Today on the pod, we're bringing you two of the liveliest panels from the 2025 Cerebral Valley AI Summit, held this week in London.
Both panels — “The Autonomous Vehicle Rollout” and “Investing in 2030” — explore one of the major themes from the event: where AI is poised to show up next in our everyday lives, beyond the chatbot. Think voice, devices, and even your car.
First up, we'll hear from Uber CEO, Dara Khosrowshahi, and Alex Kendall, Co-founder and CEO of Wayve, who are teaming up to bring self-driving cars to the UK.
Then we turn to the investor perspective, with top European VCs — Philippe Botteri of Accel, Tom Hulme of Google Ventures, and Jan Hammer of Index Ventures — on where they see the biggest AI opportunities for founders in the years ahead.
00:00:00
This is brought to you by forethought building AI agents
00:00:03
for every customer moment. Hey, it's Madeline from the
00:00:07
newcomer podcast. We're fresh off the Serbo Valley
00:00:10
AI Summit in London, and we're going to bring you two of our
00:00:12
liveliest on stage discussions into the podcast feed.
00:00:16
Both touch on one of the biggest themes of the event, which was
00:00:18
where AI will enter our lives beyond the chatbot, whether it's
00:00:21
through voice on a device or even in our cars.
00:00:25
First up, we'll hear from Uber CEO Dara Khosrowshahi and Alex
00:00:29
Kendall from Wave, who are teaming up to bring self driving
00:00:31
cars to the UK. They actually rode in a Wave
00:00:34
self driving car to the summit, which was really cool to see.
00:00:37
The next panel is a conversation between some of the top VCs in
00:00:40
Europe, Philippe Bateri from Excel, Jan Hammer from Index
00:00:44
Ventures, and Tom Holm from Google Ventures about where
00:00:47
they're seeing the next big opportunities for AI founders
00:00:49
globally. Give it a listen.
00:00:56
Hey, well, we're down to the last two, two of the discussions
00:01:00
I'm most excited about. So excited to be here with Dara
00:01:04
and Alex. I, you know, covered all the
00:01:08
Uber periods. I remember riding in an Uber
00:01:11
self driving car in 2016, pre Dara when felt like there was a
00:01:16
lot of hype, a lot of interventions and then Uber had
00:01:23
to pull them off the road because I think they maybe went
00:01:25
a little too early. And now, you know, almost a
00:01:28
decade later, I think I've gone from being a self driving bear
00:01:33
to being really excited about the opportunity.
00:01:36
So I just want to start off. I mean your two companies are
00:01:39
are working together to bring self driving cars to London.
00:01:44
Dara, you want to start off why? Why work with Wave and what is
00:01:47
the partnership going to look like here?
00:01:48
Well, we're working with Wave because they're an absolute
00:01:51
leader in the field. Our strategy at at Uber is we
00:01:56
are no longer developing our own self self driving technology.
00:01:59
We want to really work with the ecosystem and look for the best
00:02:03
and brightest. We're developing self driving
00:02:04
technology. We've built a platform that
00:02:08
brings incredible demand on a global basis.
00:02:11
We can help these highly technical companies with
00:02:14
operations, local operations, etcetera and we can be a vehicle
00:02:20
to bring the self driving technology and introduce it to
00:02:23
consumers. And so we have a rare view into
00:02:30
all of the development that is happening in the industry and
00:02:35
Wave is unquestionably one of the leaders.
00:02:37
I think they're taking A and Alice can talk to it kind of a
00:02:41
bit of a differentiated approach in terms of an end to end large
00:02:47
model. And once we saw what they were
00:02:50
doing, once we experienced the the driving efficacy of of their
00:02:56
driver and how it can generalize.
00:02:58
Ashley, Alice and I were driven here by a wave driver from Wave
00:03:03
headquarters. With a safety.
00:03:05
Driver right with a safety driver.
00:03:07
Any interventions or 0. How was the drive?
00:03:10
Perfect, perfect. And, and there were some really,
00:03:12
really complex maneuvers that we had to go through construction
00:03:17
trucks parked on the left, so it had to go around trucks.
00:03:23
Pedestrians in London being pedestrians in London.
00:03:26
It was fairly. I love all all the look left,
00:03:29
look right on the ground. You really need the
00:03:31
instructions. But, but these guys have been a
00:03:33
real leader and so we, we very much wanted to work with them.
00:03:36
We've invested in the company and and we're very much looking
00:03:38
forward to bringing their technology to London and then
00:03:43
and then beyond. Maybe you can talk a little bit
00:03:44
about that. And.
00:03:45
I mean, regulators have said, oh, you could do it next year.
00:03:48
I'm sure you guys were supportive of that.
00:03:49
But are you prepared to do it next year?
00:03:52
How do you feel? How ready are you to deliver
00:03:56
self driving cars with Uber in London?
00:03:58
Well, look, I think the thing that's very exciting for us is
00:04:01
the confidence and clarity we have from regulators here.
00:04:04
So we're working as hard as we can to get this launched in a
00:04:07
way that scales of course in London, but really worldwide.
00:04:11
And I, I can't wait to do that on the, on the Uber network.
00:04:14
But the, the clarity and conviction that the UK
00:04:17
government has said, look, we want to make this possible.
00:04:19
We want to accelerate regulation by a bit over a year and make
00:04:23
this possible from next year. It's, it's, it's great for us
00:04:25
because we can say, OK, we can commit to building this out here
00:04:28
and it won't, you know, it won't be a full service all of a
00:04:32
sudden from day zero. There will be of course steps
00:04:34
that we take in a controlled and responsible way.
00:04:37
But ultimately. You know what neighborhood will
00:04:38
be first. Well, I, we've all got our
00:04:41
favourites, but I that, that, that debate's still ongoing.
00:04:45
But I think being able to what we enable with end to end AI is
00:04:49
of course to move away from the concept of high definition maps.
00:04:52
You don't need to map each region.
00:04:54
Actually, the AI has the kind of intelligence to understand
00:04:57
scenarios it's never seen before.
00:04:58
My favorite experience from our drive, we went through this
00:05:00
intersection where there was a, a car that stopped in front of
00:05:03
us and we had to navigate. I wait for an oncoming car and I
00:05:06
lift safely, change lanes to get around them and keep the traffic
00:05:09
flowing. And all of this kind of
00:05:11
complexity our AI can handle naturally.
00:05:13
And I think it's not just about having a a, you know, AI driver
00:05:18
that gets you from A to B, but keeps the traffic flowing in a
00:05:20
way that doesn't cause any road angst around you.
00:05:22
I think that's what's going to be really loved by London, as a
00:05:25
new experience that fits in with a London driving culture as well
00:05:28
as being of course. Are you using Lidar or what's
00:05:31
your philosophy in terms of sensors?
00:05:33
We build an AI foundation model that can address all ranges of
00:05:36
autonomy. So the car we drove in today was
00:05:39
camera only. Although the what?
00:05:41
Kind of car was it? It was a Ford.
00:05:43
Marquee. But the vehicles we target for
00:05:45
the Level 4 robotaxi applications, they're going to
00:05:49
be have redundant sensors. So camera, radar lighter.
00:05:51
The key thing though is that they're not going to have a
00:05:54
retrofitted spinning lighter on the roof, but integrated sensors
00:05:57
that the manufacturers can produce at mass volume.
00:06:00
And I think that's going to make this economically viable and
00:06:02
scalable. I think just one comment on
00:06:05
what's really cool about the UK and the regulators actually
00:06:07
accelerating the timeline here is there are very few times when
00:06:11
you have a true platform revolution the way that you're
00:06:15
seeing with AI. And I do think that from a
00:06:18
European standpoint, there could be the view that the US has been
00:06:24
the country that has benefited from, you know, the big platform
00:06:27
changes, Internet, broadband, mobile, cloud, etcetera.
00:06:32
It has been the US companies that have grabbed a significant
00:06:36
share of the enormous opportunity that has been
00:06:39
created as a result of these platform changes.
00:06:42
AII mean you've heard that this is something that that you're
00:06:45
deeply in represents such a platform change and the talent
00:06:51
level that we see in the UK coming out of the educational
00:06:55
institutions is incredible. And so if the UK has an
00:07:00
opportunity to be a leader here in terms of this change and kind
00:07:03
of the application of AI generally, physical AI.
00:07:06
Are we at a point yet where cities they should almost be
00:07:09
bidding for the opportunity or like, you know, in San
00:07:11
Francisco, it feels like tourists, tourists are trying to
00:07:14
go see way MO's. Like, do you do you get a sense
00:07:17
that the mood in cities has pivoted it all to, oh man, being
00:07:22
first is like a compelling pitch?
00:07:24
We're we're seeing cities leaned in now you should be leaned in
00:07:28
at the same time, safety is paramount.
00:07:32
The cost of a mistake in the real world is so much higher
00:07:36
than a hallucination in the digital world.
00:07:39
So while they are leaned in, these are long dialogues that
00:07:44
we're having with regulators to make sure that they you can be
00:07:48
leaned in but safe at the same time.
00:07:51
But so far the conversations that we're that we've had are
00:07:53
very, very constructive. But we have to meter how we how
00:07:58
we lean in here. I mean, what we see is that
00:07:59
city, this solves so many problems for cities, but it's of
00:08:02
course it comes down to the nuance of how you integrate and
00:08:05
deploy this technology. And of course, doing it in a
00:08:07
responsible way is. Critical.
00:08:09
How do you prove it to London? Or what's the path to showing
00:08:11
that you are safe? There's so many aspects.
00:08:15
To. Regulation.
00:08:17
And the key thing for us is of course we prove it before we
00:08:19
deploy. And so of course that starts
00:08:22
with very large scale offline testing and simulation, right?
00:08:25
It starts with proving that you can expose the AI to the right
00:08:29
amount of data across the right behaviors and then testing the
00:08:32
integration. I believe the Texas approach is
00:08:35
prove it after you deploy, we'll get more on that.
00:08:38
Dara, I mean you are partnering with other self driving car
00:08:42
providers, you're working with way MO like what what do you
00:08:46
think about sort of your partnership strategy and how do
00:08:49
you decide which partners to work with?
00:08:51
Well, ultimately we think this is a technology that holds the
00:08:55
promise of making the streets of the world safer.
00:08:58
That is by far the most important factor here in the US
00:09:01
alone, there are probably 35 automotive fatalities per year.
00:09:06
And if you take the opportunity that we have now, which is to
00:09:09
hold the standard of safety of these AI drivers being multiple
00:09:13
times better than humans, and Waymo certainly has demonstrated
00:09:17
that that's possible. You get to have safer streets
00:09:20
and over a longer period of time, you can bring the cost of
00:09:23
transportation down, increase the man and hopefully take a
00:09:26
chunk out of personal car ownership there.
00:09:30
So that's the that's what ultimately the opportunity to
00:09:34
represents. We don't think there's going to
00:09:36
be a winner take all in the marketplace.
00:09:38
And so we want to partner with the best and brightest.
00:09:43
We get to meet everybody in in the industry.
00:09:47
We have a safety case as well that we want to make sure is
00:09:51
fully satisfied. And Wave is just unquestionably
00:09:55
one of our top partners that we've met and we're just super,
00:09:59
super excited about building the relationship.
00:10:00
I mean you're in sort of a funny position in the like a lot of
00:10:04
people in this room and probably Alex all want the technology of
00:10:07
AI to really be the differentiator in sort of
00:10:10
pricing and you sort of want to be the demand aggregation and
00:10:14
sort of customer relationship is a differentiator.
00:10:17
Do you agree with that or what do you see as sort of your
00:10:22
leverage in in sort of deploying self driving?
00:10:26
I mean, listen, our, our leverage if you want to call
00:10:27
that is that we're operating in over 70 countries.
00:10:30
We and so we can bring instant demand to any technology
00:10:34
provider. And I think at least for the
00:10:37
next 5 to 10 years, there's going to be more demand for
00:10:41
these robot cars, if you want to call it that than supply.
00:10:46
So 10 years from now, maybe doubly competition, who's going
00:10:49
to win etcetera, and who's going to be the provider for various
00:10:51
OEMs. But we are a long way away from that.
00:10:55
And right now it's about making sure that we introduce this
00:10:58
technology safely in the streets of, you know, hopefully all the
00:11:03
countries in which we operate. So I just don't think we're
00:11:05
there yet at this point. Yeah, far away, Alex, the, you
00:11:10
know, there's a third mystery party in this partnership, which
00:11:13
is the OEM. I don't know that you're going
00:11:16
to give it up right here on stage, but like what
00:11:18
characteristics do you look for? Do you think it'll be just one?
00:11:22
And yeah, how do you see the roles of the three different
00:11:25
players? Yeah, That's a good question,
00:11:27
Eric. I mean, we work with automakers
00:11:29
around the world through Europe, Japan and the United States.
00:11:33
And I, I again, I think this is going to be a technology that
00:11:37
all OEMs will want to produce vehicles that have autonomy
00:11:40
capabilities. There will still be some
00:11:43
personal car ownership market and of course the ability to
00:11:46
drive hands off and eyes off with that is going to be a
00:11:48
really key thing for that product experience and that's
00:11:52
what our product can enable. But then the market is going to
00:11:54
move to focus on level 4 robo taxis and other autonomous
00:11:57
mobility applications. And So what we are seeing is the
00:12:00
very best automotive manufacturers are setting
00:12:02
themselves up for their future. They're working with our AI, of
00:12:05
course, building vehicles that are software defined that enable
00:12:08
you to the amazing thing now is that we're seeing manufacturers
00:12:11
bring out vehicles that you can get data off them.
00:12:13
You can we can start to aggregate data just like Uber
00:12:16
aggregates demand. We aggregate data around the
00:12:18
industry and that enables us to build a, you know, the most safe
00:12:22
and scaled AI across the market and in the vehicles, making sure
00:12:26
that there are features you need to operate a robotaxi platform.
00:12:29
One OEM per city or you think there will be multiple?
00:12:33
I think that consumers will want choice and we want to enable our
00:12:36
platform to enable you think. The cars can provide
00:12:39
differentiation too. Absolutely.
00:12:41
And I mean, if you look at our AI is flexible and that it can
00:12:44
work with different sensors, so it learns what a sensor
00:12:47
architecture can and can't see. But some OEMs may want to put a
00:12:49
camera here or slightly in a slightly different position,
00:12:52
maybe they want to use a different compute stack or of
00:12:54
course, the actual driving experience, the inputs to our AI
00:12:58
is not just the sensor input, the navigation prompt, but also
00:13:01
a style prompt. How do you want are?
00:13:04
You going to give these? Are you going to be the
00:13:05
scorekeeper of AI self driving? Like how much do you think
00:13:08
you'll? Well, we can measure all the
00:13:10
accidents. This one is slightly safer.
00:13:13
We, we do have a good viewpoint as to the efficacy of the
00:13:17
various AIS and the different approaches and there are
00:13:19
different approaches to solving the problem here.
00:13:23
And, and obviously we're going to make our bet on the
00:13:25
approaches that we think are safe.
00:13:28
And then also, obviously, like you said, there's a third party
00:13:31
in here, which is the OEMs. I do think 10 years from now, I think
00:13:36
every car sold is going to come with an L3 or L4 package
00:13:41
depending on what you're looking for.
00:13:44
I think it represents a huge profit opportunity for OEMs. I
00:13:48
mean, like the OEMs are understanding the beauty of
00:13:50
software 0 gross margin, obviously huge upfront cost in
00:13:53
terms of development. So you know, every OEM that we
00:13:58
talked to is very, very interested in this technology,
00:14:02
but at the same time it's very different tech that they that
00:14:05
they typically build. There's a big debate, do we
00:14:08
build it internally, do we go with a third party, etcetera.
00:14:11
But I think 10 years from now, every single new car sold, if it
00:14:14
doesn't have this package available, it's not going to
00:14:16
sell. And, and you think it'll you'll
00:14:19
be using sort of consumer cars as part of a Uber fleet?
00:14:23
Well, I think that. Or you think your cars will
00:14:25
drivers. Drivers, you know, they'll drive
00:14:28
our most just like popular features or cars is a Prius,
00:14:33
right? And so there'll be a driver
00:14:35
driving a Prius, whatever the next generation is, there'll be
00:14:38
a software driver, hopefully it will be a wave driver driving
00:14:42
that make of vehicle. What any thoughts on the Tesla
00:14:48
roll out in Austin and what it sort of, yeah, I mean you you
00:14:53
need to worry about I guess the overall reputation of self
00:14:56
driving and how much do you think sort of any one player
00:15:00
acting more or less recklessly will sort of blow back on the
00:15:03
rest of self driving? Yeah.
00:15:05
I mean, listen, I think that Tesla has been an unbelievable
00:15:08
innovator in EVs. They've been innovator and and
00:15:11
developing self driving software as well.
00:15:15
And and certainly I wouldn't call their start reckless at
00:15:17
all. I mean it's a it's a very small
00:15:19
operational domain. They've got safe drivers there.
00:15:24
Yeah. So I think they're, they're
00:15:25
being actually appropriately careful and, and you see the
00:15:29
challenges in, in real life driving there.
00:15:32
So at this point, again, like this is, it's a new industry.
00:15:38
We think that it's great to have so much excitement around it.
00:15:41
I think when people experience it, it kind of blows them away.
00:15:44
Then they start taking it for granted for like 2 minutes.
00:15:46
There's absolute delight. And then they're like, you know,
00:15:49
texting you on their phone. I mean, I trust it, Yeah.
00:15:51
I mean I've been mostly in San Francisco Waymo's but I I trust
00:15:55
them more than the average Uber drive.
00:15:57
Yeah, yeah. I'm curious as to what you
00:15:58
think. I mean, you, you, you're in it.
00:16:00
So I'm curious. Do you trust them more than
00:16:02
humans yet? I, I think, I think that's going
00:16:06
to be the future is that of course they are going to be in
00:16:09
so many ways, they're going to have more intelligence and be
00:16:12
safer than what what we can do when we drive cars.
00:16:14
But the, the operation today that we saw deployed in, in, in
00:16:17
Austin from, from Tesla. I, I think the key thing from
00:16:20
here is that this is going to grow.
00:16:22
It's a start and there's going to be iteration from here.
00:16:25
The critical thing is that as an industry, we need to make sure
00:16:27
that we iterate responsibly and have the right checks and
00:16:30
balances in until we prove that the level of safety can be
00:16:33
better than a safe and competent human driver.
00:16:37
Like talking about technology, like a core problem that I think
00:16:40
we've seen with AI is that like getting very close, getting to
00:16:44
like 95% feels awesome, but then it doesn't.
00:16:49
Like if it doesn't solve the case, then you can't get rid of
00:16:52
the humans and then it's not as valuable because they're human
00:16:54
drivers. I mean, well, one of the
00:16:56
interesting things there is to look at the difference between
00:16:58
cognitive and embodied AI and large language models.
00:17:01
Actually, how do you define safety is quite nebulous and
00:17:04
there's a lot of debate around what general purpose AI
00:17:06
regulation and standards should look like, but that's not true
00:17:09
in, in self driving. We have very clear industry
00:17:12
understanding of what safe is, of what expectations are.
00:17:14
And actually the way we evaluate and benchmark these systems is
00:17:17
very well understood. So that I, I, I think there is a
00:17:20
push in the, in AI, we're seeing more of a push towards safety
00:17:24
and benchmarks in embodied AI. And I think we're seeing
00:17:26
innovation about how do we understand and prove the levels
00:17:29
of these systems come out through robotics that may
00:17:33
advance us as a field here. But I feel good about this in
00:17:36
self driving and I I still think that conversation is evolving in
00:17:39
the in the language model space. But you know, I, I'm interested
00:17:42
in this beyond self driving, almost the lessons from self
00:17:45
driving to some of the pieces of AI where it feels like we're
00:17:48
getting answers, but then we're also getting hallucinations.
00:17:51
I mean in the Uber case, I mean you came in, you did spin out
00:17:55
Aurora, like how did you? We we merge ATG into Aurora.
00:17:59
Yeah. And then then became its own
00:18:00
company. How did you decide, I guess that
00:18:03
self driving wasn't there in 2017 or they this shouldn't be
00:18:07
something to invest in or how did you go into sort of all your
00:18:10
technologists saying, oh, we're just tomorrow, just tomorrow and
00:18:14
say I'm not convinced. Listen, I I think Eric was a
00:18:18
really tough decision at the time.
00:18:19
And one is we were deeply marked by the fact that we had an
00:18:23
accident, lost life there. And so the responsibility that
00:18:29
you hold in your hands when you're operating in the real
00:18:33
world hit us deeply. I think the other circumstance
00:18:38
that affected the ultimate judgement was that we, we ran
00:18:43
ATG very separately from mainline Uber because we wanted
00:18:47
the opportunity to partner with the rest of the ecosystem.
00:18:49
Again, like we, we don't just work with, you know, we don't
00:18:53
just have a Prius is on our platform.
00:18:57
We got 4th. We want to work with the entire
00:18:59
ecosystem. And when I went and talked with
00:19:02
some of the other AV developers, I'd say, you know, ATG, it's in
00:19:06
the family, but it's separate. We're going to treat you fairly.
00:19:08
We're not going to share data, etc.
00:19:10
You know, when you're having a conversation actually, like I
00:19:12
usually do with you and I'm saying stuff and you don't
00:19:14
believe a fucking can I say, pardon me?
00:19:17
That's that was the response. Like the we were having
00:19:20
conversations, but I think people didn't trust us.
00:19:23
Yeah, because they also viewed us, rightly so, as a competitor,
00:19:26
right? So we had a decision to make.
00:19:30
Do we go with a proprietary kind of vertical strategy or do we
00:19:35
truly partner with the industry? We couldn't have the best of
00:19:39
both worlds. And ultimately we've started
00:19:41
that. A partial model was the right
00:19:43
model going forward and based on, you know, basically all of
00:19:47
the leading players sans Tesla, but everyone else is open
00:19:52
working with us, etcetera. I think it was the right
00:19:53
decision to make at the time. But this was not a simple thing.
00:19:57
It wasn't on a PowerPoint. It wasn't clear which way to go.
00:20:01
But it was more it was more about the ability to partner
00:20:03
with other people than an assessment of the technology.
00:20:06
I think it was for for us it wasn't necessarily an assessment
00:20:09
of technology, but it was an assessment as to what we are
00:20:12
great at. And Uber is a very strong
00:20:16
software company, but we're not necessarily we weren't great at
00:20:19
developing self driving. We were good, but I think it
00:20:23
takes better than good to get there.
00:20:26
Alex, I mean, this is not monogamous partnership.
00:20:30
I mean he's talking about partnering with a lot of
00:20:31
different people. How do you see the partnership
00:20:33
strategy and how do you get comfort partnering with a
00:20:36
company that's like, oh, we want to partner with everybody?
00:20:40
Well, in a similar way, the the, the platform that we're building
00:20:42
is going to enable fleets and automotive manufacturers around
00:20:45
the world to build autonomous products.
00:20:47
But what we're seeing from Uber is leadership and speed that
00:20:49
really outpaces other other fleets and marketplaces.
00:20:54
Actually, it's not just the demand.
00:20:56
You think about all the aspects of running on autonomy service,
00:20:59
how you own, operate, clean, collect the data to integrate
00:21:01
it. There's so much work there to
00:21:03
take a technology works and actually scale it.
00:21:06
Now like Dara, I'm a huge believer in focus and I know
00:21:11
that our expertise is in building the AI model.
00:21:13
We don't want to build the car. We don't want to own and operate
00:21:15
it. We want to enable this AI to be
00:21:17
on as many robots as possible. And and you know, that's going
00:21:20
to go beyond cars in the future because I'm seeing things in
00:21:24
manufacturing or other mobility spaces that are a bit like self
00:21:27
driving back in 2017. And you know, it'll take some
00:21:29
time before manufacturers figure out the right software to find
00:21:32
vehicle equivalent. Then when they do, there's going
00:21:35
to be some manufacturers like Tesla is a great example of this
00:21:38
today that can build their own own vertical stack.
00:21:41
But then there's going to be a wealth of, of manufacturers and
00:21:45
operators that have the platform, that have the demand,
00:21:47
that have the product expertise, but lack the intelligence and
00:21:50
the scale that we can build there as a platform by
00:21:53
aggregating data and training a model that's larger than than
00:21:57
anyone can do on their own is going to unlock, I think the
00:21:59
broader autonomy ecosystem. So we see it in the same way.
00:22:02
But I'm, I'm so excited about, I mean, Dara's leadership and the
00:22:06
speed that Hoover is pursuing here, I think we can move
00:22:09
really, really quickly. Uber's worth 190 billion, which
00:22:13
is great, right? I think like an all time high
00:22:16
but but Tesla you know is worth a trillion you.
00:22:20
Have to spoil the. Party, don't you?
00:22:21
The do you think? I mean, how much do you think
00:22:23
you're getting credit for your work in self driving and how
00:22:25
much do you think the market should be saying Uber is a self
00:22:30
driving car company? I think we're getting zero
00:22:33
credit at this point. But but listen that when you're
00:22:36
developing new technologies as a company, you have to be willing
00:22:39
to invest aggressively for many, many years before the market
00:22:47
understands or necessarily gives credit for what you're doing.
00:22:49
So you know the market is bet against this before 190 billion
00:22:54
isn't bad. But I think if the reality that
00:22:58
we see in partnership with Wave and others and self driving
00:23:02
becomes real, not just in passenger mobility but other
00:23:07
kinds of mobility delivery, I think 190 will just be a step
00:23:12
sum for us. Alex, you know there's a risk
00:23:16
that way MO becomes synonymous with self driving cars or a lot
00:23:20
of people technologists are getting the way MO experience
00:23:23
first. Like where do you see your
00:23:26
differentiation with them? Like how quickly do you need to
00:23:29
move to sort of stay in the conversation as they keep to
00:23:32
deploy and how do you assess what they have on the market
00:23:35
today? Well, here's the interesting
00:23:36
thing. Going city by city with a
00:23:38
geofenced approach that relies on high definition maps this
00:23:41
this AV. 1.0 strategy that that the industry looks at is AV. 1.0
00:23:45
I like is. Is 11 play?
00:23:47
But actually, I think this technology is going to see mass
00:23:51
scale through personally owned vehicles first with driver
00:23:54
assistance. And that's how most people are
00:23:56
going to get to experience it. And I'm not clear if you can
00:23:59
actually build a general purpose robotaxi without that data, that
00:24:02
manufacturing integration and having our AI in millions of
00:24:06
vehicles around the world for hands off and eyes off driving
00:24:09
is going to provide the regulatory relationships, the
00:24:11
manufacturing integration, the data, and of course, the the
00:24:14
brand and the exposure around the world that we'll be able to
00:24:16
take and use that to build a general purpose AI for a general
00:24:21
purpose robotaxi. And so I think that is an
00:24:23
enormous advantage that we can actually get that scale 1st.
00:24:27
And you know, we're, we're in it for the long game.
00:24:29
We want to produce something that if you think about what the
00:24:32
future of robotics should be, it's not a, a robot that follows
00:24:35
around infrastructure and affluent places, but we want to
00:24:38
build a future where there's a level of intelligence that can
00:24:40
drive in new places. It can, you know, respond to
00:24:43
your demands, you can delegate tasks to it.
00:24:45
It can interact with you through language.
00:24:47
I mean, fundamentally, the way that's going to happen is with
00:24:49
end to end deep learning. And that's the that's the bit
00:24:51
we're taking. And I think that of course time
00:24:55
will tell, but I think that just like we've seen in many other
00:24:57
verticals from drug discovery, game playing, agents, language
00:25:01
models, robotics is going to be no different.
00:25:04
And I think we're going to see that better lesson play out.
00:25:07
So I'm all for end to end learning being the approach that
00:25:09
does scale by leveraging the scale deployment and consumer
00:25:13
vehicles. Dara, do you have a year where
00:25:15
you think the Uber platform flips from human majority to
00:25:19
self driving majority cars? I I am 100% sure that five years
00:25:23
from now we're going to need more humans on our platform than
00:25:25
we do today. Business is growing really,
00:25:27
really fast. 10 years from now, I'm not sure about that slope of
00:25:32
the curve. It's not going to flip, but the
00:25:34
increase in drivers and couriers that we need on the platform may
00:25:37
start metering out and after that, who knows?
00:25:40
But you're still saying 10 years away I.
00:25:42
Mean I'm saying 10 years away, we will still have more drivers
00:25:47
and couriers, but a higher and higher percentage of our
00:25:50
platform is going to be served by by robots, one where they're
00:25:55
delivery robots, drivers, etcetera.
00:25:58
It's just the demand for all things on demand, whether it's a
00:26:03
ride or a food delivery, grocery retail delivery, the demand is
00:26:08
just growing so quickly and I think that'll continue.
00:26:11
Do you do you think deployed self driving cars today are
00:26:14
safer than humans? Like is is that?
00:26:16
A Well, I think Waymo is demonstrating as we speak that
00:26:20
it can not only be safer than a human, but it can be multiple
00:26:22
times safer than a human. And I think we have an
00:26:25
opportunity to hold a safety bar that is very, very high and I
00:26:30
think we should take that opportunity.
00:26:32
Great, Darren. Alex, thank you very much.
00:26:34
Thank you. Appreciate it.
00:26:39
All right, back with some of the biggest venture heavy hitters
00:26:45
around, so super excited to have you guys all here.
00:26:50
Yeah, obviously sort of a lot going on.
00:26:54
This is really our look at the investing landscape.
00:26:58
I just wanted to start off, what is a domain right now, an area,
00:27:04
it probably an artificial intelligence doesn't have to be
00:27:08
that you're really excited about, that you're digging into
00:27:10
just like one lane that you're sort of excited about at the
00:27:14
moment and you're right here. So do you want to go?
00:27:17
Sure. Happy to kick off.
00:27:19
I guess we've travelled a huge distance from the launch of
00:27:22
ChatGPT building all the infrastructure and then the move
00:27:27
to applications. I mean it's, it's been year and
00:27:29
a half or so and specifically for Europe, what I'm what I'm
00:27:33
excited now that we are at this application level is actually
00:27:38
the traditional industries are getting changed,
00:27:40
pharmaceuticals, banking and finance, lending industries like
00:27:46
manufacturing, automotive, defense and so on.
00:27:49
I think those industries have historically been slower to
00:27:53
adopt any new technology, but now it's become really a
00:27:57
necessity for them to keep up, certainly on a on a global
00:28:00
scale. Applications in old school
00:28:02
businesses. Are the applications replacing
00:28:05
those businesses or selling to them, or both?
00:28:09
AI being used as an enabler. So take a pharmaceutical
00:28:13
example, which I guess Tom, you know a ton about protein
00:28:17
synthesis. You could spend $100 million
00:28:20
developing a new protein variant.
00:28:24
You could be experimenting manually, you could be in labs,
00:28:27
or you could use AI in companies like Cradle and and and and and
00:28:31
others in the industry to shortcut that that cycle, that
00:28:35
development cycle and reduce your cost by an order of
00:28:38
magnitude. Love the answer.
00:28:41
So look, the global economy is about $100 trillion and it's
00:28:44
just not fully digitised yet. AI is going to unlock that low
00:28:48
friction. So we've been asking like what
00:28:50
applications are good enough today?
00:28:53
So we started with law, invested in Harvey Law Hive coding as
00:28:58
well, which these guys have phenomenal investments in as
00:29:01
well. I'd give you a slightly
00:29:03
different answer as well, which is I think every time you have a
00:29:06
new sort of technology, there's a new interface that arises to
00:29:09
sort of leverage it fully. So with the personal computer,
00:29:12
you had the GUI, with the smartphone, you had touchscreen.
00:29:15
And I think this is clearly going to be an era of voice,
00:29:17
which is really interesting. So we've been investing in that
00:29:20
space where we think voice is really high throughput.
00:29:24
I look at my kids, they send voice notes on WhatsApp all the
00:29:27
time because it's really quick and really efficient.
00:29:29
Some of our investments in that would be really high bandwidth
00:29:32
like Neuralink at the speed of thought.
00:29:35
We have a stealth company in Israel that's doing sort of pre
00:29:37
vocalization or silent speech which is interesting.
00:29:41
And then actually investment in nothing here in London is
00:29:44
related as well. I.
00:29:45
Think a lot of this is voice without voice or sort of voice
00:29:48
is too slow. We need to be faster than that.
00:29:51
Yeah, absolutely. It's voice.
00:29:52
It's, it's actually an incredibly high throughput way
00:29:56
of communicating information. And so, you know, it just so
00:29:59
happens that it's also probably the optimum as you've seen with
00:30:03
all of the chat interfaces for LLMS for the time being.
00:30:06
So we're really bullish on that space.
00:30:08
Felipe. Yeah.
00:30:10
I mean, I, I, I call everything that has been said here, I think
00:30:13
it was very interesting with the AI to see how three years ago it
00:30:17
was all about the infrastructure and the model and now it's about
00:30:21
the applications. And I think in the application,
00:30:24
one area which I think is very exciting right now that we
00:30:27
haven't mentioned is cyber security and AI is really
00:30:31
transforming cyber. Creating the problems and
00:30:33
solving the problem. Exactly.
00:30:35
And you have companies like Sierra who are really focusing
00:30:37
on data and AI security, which really doing very well because
00:30:42
that's the number one problem for CISO.
00:30:44
It says, well, data is growing, data is even more valuable.
00:30:48
And now it's really, it's really at risk.
00:30:51
And I think after that application phase, I think
00:30:53
what's going to come I think in the next 1224 months is like
00:30:57
more the automation side and how do we start to automate task and
00:31:02
workflow in the core enterprise. And, and right now I think we're
00:31:07
seeing enterprises are dabbling with it and you have some small,
00:31:10
smaller project and Pocs and I think the next 12, you know,
00:31:13
1224 months, I would expect much, you know, bigger chunks of
00:31:18
tasks and workflow to be automated.
00:31:20
I think some same way that, you know, RPA IN281718 restarted as
00:31:25
a big way, we say, well, it's RPA still remain the best way to
00:31:29
automate repetitive tasks, but there needs to be the same.
00:31:32
But if they are the same, these are the cheapest, most efficient
00:31:36
ways to do it. But then there are other tasks
00:31:38
which are more complex where the outcome may not be always the
00:31:41
same that that requires AI. What what is what's the mood on,
00:31:46
I guess how much software is screwed right?
00:31:48
You see what we had the Klarna CEO being like, oh, we're going
00:31:51
to be able to build everything ourselves.
00:31:53
I mean, I hear even in cyber, you know, there's almost an
00:31:56
argument like we need to be so fast.
00:31:57
I think I forgot if it was abnormal security I was talking
00:32:00
to a couple months ago. It was like, oh, we need to move
00:32:02
so quickly because people are going to be able to spin stuff
00:32:05
up. I don't know just the general
00:32:07
whoever wants to jump on it, but like the view on whether you can
00:32:11
invest in software companies or your your own companies are
00:32:14
going to build the software themselves.
00:32:15
Yeah, You know, AI is here to solve a lot of these problems
00:32:19
we're discussing. But back to cyber, it's also
00:32:22
generating problems. Tom and I are fortunate to work
00:32:26
with a company called Resistant AI and actually it's recently
00:32:30
seen a massive take off because the, the surface area of fakes,
00:32:35
of fraud, of falsified documents, of falsified
00:32:38
identities has has expanded exponentially.
00:32:41
And so at some point, you know, I recall sort of the what what I
00:32:46
learned in payments a decade ago, it was it's just a cost of
00:32:49
doing business. There's a certain amount of
00:32:51
fraud that's always going to be there across credit card
00:32:54
networks. We want to optimize that the
00:32:57
total net revenue at this point, it's moved to a different
00:33:01
dimension. So the problem has grown
00:33:03
exponentially. And with that I think you need.
00:33:06
There's some more intelligent. Solution so that doesn't.
00:33:08
There's always an opportunity. Right.
00:33:10
Yeah, relative thought. I think you get phases with the
00:33:14
rollout of these technologies and at the moment much of their
00:33:17
code is quite superficial. It's almost creating an
00:33:21
explosion of technical debt. I think at the moment like you
00:33:24
see amazing examples. Like they need to buy a lot of
00:33:26
things are creating a mess inside their companies.
00:33:28
And do you? I don't believe that software
00:33:30
will be completely kind of compostable.
00:33:35
It's got to stay composable. It's got to stay so that
00:33:38
actually the foundations would be built.
00:33:39
So the next Phase I think will be people addressing the fact
00:33:42
that when you do Vibe Coder project or you do work with
00:33:46
cursor and it cross references libraries or open source,
00:33:49
they've got to be ones that we really believe in.
00:33:51
So all three of us have invested in a company here in London
00:33:54
called Tessel Guy, Pajani's new company, where he's addressing
00:33:58
exactly that. Like these projects have to be
00:34:01
reliable, they have to be built for the long term, and I think
00:34:03
that's the next phase that will fall through.
00:34:06
I think this is a real, this is a question, right?
00:34:09
10 thousand $10 billion question, which is like you look
00:34:12
at work day, for example, and can you say, well, it's
00:34:16
tomorrow, can someone do work day?
00:34:18
Or Can you imagine even a work day where basically, which is a
00:34:22
very simple software, but every user will say instead of
00:34:25
configuring a complex solution, we'll say, oh, I need to
00:34:28
integrate with this, you know, pull the data from that, show me
00:34:31
this report and dynamically the app will create the code for
00:34:35
that potentially. That's the vision of the future.
00:34:39
I think we're very, very long away.
00:34:40
From that, I want, I don't want to spend a lot of time in this
00:34:43
conference being like, why London?
00:34:44
Like I, I live in New York, I host events in San Francisco.
00:34:47
We're in London today. Like I think the pandemic made
00:34:50
everything global and it was easy to build businesses all
00:34:52
over the world. But I, I do the stats that I
00:34:55
will ask the sort of why London or what the question really is.
00:34:59
Like what you think the opportunity specific to London
00:35:01
is and where founders should really be here.
00:35:03
And then I guess the American hubris, like when should they
00:35:06
really be racing to America versus focusing and local
00:35:09
markets? Yeah, you would probably expect
00:35:12
we would say this. It's all about talent.
00:35:14
London has been the hub for not just London born and bred
00:35:18
talent, but also sort of the European capital.
00:35:22
Sorry to all all all all those from Amsterdam, Paris and
00:35:26
Stockholm, but it is a bridge across the Atlantic and.
00:35:32
How many people here are based in London?
00:35:35
Raise your hand. Oh, there we go.
00:35:37
And how many are not based in London?
00:35:40
Minority. So there you go.
00:35:42
London has a lot of talent, but also with that comes the
00:35:47
ambition and and things as has been mentioned has been
00:35:51
accelerated. So the speed with which
00:35:54
companies are funded, speed with which companies launch and the
00:35:59
speed with which they have to get to market in some in in in
00:36:03
in some sense, it is that speed of iteration that is the Moat.
00:36:07
Certainly in the early experimental cases where the
00:36:11
customer has five competing solutions and each of those five
00:36:14
providers is 1 year old, right. So you have to have speed of
00:36:17
iteration, you have to raise capital fast.
00:36:19
You don't have the luxury of time.
00:36:22
And in some sense, that forces those companies to look big from
00:36:26
the get go and, you know, get get on that bridge to you're.
00:36:30
Saying if AI in particular is a race, you need to go global
00:36:33
pretty fast. Tom, I want to ask you
00:36:35
specifically. I mean, you're, you know, part
00:36:36
of the Alphabet empire empire. Is, is DeepMind like a good
00:36:42
source of startups or like has it underperformed, I guess the
00:36:46
level of talent and sophistication in startup
00:36:49
creation or what's what's your view on the track record in
00:36:52
terms of DeepMind and spinning off startups?
00:36:56
I would say the jury is still out.
00:36:57
These companies are moving quickly, but the great news is
00:37:00
there are world class scientists, researchers and
00:37:05
applied engineers at Deep Mind that are just learning how to
00:37:09
build at the very sort of Pareto front of these models and
00:37:13
they're still most of them in Deep Mind and enjoying it.
00:37:16
Although there's a lot of movement between all these
00:37:17
companies for each of those, I think you get this kind of
00:37:20
multiplier effect when they leave.
00:37:22
So I don't think we've seen any of the sort of really big
00:37:25
companies that we would expect come out yet.
00:37:27
Maybe the one that we're most excited about is Isomorphic,
00:37:30
which is the protein folding business that we span out not so
00:37:33
long ago. But I went out with the leader
00:37:35
of DeepMind. Or a fascinating situation.
00:37:38
No, no, Demis. Well, I mean, Demis is one of
00:37:40
the smartest people I've ever met.
00:37:42
He works across both of those businesses and I'm confident
00:37:46
that clock speed won't be kind of compromised at all.
00:37:48
That's amazing. But the back to Yan's point, if
00:37:51
you look at London, I don't, you know, I think building
00:37:54
businesses in Europe is building them in hard mode versus the US
00:37:58
But London it's less hard because you know, there's great
00:38:01
capital here. The founders will start
00:38:03
businesses everywhere, but for every single great founder you
00:38:06
probably need 5 or 10 world class operators and there's more
00:38:09
of them in London than any other city in Europe at the moment.
00:38:12
Felipe, anything you'd add there, maybe either like
00:38:14
categories you think in particular there's an advantage
00:38:17
here or anything else and. Obviously, London has a
00:38:20
privileged place in Europe, but I would say it's not the only
00:38:23
ecosystem in Europe with very strong AI talents, right?
00:38:27
Look at Paris. I mean, this is where found it
00:38:31
fair with Yan, Lukun, etc. I mean, there's very, very great
00:38:36
depth of of talent. I'm not saying this because I'm
00:38:39
as you probably get, but you look at Germany, we thought
00:38:42
about. Doing this in Paris, but it just
00:38:43
seemed like too, too hard, like it was enough.
00:38:46
Different in the continent, different language seemed
00:38:49
challenging. But you look at the the LMU in
00:38:52
Munich. I mean this is a lab where
00:38:54
Stable Diffusion was invented. I mean, Stable Diffusion is at
00:38:57
the base of pretty much all the image AI model and that's the
00:39:01
discovery from, you know, from Germany.
00:39:04
So I, I think Europe has a very, very strong voice, I think to
00:39:09
play a lot to say in the AI world.
00:39:12
And you are, are you betting just like quickly on local
00:39:16
heroes or like are, you know, especially in the foundation
00:39:18
model game, there's like the bet of, OK, the government basically
00:39:22
is going to want some hometown hero and therefore there's an
00:39:25
opportunity or is that not big enough for the most part?
00:39:27
Well, I mean, I think there's defense, which is a very
00:39:30
separate sector where I think there are some very European
00:39:33
centric dynamics, which I think are, you know, are playing out
00:39:36
and where the European market is big enough to create like big
00:39:40
champion and a lot of value. I, I think if you look at Europe
00:39:44
today, yes, I mean, you have foundational model, but if you
00:39:47
look at what's being done in the rest of the world with, with,
00:39:50
you know, 10s, hundreds of billions of dollar invested,
00:39:53
we're not seeing this in Europe, but what we're seeing in Europe
00:39:56
is great companies that are building on this model.
00:39:59
I mean, you look at Synthesia, I mean Synthesia AI avatar, they
00:40:02
are the leader globally. No one has done this.
00:40:05
You have, you know, 11 labs, You have lovables like so Europe is
00:40:09
capable of generating real, like global leaders.
00:40:13
Yeah. And I think it's just the
00:40:14
beginning of the cycle. And Philip, I would, I would
00:40:16
actually include mistrial in that category and to to add to
00:40:19
the French. List to be.
00:40:21
Respectful that there is hard evidence that the regulated
00:40:25
industries they play into specifically two of those
00:40:28
banking and healthcare are siloed.
00:40:33
And as much as we would love to sort of cheerlead those
00:40:35
industries do have government regulation and they do have
00:40:38
governments overseeing and and especially as as your AI moves
00:40:43
more to the frontline. So in the case of Alan moving
00:40:46
from back office to sort of processing tasks to all the way
00:40:51
doing medical triage where the sensitive sort of HIPAA
00:40:55
equivalent sort of patient data, you really have the regulator's
00:41:00
breathing down their necks and saying this has to stay in
00:41:02
Europe. For venture to work, we need
00:41:06
exits. We've obviously got this sort of
00:41:09
very enticing exit with scale in the news and clearly there are a
00:41:15
lot of high salaries also being thrown around.
00:41:18
But it feels, you know in venture overall there haven't
00:41:21
there's never enough. We'd always like more of them.
00:41:25
Yeah. I mean, how much can you I guess
00:41:27
invest in companies on the back of these licensing Aqua higher
00:41:30
deals? And then how, what do you see as
00:41:33
sort of the best exit opportunities right now?
00:41:36
And if if you're not conflicted, you can touch on the scale thing
00:41:40
that some of you are, but I'm just throwing this out.
00:41:42
I'll take a 15. Billion dollar whatever he likes
00:41:44
and say, however. You call it deal any day.
00:41:47
Is that well, is the 15 billion? That's really the money and the
00:41:50
rest is like, maybe? I mean the rest is a stand alone
00:41:53
company that's going to continue to operate with the CEO,
00:41:57
customers etcetera. And then?
00:42:00
Tom, how much is it worth as a stand alone?
00:42:03
A lot less than 15. These guys invested in it.
00:42:07
It was a phenomenal and is a phenomenal investment.
00:42:10
And I think we've seen a new pattern of a sort of glorified
00:42:13
Aqua hire. So to me, this is a sort of $15
00:42:17
billion Aqua hire. We're seeing a lot of the
00:42:19
customers move away for obvious reasons, like snorkel in our
00:42:23
portfolios getting a lot of inbound because of it.
00:42:26
But I actually respect Meta for this move.
00:42:29
It's unbelievably, it's a critical time.
00:42:31
You're as good or as bad as the team.
00:42:33
They lost their head of AI, was it 3 weeks ago?
00:42:36
And they've now brought in a world class operator that
00:42:38
understands the secrets of, you know, LLMS and Gen.
00:42:42
AI in Alexander Wang. And $15 billion is a lot of
00:42:46
money. But what is it?
00:42:47
It's 1% dilution or something. I would take that bet if I was
00:42:50
that team. So I don't know how well the
00:42:53
business will do on an ongoing basis.
00:42:55
It might look a bit like Inflection did when they sold in
00:42:58
a similar deal. Didn't hear so much from them
00:43:00
afterwards. But phenomenal investment for
00:43:03
these guys, $15 billion. You know, if you just wind back
00:43:06
the clock five years ago and say $15 billion exit, it would have
00:43:11
absolutely blown our minds. And now we're talking about it.
00:43:14
We've just had a kind of recency bias of an open AI round where
00:43:17
it doesn't sound like a lot of money.
00:43:19
It was a phenomenal deal. Eric, let's not shop the future.
00:43:23
Time will tell. Yeah.
00:43:24
Time will tell. I love your if you're trying to
00:43:26
create a rule, are we going to have more of those?
00:43:28
We'll, we'll, we'll, we'll take your, your optimism, but you
00:43:33
know, to more than agree with your, with the premise of the
00:43:35
question. The economy does need some exits
00:43:40
and the ecosystem needs some exits.
00:43:42
And I mean, to take it even further, some sort of recycling
00:43:46
of capital and moving on forward is the essence of realizing
00:43:50
ambition for entrepreneurs and sort of, you know, putting the
00:43:54
capital and talent back into into the startup ecosystem.
00:43:57
So are we working hard to move our companies along?
00:44:01
Yes, we are. But obviously a lot of the exit
00:44:06
activity has has historically been on public markets.
00:44:09
It's been challenging looking looking backwards.
00:44:13
Let's hope it's more positive looking.
00:44:15
Forward, Yeah. What's do you want US regulators
00:44:17
to be chiller here or? Yeah.
00:44:20
Is it, is it a regulatory problem with IP OS or you think
00:44:23
it's a company quality problem? Well, it's it's been not too.
00:44:27
We don't have the time to sort of analyze.
00:44:29
That's a big question. But in in two sentences, the the
00:44:32
bar has gone up in terms of marketing because of float
00:44:36
research coverage, what's economical for the underwriters
00:44:39
to bring to market. So in some sense the the pool of
00:44:44
attention of both human time analysts by side as well as
00:44:50
capital has shrunk and and as a result the bar has risen.
00:44:54
And you know the the flip side of that is the activity in
00:44:58
private markets. Yeah.
00:44:59
Well, I think I would add the geopolitical uncertainty, which
00:45:02
is a huge thing, right. I mean, you need markets, you
00:45:05
need low volatility to go public and and right now I don't know
00:45:09
what the volatility is because we have been living in high
00:45:11
volatility for a very long time. Right.
00:45:13
What about Felipe, like private to private and specifically like
00:45:17
open AI? I mean, it's so valuable, but
00:45:19
you're getting these weird shares like I don't know, do you
00:45:22
think open AI can be part of the solution here and the and
00:45:24
private to private deals? Databricks has certainly done a
00:45:27
number of deals. I mean, I think private to
00:45:28
private are one part of the the solution.
00:45:33
But I do think we're going to, I think we need the public markets
00:45:37
and the geopolitical environment to stabilize.
00:45:39
Like hopefully there's signs that this is going to start to
00:45:43
happen. I think when this happened, I
00:45:45
think there's a very high number of high quality companies that
00:45:48
are just waiting to go public and that's going to provide a
00:45:51
good amount of liquidity in the system.
00:45:54
I think for the strategic acquisition in AI, I think we're
00:45:56
going to continue to see more because a lot of the big
00:45:59
companies are seeing that they are missing these AI talents.
00:46:02
And even if Meta is ready to some bold move and Meta is, I
00:46:05
mean everybody says, oh, you know, they're behind, but
00:46:08
they're still very good in the high, right.
00:46:11
And so we'll see a lot of other companies saying, OK, now.
00:46:14
We really need to clearly needs to make a move.
00:46:17
Last question quickly. You know, technology is a boom
00:46:21
and bust sort of business with hype cycles and corrections
00:46:25
where anyone brave enough to sort of chart out or say where
00:46:28
you think we are in this boom and bust cycle, How much do do
00:46:31
we have a lot of juice over the next 12 months to keep investing
00:46:34
aggressively on the AI thesis? Definitely an optimist on this
00:46:39
one. I I go go back to the
00:46:41
traditional industries, you know, fintech, banking, lots,
00:46:44
lots to be done. Checkbooks open.
00:46:45
You agree? Yeah.
00:46:46
It's like the drivers for me are, is there a capital?
00:46:49
There absolutely is. And if the IPO windows open,
00:46:51
they'll be more in the private market.
00:46:52
And then secondly, do we think that the models and the tech is
00:46:56
getting better? And I think there's no end
00:46:57
insight pre training and post training and then now with
00:47:01
reasoning there's more progress. So right at our.
00:47:04
Conference in November, we were all worried about a wall and
00:47:07
then post training I think in particular showed with reasoning
00:47:09
models there's plenty of improvement.
00:47:11
Philippe last. Year 100%, yeah, I think the
00:47:13
there there are two things. There is a cycle in prices which
00:47:16
may go ups and down, but I think the secular shift with AI is
00:47:19
here for the next 10 years. I think like we'll have a lot to
00:47:22
invest in for the next 10 years plus.
00:47:24
Great. All right, that's our panel for
00:47:25
the morning. Riley is going to explain what's
00:47:27
next.