low
00:43:45
volumes and be accurate and measure
00:43:47
these things. Technology wasn't there.
00:43:48
So when you going back to our hardware
00:43:50
manufacturing innovations, I think they
00:43:52
will catch on to enable this and and you
00:43:54
want this you you want this to be that
00:43:57
uh you know you can have real-time
00:43:58
diagnostics think about a modern
00:44:00
physical and be much more preemptive
00:44:02
about healthcare like pervasive
00:44:04
effective capabilities like this these
00:44:06
endpoints will be useful for that
00:44:07
>> and you have function health you have
00:44:08
superpower now doing I don't know if you
00:44:11
guys use either of those products
00:44:12
>> but getting your blood work done every
00:44:14
year having you know a concierge talk to
00:44:17
you about it for butt 800 $100 a year,
00:44:19
$600 a year. Obviously, consumer-ledd
00:44:21
healthcare and the Theronos vision. I
00:44:23
>> I think there's a growing movement
00:44:24
around longevity. It's like become a
00:44:26
cultural phenomenon. And so that's first
00:44:29
of all, the fact that consumers have
00:44:30
propensity to pay. We have a company
00:44:32
called RO for example that focus on
00:44:33
GLP1s because there's that propensity,
00:44:36
it drives innovation to create more uh
00:44:38
products like this that are focused on
00:44:40
keeping you healthy.
00:44:41
>> How many people owned one of these?
00:44:43
Raise your hand. All right. And how many
00:44:45
of people have three of these in their
00:44:46
closet that they can't throw away?
00:44:49
>> I mean
00:44:49
>> the keyboard.
00:44:50
>> This was the greatest product ever.
00:44:52
>> So we we so I did a startup out of
00:44:54
college. One of the very first apps that
00:44:56
was non- email on this. We wrote that
00:44:58
and it was a merch merchandising app for
00:45:00
Red Bull. So they could actually do
00:45:02
inventory tracking in a store. And this
00:45:04
was like this is amazing product. You
00:45:06
know the
00:45:06
>> I'm I'm still faster on this keyboard,
00:45:09
>> right? I mean this was like for McKenzie
00:45:12
this was your cocaine.
00:45:14
>> We we had some this was and we had some
00:45:17
very senior people even when we might
00:45:18
that wouldn't give up.
00:45:20
>> I have to tell you a story. It just
00:45:21
gives me anxiety. I used to I I grew up
00:45:24
writing apps on this and then in 2011 I
00:45:27
moved to the valley and I had my
00:45:29
Blackberry. I put on a table like this.
00:45:31
I met with somebody who was a well-known
00:45:32
person in the valley. We had a good
00:45:34
conversation. At the end of it he said
00:45:36
you still use a Blackberry. I was like
00:45:38
yeah. He's like, "Stop doing that. You
00:45:40
were judged in this meeting."
00:45:42
>> I kid you not. Like, okay.
00:45:44
>> Well, I mean, just think about
00:45:45
>> I don't want to touch it. You were uh
00:45:47
you were holding it away from me.
00:45:49
>> Just think about how many carpal tunnel
00:45:51
surgeries this created.
00:45:53
>> Oh, absolutely.
00:45:53
>> I mean, this was great for the economy.
00:45:55
>> This is an interesting one. How many
00:45:56
people owned
00:45:58
>> Palm
00:45:58
>> a pilot? Yeah.
00:45:59
>> How many people own one of these?
00:46:01
>> Incredible, right? And it this one I
00:46:03
have a stylus antenna.
00:46:04
>> No, the stylus isn't here. We got this
00:46:06
off of eBay thanks to my friends at CES.
00:46:09
Um, but you got to learn script and you
00:46:11
would be very good at, you know,
00:46:12
spending at a party three or four
00:46:14
minutes typing in some
00:46:15
>> and you'd have to have your phone
00:46:16
separately, right? These are two
00:46:17
different devices. So,
00:46:18
>> and if you really wanted to be like have
00:46:21
a lot of swagger and a lot of RZ, you
00:46:23
would have this on one side of your
00:46:24
belt. I know you had this problem. You
00:46:26
did have, didn't you?
00:46:28
>> You have to be equal.
00:46:29
>> And the Blackberry
00:46:30
>> on the other side.
00:46:30
>> That was like you were like a
00:46:31
gunslinger.
00:46:32
>> Yeah. And then in the early days when
00:46:33
the Blackberry didn't have the phone,
00:46:34
then you had the phone, too. So then you
00:46:36
you look like a utility guy.
00:46:39
>> Hey Mont, I know that in college you
00:46:41
lost a lot of brain cells to this one.
00:46:44
>> The first ad on the internet
00:46:46
>> was a banner ad for Zema.
00:46:49
>> Oh boy.
00:46:50
>> How many people have had a Zema?
00:46:52
>> Oh, too many.
00:46:53
>> These are headaches.
00:46:54
>> This was the most repulsive drink in the
00:46:56
world. We got a empty can of it. Um,
00:46:59
it's still available, I think, in
00:47:01
Sweden. I think there's one place that
00:47:03
still has the license and produces this
00:47:05
horrific beverage.
00:47:06
>> Um,
00:47:07
>> but you know, you look at all the
00:47:09
carbonated stuff now, like
00:47:10
>> I mean, I think version of this
00:47:12
>> White Claw is there.
00:47:14
>> Yeah, I think that's that generation's
00:47:15
>> You want No thanks.
00:47:20
>> I I actually ran a marathon with one of
00:47:22
these on my waist in New York City. The
00:47:25
Sony Discman. It didn't skip when you
00:47:28
were running.
00:47:28
>> You see, this is a very good point. I
00:47:30
had the advanced one that had 10,
00:47:33
>> it had a 10-second buffer.
00:47:35
>> Oh,
00:47:36
>> this was elite at the time. It's an
00:47:38
extra 50 bucks, but it would buffer 10
00:47:40
seconds.
00:47:42
>> And then obviously the iPod came out.
00:47:44
>> What What do we think in terms of the
00:47:47
limited capabilities, but the
00:47:49
inspiration of this
00:47:51
will we look back on at this moment in
00:47:53
time? In other words, a device that
00:47:57
could go a thousandx in its capability,
00:48:00
but providing the same similar
00:48:02
functionality. This case, being able to
00:48:04
have portable music. That's interesting
00:48:06
because, you know, you think of the
00:48:08
Walkman before this, right, which was
00:48:10
the cassette,
00:48:10
>> right?
00:48:11
>> The wooden skip that was durable.
00:48:12
>> Durable.
00:48:13
>> Advance in technology and moving from
00:48:15
analog to digital, but less durable.
00:48:17
>> Yeah. But fidelity,
00:48:18
>> better fidelity. Transition to iPod,
00:48:21
whatever. Right. that then solve both of
00:48:23
the the equations. And you know, it gets
00:48:25
you think what what are the transition
00:48:27
technologies we're in right now? And one
00:48:29
of the places I come back to is is um
00:48:32
health wearables. Um so many different
00:48:35
health wearables out there and they're
00:48:36
all attacking the problem from slightly
00:48:38
different angles.
00:48:39
>> Yes,
00:48:40
>> some advances, but I think we're on the
00:48:41
cusp. I go back to marrying this plus
00:48:44
wearables to having more continuous
00:48:46
monitoring and data. We might be this
00:48:49
might be the transition step on uh
00:48:51
>> between your eight sleep, your aura,
00:48:53
your whoop, all of that, your blood work
00:48:56
coming together and giving you
00:48:57
customized medicine.
00:48:58
>> I think that's a better answer than I
00:49:00
was going to give, but my answer is the
00:49:02
LLM hallucinations
00:49:03
>> because when you think about the
00:49:05
intelligence, it's actually unreliable
00:49:07
in a lot of ways
00:49:08
>> just like the music was unreliable with
00:49:10
this and is that going to change
00:49:11
fundamentally over the next
00:49:14
one. This was a very interesting device
00:49:17
because for people who don't know, this
00:49:19
one might have text messaging on it, but
00:49:21
it used to just tell you the phone
00:49:22
number of the person who text called
00:49:24
back. So now if you were dating and you
00:49:26
were in the dating pool and you got that
00:49:28
text from that special number, you're
00:49:29
like, "Oh, how many minutes before I
00:49:31
call back? I got to go find a pay phone
00:49:34
and call back." But you used to be able
00:49:36
to give a number. So after you page
00:49:38
somebody, you could put in a couple of
00:49:40
digits code. So, we started to have our
00:49:42
own vernacular uh 411 or 911. You could
00:49:47
append to your beep some numbers like
00:49:51
maybe your location, etc., the street
00:49:54
number you were on, etc. Really an
00:49:55
interesting product in how
00:49:59
we never got to turn off work that
00:50:04
led to always on
00:50:07
>> doom scrolling. the neverending nature
00:50:09
of, you know, our commitment to work.
00:50:12
And in some ways now we're starting to
00:50:14
see a reverse of that. People are buying
00:50:16
phones. I I understand a lot of
00:50:18
millennials now are buying digital
00:50:20
cameras so they can leave their phone at
00:50:21
home and they're getting flip phones
00:50:24
>> so they've unbundled it. Really
00:50:25
interesting. Any memories of the uh
00:50:27
pager for you?
00:50:28
>> Yeah. Well, first of all, you know, they
00:50:29
always say all the money is made in
00:50:30
bundling and unbundling and that is
00:50:31
happening. And I think it is about if
00:50:34
you're going to say the equivalent of
00:50:35
this which is about how do we go back to
00:50:37
human connection uh and engaging in
00:50:40
person as opposed to trying to you know
00:50:41
be lonely online being fulfilled
00:50:44
offline. That's probably the behavioral
00:50:45
change that's going to happen. What what
00:50:48
enables that I think is probably there
00:50:49
is some social engineering that's going
00:50:50
to drive that.
00:50:51
>> This has been an amazing hour. Well done
00:50:53
gentlemen. Big round of applause for our
00:50:55
guest.
00:50:56
>> Thank you so much for hosting us. This
00:50:58
is incredible. Thank you. been a great
00:51:00
audience. Thank you for
