July 10, 2025 • 43 mins
Bloomberg’s Caroline Hyde discusses xAI's rollout of Grok 4 as Elon Musk says Grok will come to Tesla vehicles next week. Plus, we hear from LinkedIn co-founder Reid Hoffman as the AI talent war draws big pay packages. And Varda Space raises $187 million in a series C funding as they look towards advancing space medicine manufacturing.
See omnystudio.com/listener for privacy information.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:02):
Bloomberg Audio Studios, podcasts, radio news. Bloomberg Tech is alive
from coast to coast with Caroline Hide in New York
and Eva Though in San Francisco.
Speaker 2 (00:21):
This is Bloomberg Tech coming up.
Speaker 3 (00:23):
Elon Musk's AI startup Xai is rolling out Grock four
and it will be coming to Tesla vehicles quote very soon.
Plus Graylock partner and LinkedIn co founder Reid Hoffman wis
in on the AI talent wars.
Speaker 2 (00:36):
Why hundreds of millions for salaries is not crazy?
Speaker 3 (00:40):
And Varda's Space raises one hundred and eighty seven million
dollars in a series funding to produce medicines in low
Earth orbit. But first we check in on these markets
that just pull back from those near term highs.
Speaker 2 (00:53):
Again, we're actually bouncing off of our lows.
Speaker 3 (00:55):
We're seeing the Nasdaq one only four tenths percent had
sold off.
Speaker 4 (00:58):
More.
Speaker 2 (00:58):
We've got a big bond cell today.
Speaker 3 (01:00):
That's a test of appetite on the long end of
the curve, and that a lot has been dictating what's
happening in the equity markets. We take a pause amid
the ongoing barrage of news when it comes to Taris
Brazil the latest in the line of fire. But we're
looking at cryptos still catching a bid. So because this
were one hundred and eleven thousand, we exceeded one hundred
and twelve thousand yesterday. So this asset of choice at
the moment is up more than twenty percent year to date.
(01:22):
Move on, have a look at some of the individual
moves that dictate the.
Speaker 2 (01:25):
Move of travel.
Speaker 3 (01:25):
And then as that one hundred in videos off by
three ten percent, coming off of that all time high
that we on an intry day basis hit yesterday of
four trillion dollars, we pulled down.
Speaker 2 (01:34):
So therefore than aw's that one hundred poorly.
Speaker 3 (01:35):
Lower TSMC, though actually the ADRs are in the red.
But more broadly, Alyss is saying that the latest June
numbers on sales will then being hit more by FX woes.
This is a strong set of numbers coming from the
AI demand and Apple and in video demand for the
chip making juggernaut that is TSMC, and I'm looking at
Tesla up three percent.
Speaker 2 (01:56):
In excess of this is as we get.
Speaker 3 (01:58):
Not only hints fromine on Mask on his platform X
that we'll see Robotaxi coming out in California in San
Francisco relatively soon. But also Grock four, the announcement of
the chatbot of course that was showed off yesterday, also.
Speaker 2 (02:11):
Coming at Tesla in the near term.
Speaker 3 (02:13):
Let's dig into that with Jackie Devolas and Jackie this
really is moving to stop people anticipating what grop for
in the chat bolt could do when you're in the car.
Speaker 5 (02:23):
That's exactly right, And because with Grock four, Musk actually
presented these two new models kind of as a breakdown
of the release. It had grock four and grock four Heavy,
and the ladder is really the company's this multi agent
version that offers this advanced performance. Musk said that GROK
(02:44):
Heavy could actually deploy multiple agents to work on a
problem and that they can all compare the work almost
like a study group. And it's said that Grock four
actually shows frontier level performance on several benchmarks. So this
is a really exciting moment for the company, especially when
you think about just kind of what a tumultuous week
it has.
Speaker 2 (03:03):
Been for x and XAI. But as for Tesla drivers.
Speaker 5 (03:08):
You know, this has been kind of something that's been
teased and really anticipated just given how intense these AI
wars are coming, and the more applications we can find
for them, companies are really eager to find ways to
monetize that.
Speaker 3 (03:23):
Precisely, many have seen the intertwining of x and XAI,
and then at some point ad Or Musk even saying
maybe Tesla should invest in Xai more broadly because of
the dividends that will pay for the company. Remind us
of the tumultuous nature of this week, though, because we
have glock Falls unveiling hot on the heels of some
pretty difficult content that it was spewing out across the platform.
Speaker 5 (03:46):
That's right, and you know it was related to some
replies to user posts that those replies had anti Semitic tropes,
it was glorifying Adolf Hitler in some others, and XAI
the company was forced to remove these inappropriate posts, and
the company also acknowledged that, you know, they should n't
(04:07):
have been up there.
Speaker 2 (04:07):
But Musk himself kind of.
Speaker 5 (04:09):
Almost avoided explicitly referring to them, but just said that quote,
we need to make sure.
Speaker 2 (04:17):
That AI is good AI. And he he said that
on Wednesday night, so.
Speaker 5 (04:21):
Of course, probably referring to some of those posts in
advance of this release because of course, as he's about
to release the latest and greatest model, the top question
for users broadly is how do we know that this
isn't just going to spew similar content? And then of
course Linda Yakarino stepping down as CEO of x so
(04:42):
there's a lot going on at x and XAI. But
you know this, this latest news that Tesla might have
this latest model is pretty exciting for people out there.
Speaker 3 (04:52):
And helps it inchback on some of the losses experience
earlier in the week because of that ongoing feud between
Elo Musk and President Trump. A caudivolous with all the
laatest on groc four. We so appreciate it. Meanwhile, China's
Huawei is trying to export its AI chips to the
Middle East and Southeast Asia to establish a foothold in
markets dominated by Nvidia's despite ongoing manufacturing challenges for Huawei.
Speaker 2 (05:15):
It's all according to sources our own ed Ludlow is
standing by in Sun Valley and I.
Speaker 3 (05:19):
Want to dig into this with un because it's just
a small quantity of the chips we understand thus far
tentatively going to these places.
Speaker 4 (05:30):
Yeah, low volumes, targeting Saudi Arabia and UAE in the
goal for then Malaysia, in Thailand in Southeast Asia. But
what our reporting tells us, and what our sources are
informing us of, is that Huawei wants to give access
to nine ten B prior older generation AI Accelerator card
so that customers in those regions can experience the technology
(05:50):
and then maybe in the long run look more deeply
at what Huawei has to offer. Completely right, that it's
low volumes in the thousands is what we're reporting, and
Huawei is supply constrained, it has a manufacturing cap and
what it's able to produce. But of course this is
politically difficult as well. It looks like this is Huawei
going out to the market and that there isn't actually
(06:11):
that much reciprocal interest.
Speaker 6 (06:13):
Saudi Arabia, based.
Speaker 4 (06:14):
On our reporting, seems the most likely to look at
those prior generationships. Saudi has a long standing relationship with Huawei,
but has also said that if America turned around and
was like, you're not doing business there, then Saudi cooperates.
So it's a dynamic situation. But Huawei flexing its muscles
a little, and.
Speaker 3 (06:31):
Some dynamic conversations you're having over there in some valley.
How much has everyone been focused in on China US relations?
And indeed the big four trillion number hit by Nvidio yesterday.
Speaker 4 (06:44):
Yeah, I mean, geopolitics and trade are common conversations here.
Speaker 6 (06:48):
You expect that, right.
Speaker 4 (06:50):
You have some of the CEOs of the world's largest
technology companies here who do a lot of business across borders.
Scott Besson, the Treasury Secretary, made a presentation there was
a model rated Q and A late yesterday morning where
he took questions from the audience. Some of those questions
related to tariffs and trade.
Speaker 6 (07:07):
He also kind of.
Speaker 4 (07:07):
Gave his kind of longer term view on the global
economy and on in video. Four trillion. Sure, you know,
it's kind of like a market moment in history. What's
so interesting? And it was part of the conversation that
I had with both Reid Hoffman and Steve Pauuka last night,
which I think we'll play a bit of later in
the show. Is four trillion just the beginning? And they
would argue, yes, it is, because it even though the
(07:31):
cost of inference is coming down, it's in video that
still has a hold on that and that market and
there is more room to run because you have frontier
models like you just discussed with Jackie Davilos on Grock four.
But there are other vertical models, very specific use case
models where we're just getting started in the development process.
Speaker 6 (07:50):
So I found that really interesting.
Speaker 3 (07:52):
Ed Ladlow will be going back to you and some
of your conversations at the Allen and Co.
Speaker 2 (07:56):
Event. Thanks so much.
Speaker 3 (07:57):
Now, look, let's stick with tech chips and in video
being the winning formula with epek Oskodeshka of Swiss quote
and it feels as though you probably are going to
echo what Steve paal Yuka, what read Hoffman, what Ed's
just been saying that we're just at the beginning with
this four trillion number.
Speaker 7 (08:14):
I mean, definitely, the numbers are not surperviising right now,
and when you're looking at the valuation basis, Nvidia is
not necessarily much more expensive today than.
Speaker 2 (08:24):
It was a year ago.
Speaker 7 (08:25):
The pe ratio is totally acceptable, is around fifty two,
but with the upcoming earnings, we think that's going to
be pulled back to a low forties and four trillion number.
It incorporates not only what MVDA achieved so far. Obviously, yes,
but also what it will be achieving and their proactively
searching clients and expanding their client book to other continents
(08:49):
and to other sectors like robotics is very interesting. We
actually think that robotics is going to be the next
big milestone because giving AI hands, feet, and a body
is just going to be amazing. We're just at the
beginning of this AI revolution is just so exciting.
Speaker 2 (09:04):
And other companies benefit too.
Speaker 3 (09:06):
We think of TSMC, which is the chip maker of
choice Foreign Video their numbers once again according to the
Bloomberg Mathematics it is about thirty nine percent increase in
revenue in the previous quarter just gone epec how much
we likely to see other chip makers really fueled higher.
Speaker 2 (09:21):
AMD doing well today on the back of it, well,
we have been.
Speaker 7 (09:25):
Quite positive for AMD hasn't done perfectly well over the
past few months. Now it is back to the bullish consolidation.
So now we think that companies AMD alternatis are cheaper versions.
Will be quite interesting when the AI adoption is going
to go wider because companies, some companies will be looking
for these cheaper alternatives. This thing said, MVD is not
(09:47):
only offering chips, they're offering ecosystem, and this is where
MVDA is still beating this competition. This is why probably
we see the growth and an a in AMD and
investor interest in AMD remain quite though, but we think
that AMD's time will come.
Speaker 3 (10:03):
Both AMD and Nvidia still want to have a presence
in China as an end market, and actually there's reporting
that Jenson Wang is going out to China very soon
because he wants to be making yet another specific chip
that is allowed to be exported. How much is that
a headwind for you or indeed a long term tailwind
whether or not they will have access to this market.
Speaker 7 (10:25):
Well, China is a very sensitive as a politically and
a geopolitical sensitive market, so when Nvidia goes into that market,
we know that the risk on that specific.
Speaker 2 (10:35):
Segment is higher.
Speaker 7 (10:36):
But the company's efforts in expanding beyond China and the
US markets like Europe for example, and Middle East, which
do have deep pockets and very big potential, actually delutes
the risk of while not doing business in China, we
think that China is still worth trying. As Jensen Huang says,
it's not necessarily a good idea to completely ben chips
(10:57):
from China because it will only push them to do better.
So I think that China is.
Speaker 2 (11:01):
Still interesting market. But if the risk could.
Speaker 7 (11:04):
Be diluted like MVDA is doing today, then it's going
to be only a tailwind rather than a headwind.
Speaker 2 (11:10):
Just putting it in the context, and your note does
this so well for us Epek, I love reading them.
Speaker 3 (11:15):
That basically in video is up one thousand percent since
CHATCHIBT came to all of US AI general to AI
was put in yours in my hands.
Speaker 2 (11:24):
How much your client's.
Speaker 3 (11:26):
Bulking at the fact that we have seen this run.
I know you talk about the pes actually big, relatively appetizing.
We are at a technically overbought level, though if you're
looking at the RSI, well yes.
Speaker 7 (11:37):
In the short one, we are at overbought markets conditions,
meaning that NVDA stocks may have been built too quickly
in a too short period of time. But what investors
look with Nvidia is not necessarily the short term dynamics
and short term indicators. They actually do have this very
long term perspective on what Nvidia will be involved in
in the next decade. Nvidia is a company that is
(12:00):
going to be at the heart of AI and robotics,
for the next decade.
Speaker 2 (12:04):
So if the price is over.
Speaker 7 (12:05):
Today, it is not necessarily a major significant downside correction.
Speaker 2 (12:13):
Downside correction potential for the long term.
Speaker 7 (12:15):
Investors are really looking at what Nvidia will mean in two, three, five,
ten years from now, and when you look at that perspective,
the company still has so much to offer.
Speaker 3 (12:26):
We've talked about Asian companies TSMC, for example, you're sat
in Switzerland. We've got a European tech show recently launched
as well. How much are you thinking about the European
opportunity here? Are there any names other than the ASML
that have been coming to light when it comes to.
Speaker 2 (12:40):
AI, Well, we do have a couple of names.
Speaker 7 (12:43):
We obviously don't have names like you do in the US,
like Facebook, Meta, Amazon. So what's going to be interesting
in Europe in terms of AI, I think is going
to be more specific models, sectors specific models, business specific models,
so that they can find the data that they need
in order to train these models. We do not have
(13:04):
these huge AI labs, these huge platforms that could help
building these models. So it's going to be a very
different play in Europe, and European companies will be going
case by case and they will be building smaller, more
specific models in order to benefit from AI's AI benefits.
So it's going to be very different landscape of AI
(13:27):
when you're looking at Europe, rather than what we see
in the.
Speaker 2 (13:29):
US today, the world wide big.
Speaker 7 (13:32):
Companies reaching out to everyone. I think it's going to
be much more B to B business based.
Speaker 3 (13:38):
Ibek PSKODESHKAA great to have some time with you a
Swiss code.
Speaker 2 (13:42):
We appreciate it.
Speaker 3 (13:43):
Like coming up Prime Day, it enters its third day
of deals. People want anything, yet it'll take a look
at whether it's managing to entice shoppers.
Speaker 2 (13:51):
This is Blue Bag Tech.
Speaker 8 (14:07):
We're really pleased by the engagement that we're getting from
our customers and our members. Not only is the event
twice as long, so four days, as you mentioned, but
we're also offering more than double the number of deals
that are fifty percent off or more. That's striving engagement
across the millions of deals we have.
Speaker 6 (14:23):
In thirty five categories.
Speaker 8 (14:24):
So we're really pleased.
Speaker 6 (14:25):
But it's very early.
Speaker 3 (14:26):
Of course, Amazon Prime Vice President Jamil Ghani there weighing
in on the early numbers trickling in yesterday. Let's get
an update now, with Bluemberg Spenser Sopa on day three
of Prime Day. How hard is it to really discern
whether this is going well or not?
Speaker 2 (14:41):
Spencer, Yes, it's very tricky.
Speaker 9 (14:44):
You know, you're comparing a two day event last year
last year to a four day event this year. Amazon's
experiment is like, hey, what if we kind of slow
the sales cadence down a little bit but spread it
over a longer period of days, will that be.
Speaker 6 (14:56):
Better for us?
Speaker 9 (14:57):
And that's the experiment, So that's where these next two
days are really going to be critical. Can Amazon kind
of sustain this kind of more muted momentum over four
days or do people just trail off? So that's really
what we're what we're looking to see on days three
and four.
Speaker 3 (15:12):
I mean talking to momentum. Momentum Commerce was an advisor
that you went to on a data analysis service that
seems to be showing that the numbers were pretty poor
at the start of things.
Speaker 9 (15:24):
Yeah, and again let's emphasize that was the first day
of what is now a four day event as opposed
to the first day last year when it was a
two day event. So when you prolong an event, you
lose that sense of urgency. People aren't necessarily shopping, people
aren't necessarily compelled.
Speaker 6 (15:40):
To act immediately.
Speaker 9 (15:41):
And you know when you had the Amazon executive on yesterday,
the key word there was engagement.
Speaker 6 (15:46):
You know, they're pleased with engagement.
Speaker 9 (15:48):
That means people are kind of poking around, but they're
not really buying, so everything kind of tracks. It's just
a question of you know, are people going to keep
buying through days three and four? And if they do,
Momentum said, you know, the four day event should yield
a lift compared to the two day event last year.
Speaker 6 (16:04):
But that's the big question.
Speaker 9 (16:05):
We'll see if Amazon's experiment is working today and tomorrow.
Speaker 2 (16:09):
Big bet.
Speaker 3 (16:10):
Spenser, Sopa, you stay on it, Thank you so much.
Now coming up, Gecko Robotics says twenty twenty five is
the Year of the robot. We speak with a CEO
live from some Vallely, Idaho. Next this is Bluemberg Tech.
(16:35):
Let's get back out to sun Valley, Idaho, where Ed
Ludlow is standing by with.
Speaker 2 (16:40):
A special guest.
Speaker 10 (16:41):
Ed.
Speaker 4 (16:42):
Yeah, let's talk about robotics and we're not talking humanoid rebulics.
Actually really interesting. Gecko Robotics and his co founder CEO
Jake Lucerian, There's a lot of attention on you. And
one of the reasons why is this super interesting business model.
The robots basically can assess the structural integrity of various facilities,
cool energy infrastructure. But actually a big part of your
(17:04):
business is AI and data. You sell them as a package.
Give us the basics on Gego Robotics and we'll go
from that.
Speaker 11 (17:10):
Yeah, thanks, Ed, Well, it's the company was started out
with this mission of diagnosing the health of the built world,
and so I started out of a college dorm with
that with that ambition emission, and what we were able to
understand and determine is we start with the foundation of
understanding the health of built structures, whether it's a dam
or a bridge, a vessel, of oil and gas refinery.
Speaker 6 (17:31):
Whatever the physical structure is.
Speaker 11 (17:32):
If you understand the health of that, you begin to
get really good at understanding how these facilities work and function.
And then what we've done is we've used all that
information data and the processing data from the facilities and
also the smart centors that we've built to be able
to create an operating platform to help those facilities run
in some cases twenty percent less cost for CAPEX and
(17:53):
running those facilities, but also operate fifteen percent more efficient
in some cases. And so the applications of the fence
and commercial are enabled because.
Speaker 6 (18:02):
We have an unfair data advantage.
Speaker 4 (18:05):
Someone that's never seen a Geho robotics robot scaling a wall,
explain what it looks like, how many of them are
out there and deployed in the real world, and what
kinds of commercial businesses are using them right now.
Speaker 11 (18:17):
Yeah, So the robots feed into the software operating platform
called Candilever, and the robots can look like the size
of a briefcase, climbing up a wall, climbing around, piping.
We also have drones, were walking robots, or smart sensors.
But all these tools are gathering information and data and
plugging into that central software.
Speaker 6 (18:36):
And then what we've been able to.
Speaker 11 (18:37):
See is that customers are actually asking can we put
our operational or our IoT sensor data into that platform
as well to derive better results, outcomes or decisions on
the ground. So what we've done is we've begun to
enable the folks on the ground to become not just
blue collar workers, but engineers because they can use the
LLLM that we have inside of Candile Leaver to be
(18:57):
able to understand what sort of decisions to make because
they have the data that allows them.
Speaker 6 (19:02):
To be superpowered in that way.
Speaker 4 (19:03):
So the energy sector is somewhere where you're very focused.
Next week, in your hometown, home state, there is a
gathering of political gathering with the energy industry. You work
with Legacy Energy to basically say here's how you could
be better, just explain how the type works.
Speaker 11 (19:19):
Yeah, So they're right in Pittsburgh next week, the President
and put on by Senator McCormick, is bringing some of
the biggest names in energy and AI to the city
of Pittsburgh.
Speaker 6 (19:29):
Where Get Goobrobotics is based.
Speaker 11 (19:32):
And so what it's going to be all about is
not just what's the state of things and what the
infrastructure for AI look like, But more importantly, the message
that I'm delivering and I'm speaking on in particular with
some of those leaders is how we should begin to
judge AI's impact and energy by how much AI can
actually create energy. Example, one of the biggest customers that
(19:53):
we have that helps to run five percent of the
power in the US, the five facilities that we've been
able to deploy our technology stack and operating platform at
have been able to show a huge increase to the
heat rate of these natural gas power plants. The extrapolated
impact is about ten gigawatts of more power without having
to create a single new power point. This is a
(20:14):
huge outcome as it relates to how do you combat
this issue. We don't have enough energy to power the
US In particular, to be the leader in artificial intelligence,
energy is a constraint, and so AI should be able
to unlock.
Speaker 6 (20:27):
Energy, not always the other way around.
Speaker 4 (20:29):
And what's so interesting is we're talking about some carbon
based legacy source of energy.
Speaker 6 (20:33):
There's a lot of.
Speaker 4 (20:34):
Focus on you right now because last month you hit
a one point two five billion dollar valuation.
Speaker 6 (20:39):
You want to scale.
Speaker 4 (20:40):
I know that you're kind of very focused on just
keep continuing growing where we see a follow on round.
Speaker 6 (20:46):
Just to keep that on the hardware side going.
Speaker 11 (20:49):
We really so the way that customers buy and work
with get Go is they buy the operating platform and
all the robots and centsives just by the rover.
Speaker 6 (20:57):
They don't buy any of that.
Speaker 11 (20:58):
It all feeds into the outcomes that we are deliver
with the operating platform, and so the biggest oil and
gas companies in the world will be unveiling things in
a couple quarters. Same with manufacturing facilities that are making
steel or making copper. We go into these old sectors
as Silicon Valley has very much forgotten about, and what
that means for us is that the scale and the
(21:18):
impact of pragmatic artificial intelligence robotics.
Speaker 6 (21:21):
Is hitting this inflection point.
Speaker 11 (21:23):
And you think about physical intelligence, and Jensen talks a
lot about this physical intelligence in these built environments, whether
it be in the defense sector that we work in
or it's in the industrial there is very little physical
intelligence and so we are the leading company as it
relates to creating that which enables a moat for our
artificial intelligence to be exceptionally better than those that are
(21:43):
coming with a software only solution. Meaning to the answer
to your question, scaling and increasing the amounts of fuel
to fire to that fire is something that we're going
to need to do the next time we.
Speaker 4 (21:56):
Have fifteen seconds. But your relationship with Nvidia, you're happy
with them.
Speaker 6 (22:00):
Yeah, I'm really happy with them. There have been great but.
Speaker 11 (22:05):
Looking at physical intelligence for these industrial sectors, there's a
lot more to go and so hopefully we can help out.
Speaker 4 (22:12):
Jake Luceerrian is the Gecko Robotics co founder and CEO
out here in some valley, and a lot of people
were interested in talking to this guy, Karro.
Speaker 3 (22:19):
I bet it really does feel like the year of
the robot. Great conversation. I thank you. Coming up, our
analysts say Apple should consider replacing Tim Cook. More on
that note. Next, this is bluebg Tech. Welcome back to
(22:42):
Blue beg Tech. Let's take a quick check on these
markets as we are open for trade, and.
Speaker 2 (22:46):
Then as that one hundred has been flitting.
Speaker 3 (22:48):
Between gains and losses. Earlier in trade, we went down
in the red.
Speaker 2 (22:51):
We're off of our lowers. We're off about a quarter
of a percent.
Speaker 3 (22:54):
But when you've got in video in the red, of
course that ples us a little bit lower. But more generally,
we're waiting on a big bond orction. A lot of
the yield movements been dictating what's happening in the stock
market right now, but we are still near all time highs.
Let's move on and think about what is very close
to an all time high. We hit in excess of
one hundred and twelve thousand dollars for bitcoin yesterday. We're
(23:15):
at sixtens a percent on the interday basis, one hundred
and eleven thousand, ninety five is where we trade, and
it really does seem that in mid the trade anxiety
there still is a risk asset of choice.
Speaker 2 (23:25):
Of course, apparently next week is crypto week according to
the House.
Speaker 3 (23:30):
In terms of Washington, we'll see how those changes in
rules and regulations continue to transpire. But let's get back
to some of the big tech news here. When it
comes to AI, Google's Gemina Gemini AI gets a new feature.
The company says it's the latest AI update, can now
produce eight second videos out of ordinary still pictures if
you're a paid user.
Speaker 2 (23:50):
For all, Bloomberg's Dana Woman is here to join us.
Speaker 3 (23:53):
This is really AI ultra AI pro users of Google
being able to keep up with what perhaps you're already
getting on an open AI or with a runway for example.
Speaker 12 (24:03):
Yes, this is a feature that will be available to
paid Gemini users, and it is a technology that Google
first launched at its Io developer conference back in May.
But this makes it available to a much broader swath
of users, allowing it, as you said, to better keep
pace with rivals.
Speaker 2 (24:18):
And it does have many rivals in this space, it
does have rivals.
Speaker 3 (24:21):
It does also have a stop start feeling to Google
because many would say, look, there were the reason they're
underlying research is why we've got general to AI in
the first place with the transformer perspective. But they had
this sort of inevitable feeling that they were losing pace
with rivals. Then they had the issue about whether or
not there was some generative AI image features that change
(24:43):
people's race, or there was a sort of a concern there.
Speaker 2 (24:47):
This is being replicated again in the video or no.
Speaker 12 (24:51):
Yes, So Google had a while ago pause certain generative
AI features just for that reason. But here it there
are some guardrails in place that said. We tested it
ourselves and the results were both fascinating and somewhat mixed.
For instance, reporter Natalie Lung, who is Asian, she did
(25:12):
include a photo of herself and the AI did change
her race in one of the videos without her asking it.
To other goofs, we did ask for a video of
someone breakdancing, we did not see any break dancing. Sadly,
the AI did follow other instructions we gave faithfully, but
not that one. So the results were mixed and we
do expect the technology to improve.
Speaker 3 (25:34):
And people do seem to really love what's been coming
out in terms of image generation coming from Google and Gemini.
Speaker 2 (25:40):
Writ large of late.
Speaker 3 (25:41):
But meanwhile Apple everyone still thinks is behind the curve
when it comes to AI.
Speaker 2 (25:45):
Is interesting.
Speaker 3 (25:46):
We've got a note out today with analysts really talking
about how maybe it's time to replace Tim Cook even
we were hearing some view that really the AI focus
is what needs to win here from a product perspective,
rather than having someone in charge who.
Speaker 2 (26:00):
Owns the supply chain. But the fierce talent for AI.
Speaker 3 (26:03):
The wall going on is an amazing scoop that Mark
German's been bringing us all week. And now we understand
two hundred million might have been parted with to.
Speaker 2 (26:11):
Nab the head of AI.
Speaker 12 (26:13):
A pay package worth more than two hundred million. And
I should say that this package does include not just
a base salary and assigning bonus, but stock options. And
there are a few things tied to that. One, someone
would have to stay with the company for a certain
period of time, so they wouldn't get all that money
if they left early.
Speaker 2 (26:28):
And it also is tied.
Speaker 12 (26:30):
To the performance of the stock. So it does assume
that the stock would reach certain levels, so if it doesn't,
someone wouldn't receive that quite that large of a pay
package in the end.
Speaker 3 (26:39):
And that is, of course all as we see transfer
of talent going from Apple to Meta or everyone to Meta.
As Mark Zuckerberg is all in on winning the superintelligence race.
Where are we though, with the feeling around Apple right
now as to ultimately whether it's business model is really
necessary to be the generative AI developer of choice, So
whether it ultimately can be the generator of AI products
(27:01):
user of choice, we'd just go to it to use
other AI offronts.
Speaker 12 (27:05):
So just given its overall stack performance and its market cap,
I really hesitate to call it struggling, but I think
it struggles in AI specifically have been well documented.
Speaker 2 (27:14):
I think it's.
Speaker 12 (27:15):
Efforts to have a new blockbuster hardware hit have been
well documented. I think we know that the Vision pro
has not been a great seller, even though reviewers and
sort of tech enthusiasts have praised it for being innovative.
And yes, the company does have a way to go,
ways to go in AI, and certainly the poaching is
(27:36):
yet another setback in that space.
Speaker 3 (27:39):
And maybe that's why hardware focusing with Apple Senior VP
of Hardware Engineering, John Turnus's why he's being singled out
by the note coming from Lightshed analysts as to whether
he'd be the right replacement for Tim Cook Data Wollman.
Fantastic to have her on all things Google and Apple. Now,
let's return to that talent war that we were just
talking of. Gray Oxweed Hoffmann calls multimillion dollar offers for
(28:02):
a talent quote not crazy when such investments are expected
to generate billions for a company. Now, he weighed in
on the escalating AI talent war between Apple and Meta
their own ed Ludlow over in Sun Valley.
Speaker 13 (28:14):
Well, so, essentially, one of the things that's good about
having an open marketplace of anything, including labor, is that
you essentially get multiple bidders, and that's what sets a price.
And so the fact that someone goes, I'm willing to
spend one hundred million dollars in order to get this talent.
Speaker 6 (28:28):
You think that's crazy.
Speaker 13 (28:29):
Why would it possibly do? That's like higher than the
vast majority of CEOs the construct, Well, if that one,
what you're judging is that one talent will potentially create
billions of dollars of value for you, whether it's in
a social network, whether it's in you know, kind of
you know, a search engine, anything else as a way
of doing it. And so if you think that, then
(28:52):
that's just a bold bet.
Speaker 4 (28:54):
What Sam Oltman, the CEO of open ai, said was
that that particular figure which he had accused Meta of
offering to some open ai staff, what he claimed was
it was crazy. That it sounds like it's not crazy.
Speaker 13 (29:05):
Really, well, I don't think it's crazy when we know
that AI is going to transform all businesses in all industry,
in the entire tech industry, and these are trillion dollar
companies et cetera at the high end and the whole
stack all the way through undown in.
Speaker 6 (29:19):
A billion today.
Speaker 4 (29:20):
For a brief moment, one of them was a four
trillion dolesmpany exactly, please continue.
Speaker 13 (29:23):
Yes, Nvidia, and so the you know so saying hey,
this piece of talent is a risk adjusted bet. That
that makes a difference in that that's just a risk bet.
You might say it's a dumb risk bet, but it's
not done to do if those are the stakes you're
playing for.
Speaker 3 (29:44):
Gray Lock partner and LinkedIn co founder Reid Hoffman with
ed Ludlow. Coming up, we speak with the CEO and
co founders Varda Space the latest series funding.
Speaker 2 (29:54):
Again, we're live from some valley. Stick with us. This
is really bad time.
Speaker 3 (30:09):
Welcome back to Bloomberg Tech. Now, Varda Space Industries has
announced it's raised one hundred and eighty seven million dollars
in the Series C funding, bringing its total capital raise
to three undred and twenty five million dollars. Then the
company aims to produce medicines in low Earth orbit. And
(30:29):
Ludlow is standing by in Sun Valley with Delian Asparov is,
co founder of Varda, and we're also joined by will Brewy,
Varda CEO. And will I start with you the money,
what will you use it for?
Speaker 14 (30:43):
Oh, we have big plans. The Series C raise is
great because it allows us to open up our biologics
lab down the street, and the entire purpose of that
is to prepare biologics formulations, so medicines for flight in
outer space. And the entire purpose of this is micro
gravity allows us to create new drug formulations that otherwise
(31:03):
would be impossible to create on Earth. That's the whole
purpose of Varda. And so with this Series C funding
we're able to build out a large lab space to
vet those drug formulations before flying them and create an
entire pipeline on the way to space.
Speaker 4 (31:19):
Delian, I just want to point something out, which is
that you were also a partner at Founder's Fund and
this has worked exactly the same way it always works.
You recuse yourself from the round and you represent VARDA,
but they participated as depeated t as an individual. Just
expand a little bit in what we were saying. You know,
one of the big changes that you guys have made
over the last yearies to start manufacturing yourself away from
(31:41):
Rocket Lab who used to do that as kind of
the third party for you. Why is that significant and
how does that ramp look?
Speaker 10 (31:47):
Yeah, I mean they were you know, sort of phenomenal
partner and were deeply grateful you know too them and
there they got us off the ground, you know, very quickly.
But as we were looking to you know, sort of
scale up our cadence of flights and ability to truly
control our supply chain, we had.
Speaker 6 (31:58):
To go more full stacking.
Speaker 10 (32:00):
We see this in a variety of different aerospace companies,
where as they scale up, they tend to go more
full stack over time. So that's going only to allow
us to you know, fly four times next year, and
then after that ideally start doubling basically every single year
and be able to control that doubling entirely in house.
Speaker 6 (32:14):
And then it just expands the.
Speaker 10 (32:15):
Capabilities that we have on the spacecraft in terms of
you know, biopharmaceutical production. By owning more of it, we're
able to expand the capabilities of what we can have
on board.
Speaker 4 (32:23):
Well, Dell's talking about all of the rampant launch getting
to orbit. The big beautiful benefit is zero gravity right
in the pharmaceuticals context. But then you've got to get
back down to Earth and that's been one of the
really exciting parts to track with Vada. The video in
particular explain how that process works and also how it's
(32:44):
going to go going forward.
Speaker 14 (32:46):
Yeah, so you hit the nail on the head. The
entire reason for going to space is for micro gravity.
It's just a fundamental force of physics that you can't
change on Earth. You know, we don't want to go
to space to have to make these new drugs. You
just have to because of the laws of physics, and
so our spacecraft allows us to do that quickly and cheaply,
and a lot and a part of that is once
we make those drugs in orbit in microgravity so that
(33:09):
they're different than they would be made if they were
on Earth, we have to bring them back and that's
certainly the most exciting part for the aerospace engineers here.
You know, my background is also airspace engineering. Like you said,
the video is phenomenal. We hit the atmosphere going mock
twenty five, So that's twenty five times the speed of sound,
twenty seven times the speed that you go on a
commercial airliner, and so you rip apart the atmosphere and
(33:31):
you create this beautiful plasma of a shooting star for
over a minute across the sky. It's beautiful on the
way back.
Speaker 3 (33:38):
The passion is real, Delian, the demand is real as well.
I'm interested in the DoD and the government perspective here
because we have seen so much more focus coming on
space and defense more broadly.
Speaker 10 (33:50):
Yeah, you know, sort of new Secretary of Defense in
the early days of the Trump administration came out with
what his national security priorities were going to be a
top three where nuclear weapons modernization, the development of next
generation hypersonic capabilities, and shipbuilding and you know, while VARDA
doesn't really you know, sort of build chips, you know,
we do you help the DoD you know, sort of
(34:11):
with the other two and have a variety of different
you know, sort of collaborations across groups from the Air Force, Navy,
and NASA. What they're primarily interested in is as well
mentioned as we come into the atmosphere we're going mock
twenty five. It's regime that typically is very difficult to
study on the ground and really understand what the flight
dynamics will be like. And so you can think about
the value proposition for VARDA to the Defense Department as
(34:33):
effectively like a hypersonic wind tunnel. They can attach net
new materials, wing shapes, sensors onto the vehicle as we're
coming in as a way to validate the subsystems performance
before they go and take that and integrate that into
the ultimate you know, sort of production system. And so
by enabling faster testing, more frequent testing, cheaper testing, it
(34:53):
allows us to expand on our capabilities and ideally both
catch up to you and surpass some of our new
europeer adversaries like China in Russia that have demonstrated some
pretty phenomenal hypersonic boost glide capabilities.
Speaker 3 (35:04):
I also think that the competition is somewhat closer to home,
or maybe it's a frenemy situation. Will you have got
such a broad experience, You've been flying the Dragon yourself
when you were working at SpaceX. You've also worked in banking,
you've founded your own venture fund in many ways, But
will I ask you now, what is your biggest headache?
(35:26):
Is it talent? Is it the race for talent? How
do you expand?
Speaker 9 (35:31):
Well?
Speaker 14 (35:31):
Believe it or not, there's not that much competition in
re entry because we're so much different than other commercial
re entry vehicles like SpaceX's Dragon or Boeing Starliner. We
are much smaller, we're off access in almost every way.
We're much smaller, we're higher cadence, we're not human rated,
and we're lower cost. So it's more of a complementary situation.
As far as headaches go, you know, a lot of
(35:53):
them are gone now thanks to a working spacecraft and
funding to grow. It's more about enthusiasm and exciting here.
I would say the challenge over the next couple of
years is really to scale up that cadence because as
we scale cadence. We begin to look more and more
just like a zero gravity oven. To the pharmaceutical industry,
they they quite frankly, don't care that we're going to
(36:16):
outer space. They only care that we can deliver new
drug formulations, and we just happen to have to do
that by going the space and back. So the more
often and the more frequent we can do that, the
more easily it is for us to integrate into the
pharmaceutical supply chain. And so that's the big challenge over
the next couple of years is increase cadence, fly more
spacecraft and fly more, develop more drugs on board.
Speaker 4 (36:39):
Dellian, The misconception maybe or misunderstanding about what Will's talking
about ramping the cadence is the cost consideration flying SpaceX
on transport emissions. Actually, my understanding is it's not that
great a cost. But would you talk a bit about
the relationship and how it works hitching a ride on
Falcon nine.
Speaker 10 (36:59):
Yeah, I mean to give you a send. It's like,
of our total all in mission costs that like space X,
like launch costs represent about twenty twenty five percent of
our costs.
Speaker 6 (37:05):
So it's not like the.
Speaker 10 (37:07):
You know, majority or anything. It's also you know, definitely
one of the larger line on it was, but there's
a lot of other things that go into operating either
of our business. Our business could not exist without the
you know sort of Falcon nine being reusable and that
you know sort of transporter program. If you think about
even you know, sort of five six years ago, your
only real option to get up to orbit was having
to buy a full dedicated rocket or find some way
to squeeze yourself onto a large geosats you know, sort
(37:29):
of launch and you know be on board, but you know,
find your own way to orbit. Now, these transporter missions,
they're doing them, I think on the order of three
times a year. They start to add on you know,
different inclinations in orbit. They have these bandwagon bandwagon missions
that go to you know, sort of polar orbit, and
so it's been a really you know, sort of critical
you know, sort of part of our ability to actually succeed.
Speaker 6 (37:47):
As a business.
Speaker 10 (37:47):
I think there's just this interesting you know of pattern
and technological history where when certain infrastructure layers get built up,
it enables these net new you know sort of business
opportunities right in some ways for a business like Uber
Airbnb to get built, you kind of needed mobile to
be very distributed and have the app store, and that's
what really enabled those business models. We think of, you know,
sort of the access to space and what SpaceX provides
us in the same way. It's a fundamental infrastructure layer
(38:10):
that enables our business and it wouldn't have been possible,
you know, sort of without them, and so we're thrilled
to have them, you know, you know sort of partners
and you know, have lots and lots of you know,
rockets out you know, sort of bud with the SpaceX.
I think, you know, we've booked out all the way through.
I think like Q three twenty seven or something like that.
Speaker 4 (38:24):
You know, will we recognize here on Bloomberg Tech there
are many in the space industry who have looked to
the defense apparatus of this country and the defense use
case to kind of start to get off the ground
because of grants and because of the availability of freedom
to move. But long term, what is the business model?
Who will your customers be when you're able to bring
(38:46):
payload back down to Earth. Pharmaceuticals developed in that zero
gravity environment.
Speaker 14 (38:52):
Yeah, it's it is a new technology and what's cool
going to the defense aspect and Deli and hinted towards us,
is that because reusable andmercial launch is a unique American asset,
VARDA is also a unique American asset on top of that,
so it's almost a leap frog above everyone else. And
so we have this cool armada of small, low cost,
(39:13):
high cadence re entry vehicles that can provide a service
continuously for the Department of Defense as a hypersonic wind
tunnel like Dally mentioned. So to answer your question, what
that means is now the Department of Defense can test
things like sensors, new materials for heat shields, they can
test flight computers and RF or radio frequency systems navigation systems.
(39:36):
So very similar to like a wind tunnel where it's
acting as infrastructure so that the Department of Defense can
advance new technologies faster and cheaper.
Speaker 2 (39:46):
Fascinating conversation.
Speaker 3 (39:48):
Thank you for all joining us, Will Brewy, Delian, Asmare
Evada and of course Ed stick.
Speaker 2 (39:53):
With us out there in some valley. We thank you.
Speaker 3 (39:55):
Now, coming up, we're going to hear from s Pat
Yukaro Bain Capital and his outlook the AI sector. This
is Bloomberg Tech let's head back to some value, where
(40:18):
the future of AI is one of the hot topics
among tech leaders. Steve Baruca of main Capital sat down
with Ed Ludlow to discuss well why the AI revolution
reminds him of the early days of the Internet and
actually talk about the potential for AI to transform productivity.
Speaker 15 (40:33):
It reminds me a lot of the end of the
period of kind of ninety eight, ninety nine, two thousand
and two thousand.
Speaker 6 (40:38):
And one in the Internet. That's worrying. Well, no, no, no, no.
Speaker 15 (40:43):
There were some fits and starts, and there'll be some
fits and starts in AI, but I think we're past
the almost past the fits and start stage. You've got
kind of nine global foundation model companies the Internet out there,
perplexity of open AI. We actually funded a startup from
MIT that has modeled the new AI system on a
(41:04):
worm's brain.
Speaker 6 (41:05):
Believe it or not?
Speaker 15 (41:06):
Do you know do you know how many neurons and
worms brain hasn't it?
Speaker 6 (41:08):
Educate me?
Speaker 15 (41:09):
Three hundred and five?
Speaker 4 (41:10):
Okay?
Speaker 15 (41:10):
How many? How many neurons do you think a human
brain hasn't many?
Speaker 6 (41:13):
Many more?
Speaker 15 (41:14):
Eighty six billion?
Speaker 6 (41:15):
Right? And you're smarter than most. Maybe you have eighty
seven billions kind of you.
Speaker 15 (41:18):
But they they these wonderful scientists that a MIT came
up with this model called the neural network model. Of
those nine companies, they're one of those nine companies out
there that are state of the art. They actually because
they they've used the neural network model. They use eighty
percent less GPUs to build the model, and therefore they'll
(41:38):
be perfect for edge devices like phones, because you don't
need the power and so a chet GPT model costs
about five hundred million dollars of GPU to load it.
You can do that for eighty five or ninety percent less,
and this model can sell on phones. And they're they're
just they're they're signing deals with edge device companies, you
know as we speak. So it's very exciting, and all
the costs are going down for the chips, for the
(41:59):
for the for the model loading. And as opposed to
the Internet, which there were some very few big winners,
I think there will be a lot of winners here
because there's going to be vertical models and foundational models,
and the vertical models will have the chuck full of expertise.
First example, a medical diagnosis model that focuses on disease
states and giving you a diagnosis. And some of these
(42:19):
models are proving much better than humans because it's a
game of elimination. Once you put the right data in,
it gives you the diagnosis. The same thing reading x rays. Right,
the models AI models reading x rays are now beating
the top humans in the field. So really it's really
we're only still in the early stages. It's going to
revolutionize productivity. It's going to have the same impact that
(42:40):
the Internet had twenty years ago, and it's going to
be a twenty year run. And that's why Navidia's values
so high, because they're in the cat bird seed of
the chips needed to power that revolution.
Speaker 3 (42:50):
Steve Padiuca a baying capital and the Boston Celtics. Now
that does it for this edition of Bloomberg Tech. Next week,
bloombag Tech will be live from Panasics new EV battery
plant in Kansas. Tune in for an exclusive interview with
the Panasonic North America CEO on Monday.
Speaker 2 (43:08):
And meanwhile, len't forget to check out our podcast.
Speaker 3 (43:10):
We'll find an on the terminal as well as online
on Apple, Spotify, and iHeart from the Sun Valley, Idaho,
or from New York.
Speaker 2 (43:17):
This is Bloomberg Tech
Bloomberg Audio Studios, podcasts, radio news. Bloomberg Tech is alive
from coast to coast with Caroline Hide in New York
and Eva Though in San Francisco.
Speaker 2 (00:21):
This is Bloomberg Tech coming up.
Speaker 3 (00:23):
Elon Musk's AI startup Xai is rolling out Grock four
and it will be coming to Tesla vehicles quote very soon.
Plus Graylock partner and LinkedIn co founder Reid Hoffman wis
in on the AI talent wars.
Speaker 2 (00:36):
Why hundreds of millions for salaries is not crazy?
Speaker 3 (00:40):
And Varda's Space raises one hundred and eighty seven million
dollars in a series funding to produce medicines in low
Earth orbit. But first we check in on these markets
that just pull back from those near term highs.
Speaker 2 (00:53):
Again, we're actually bouncing off of our lows.
Speaker 3 (00:55):
We're seeing the Nasdaq one only four tenths percent had
sold off.
Speaker 4 (00:58):
More.
Speaker 2 (00:58):
We've got a big bond cell today.
Speaker 3 (01:00):
That's a test of appetite on the long end of
the curve, and that a lot has been dictating what's
happening in the equity markets. We take a pause amid
the ongoing barrage of news when it comes to Taris
Brazil the latest in the line of fire. But we're
looking at cryptos still catching a bid. So because this
were one hundred and eleven thousand, we exceeded one hundred
and twelve thousand yesterday. So this asset of choice at
the moment is up more than twenty percent year to date.
(01:22):
Move on, have a look at some of the individual
moves that dictate the.
Speaker 2 (01:25):
Move of travel.
Speaker 3 (01:25):
And then as that one hundred in videos off by
three ten percent, coming off of that all time high
that we on an intry day basis hit yesterday of
four trillion dollars, we pulled down.
Speaker 2 (01:34):
So therefore than aw's that one hundred poorly.
Speaker 3 (01:35):
Lower TSMC, though actually the ADRs are in the red.
But more broadly, Alyss is saying that the latest June
numbers on sales will then being hit more by FX woes.
This is a strong set of numbers coming from the
AI demand and Apple and in video demand for the
chip making juggernaut that is TSMC, and I'm looking at
Tesla up three percent.
Speaker 2 (01:56):
In excess of this is as we get.
Speaker 3 (01:58):
Not only hints fromine on Mask on his platform X
that we'll see Robotaxi coming out in California in San
Francisco relatively soon. But also Grock four, the announcement of
the chatbot of course that was showed off yesterday, also.
Speaker 2 (02:11):
Coming at Tesla in the near term.
Speaker 3 (02:13):
Let's dig into that with Jackie Devolas and Jackie this
really is moving to stop people anticipating what grop for
in the chat bolt could do when you're in the car.
Speaker 5 (02:23):
That's exactly right, And because with Grock four, Musk actually
presented these two new models kind of as a breakdown
of the release. It had grock four and grock four Heavy,
and the ladder is really the company's this multi agent
version that offers this advanced performance. Musk said that GROK
(02:44):
Heavy could actually deploy multiple agents to work on a
problem and that they can all compare the work almost
like a study group. And it's said that Grock four
actually shows frontier level performance on several benchmarks. So this
is a really exciting moment for the company, especially when
you think about just kind of what a tumultuous week
it has.
Speaker 2 (03:03):
Been for x and XAI. But as for Tesla drivers.
Speaker 5 (03:08):
You know, this has been kind of something that's been
teased and really anticipated just given how intense these AI
wars are coming, and the more applications we can find
for them, companies are really eager to find ways to
monetize that.
Speaker 3 (03:23):
Precisely, many have seen the intertwining of x and XAI,
and then at some point ad Or Musk even saying
maybe Tesla should invest in Xai more broadly because of
the dividends that will pay for the company. Remind us
of the tumultuous nature of this week, though, because we
have glock Falls unveiling hot on the heels of some
pretty difficult content that it was spewing out across the platform.
Speaker 5 (03:46):
That's right, and you know it was related to some
replies to user posts that those replies had anti Semitic tropes,
it was glorifying Adolf Hitler in some others, and XAI
the company was forced to remove these inappropriate posts, and
the company also acknowledged that, you know, they should n't
(04:07):
have been up there.
Speaker 2 (04:07):
But Musk himself kind of.
Speaker 5 (04:09):
Almost avoided explicitly referring to them, but just said that quote,
we need to make sure.
Speaker 2 (04:17):
That AI is good AI. And he he said that
on Wednesday night, so.
Speaker 5 (04:21):
Of course, probably referring to some of those posts in
advance of this release because of course, as he's about
to release the latest and greatest model, the top question
for users broadly is how do we know that this
isn't just going to spew similar content? And then of
course Linda Yakarino stepping down as CEO of x so
(04:42):
there's a lot going on at x and XAI. But
you know this, this latest news that Tesla might have
this latest model is pretty exciting for people out there.
Speaker 3 (04:52):
And helps it inchback on some of the losses experience
earlier in the week because of that ongoing feud between
Elo Musk and President Trump. A caudivolous with all the
laatest on groc four. We so appreciate it. Meanwhile, China's
Huawei is trying to export its AI chips to the
Middle East and Southeast Asia to establish a foothold in
markets dominated by Nvidia's despite ongoing manufacturing challenges for Huawei.
Speaker 2 (05:15):
It's all according to sources our own ed Ludlow is
standing by in Sun Valley and I.
Speaker 3 (05:19):
Want to dig into this with un because it's just
a small quantity of the chips we understand thus far
tentatively going to these places.
Speaker 4 (05:30):
Yeah, low volumes, targeting Saudi Arabia and UAE in the
goal for then Malaysia, in Thailand in Southeast Asia. But
what our reporting tells us, and what our sources are
informing us of, is that Huawei wants to give access
to nine ten B prior older generation AI Accelerator card
so that customers in those regions can experience the technology
(05:50):
and then maybe in the long run look more deeply
at what Huawei has to offer. Completely right, that it's
low volumes in the thousands is what we're reporting, and
Huawei is supply constrained, it has a manufacturing cap and
what it's able to produce. But of course this is
politically difficult as well. It looks like this is Huawei
going out to the market and that there isn't actually
(06:11):
that much reciprocal interest.
Speaker 6 (06:13):
Saudi Arabia, based.
Speaker 4 (06:14):
On our reporting, seems the most likely to look at
those prior generationships. Saudi has a long standing relationship with Huawei,
but has also said that if America turned around and
was like, you're not doing business there, then Saudi cooperates.
So it's a dynamic situation. But Huawei flexing its muscles
a little, and.
Speaker 3 (06:31):
Some dynamic conversations you're having over there in some valley.
How much has everyone been focused in on China US relations?
And indeed the big four trillion number hit by Nvidio yesterday.
Speaker 4 (06:44):
Yeah, I mean, geopolitics and trade are common conversations here.
Speaker 6 (06:48):
You expect that, right.
Speaker 4 (06:50):
You have some of the CEOs of the world's largest
technology companies here who do a lot of business across borders.
Scott Besson, the Treasury Secretary, made a presentation there was
a model rated Q and A late yesterday morning where
he took questions from the audience. Some of those questions
related to tariffs and trade.
Speaker 6 (07:07):
He also kind of.
Speaker 4 (07:07):
Gave his kind of longer term view on the global
economy and on in video. Four trillion. Sure, you know,
it's kind of like a market moment in history. What's
so interesting? And it was part of the conversation that
I had with both Reid Hoffman and Steve Pauuka last night,
which I think we'll play a bit of later in
the show. Is four trillion just the beginning? And they
would argue, yes, it is, because it even though the
(07:31):
cost of inference is coming down, it's in video that
still has a hold on that and that market and
there is more room to run because you have frontier
models like you just discussed with Jackie Davilos on Grock four.
But there are other vertical models, very specific use case
models where we're just getting started in the development process.
Speaker 6 (07:50):
So I found that really interesting.
Speaker 3 (07:52):
Ed Ladlow will be going back to you and some
of your conversations at the Allen and Co.
Speaker 2 (07:56):
Event. Thanks so much.
Speaker 3 (07:57):
Now, look, let's stick with tech chips and in video
being the winning formula with epek Oskodeshka of Swiss quote
and it feels as though you probably are going to
echo what Steve paal Yuka, what read Hoffman, what Ed's
just been saying that we're just at the beginning with
this four trillion number.
Speaker 7 (08:14):
I mean, definitely, the numbers are not surperviising right now,
and when you're looking at the valuation basis, Nvidia is
not necessarily much more expensive today than.
Speaker 2 (08:24):
It was a year ago.
Speaker 7 (08:25):
The pe ratio is totally acceptable, is around fifty two,
but with the upcoming earnings, we think that's going to
be pulled back to a low forties and four trillion number.
It incorporates not only what MVDA achieved so far. Obviously, yes,
but also what it will be achieving and their proactively
searching clients and expanding their client book to other continents
(08:49):
and to other sectors like robotics is very interesting. We
actually think that robotics is going to be the next
big milestone because giving AI hands, feet, and a body
is just going to be amazing. We're just at the
beginning of this AI revolution is just so exciting.
Speaker 2 (09:04):
And other companies benefit too.
Speaker 3 (09:06):
We think of TSMC, which is the chip maker of
choice Foreign Video their numbers once again according to the
Bloomberg Mathematics it is about thirty nine percent increase in
revenue in the previous quarter just gone epec how much
we likely to see other chip makers really fueled higher.
Speaker 2 (09:21):
AMD doing well today on the back of it, well,
we have been.
Speaker 7 (09:25):
Quite positive for AMD hasn't done perfectly well over the
past few months. Now it is back to the bullish consolidation.
So now we think that companies AMD alternatis are cheaper versions.
Will be quite interesting when the AI adoption is going
to go wider because companies, some companies will be looking
for these cheaper alternatives. This thing said, MVD is not
(09:47):
only offering chips, they're offering ecosystem, and this is where
MVDA is still beating this competition. This is why probably
we see the growth and an a in AMD and
investor interest in AMD remain quite though, but we think
that AMD's time will come.
Speaker 3 (10:03):
Both AMD and Nvidia still want to have a presence
in China as an end market, and actually there's reporting
that Jenson Wang is going out to China very soon
because he wants to be making yet another specific chip
that is allowed to be exported. How much is that
a headwind for you or indeed a long term tailwind
whether or not they will have access to this market.
Speaker 7 (10:25):
Well, China is a very sensitive as a politically and
a geopolitical sensitive market, so when Nvidia goes into that market,
we know that the risk on that specific.
Speaker 2 (10:35):
Segment is higher.
Speaker 7 (10:36):
But the company's efforts in expanding beyond China and the
US markets like Europe for example, and Middle East, which
do have deep pockets and very big potential, actually delutes
the risk of while not doing business in China, we
think that China is still worth trying. As Jensen Huang says,
it's not necessarily a good idea to completely ben chips
(10:57):
from China because it will only push them to do better.
So I think that China is.
Speaker 2 (11:01):
Still interesting market. But if the risk could.
Speaker 7 (11:04):
Be diluted like MVDA is doing today, then it's going
to be only a tailwind rather than a headwind.
Speaker 2 (11:10):
Just putting it in the context, and your note does
this so well for us Epek, I love reading them.
Speaker 3 (11:15):
That basically in video is up one thousand percent since
CHATCHIBT came to all of US AI general to AI
was put in yours in my hands.
Speaker 2 (11:24):
How much your client's.
Speaker 3 (11:26):
Bulking at the fact that we have seen this run.
I know you talk about the pes actually big, relatively appetizing.
We are at a technically overbought level, though if you're
looking at the RSI, well yes.
Speaker 7 (11:37):
In the short one, we are at overbought markets conditions,
meaning that NVDA stocks may have been built too quickly
in a too short period of time. But what investors
look with Nvidia is not necessarily the short term dynamics
and short term indicators. They actually do have this very
long term perspective on what Nvidia will be involved in
in the next decade. Nvidia is a company that is
(12:00):
going to be at the heart of AI and robotics,
for the next decade.
Speaker 2 (12:04):
So if the price is over.
Speaker 7 (12:05):
Today, it is not necessarily a major significant downside correction.
Speaker 2 (12:13):
Downside correction potential for the long term.
Speaker 7 (12:15):
Investors are really looking at what Nvidia will mean in two, three, five,
ten years from now, and when you look at that perspective,
the company still has so much to offer.
Speaker 3 (12:26):
We've talked about Asian companies TSMC, for example, you're sat
in Switzerland. We've got a European tech show recently launched
as well. How much are you thinking about the European
opportunity here? Are there any names other than the ASML
that have been coming to light when it comes to.
Speaker 2 (12:40):
AI, Well, we do have a couple of names.
Speaker 7 (12:43):
We obviously don't have names like you do in the US,
like Facebook, Meta, Amazon. So what's going to be interesting
in Europe in terms of AI, I think is going
to be more specific models, sectors specific models, business specific models,
so that they can find the data that they need
in order to train these models. We do not have
(13:04):
these huge AI labs, these huge platforms that could help
building these models. So it's going to be a very
different play in Europe, and European companies will be going
case by case and they will be building smaller, more
specific models in order to benefit from AI's AI benefits.
So it's going to be very different landscape of AI
(13:27):
when you're looking at Europe, rather than what we see
in the.
Speaker 2 (13:29):
US today, the world wide big.
Speaker 7 (13:32):
Companies reaching out to everyone. I think it's going to
be much more B to B business based.
Speaker 3 (13:38):
Ibek PSKODESHKAA great to have some time with you a
Swiss code.
Speaker 2 (13:42):
We appreciate it.
Speaker 3 (13:43):
Like coming up Prime Day, it enters its third day
of deals. People want anything, yet it'll take a look
at whether it's managing to entice shoppers.
Speaker 2 (13:51):
This is Blue Bag Tech.
Speaker 8 (14:07):
We're really pleased by the engagement that we're getting from
our customers and our members. Not only is the event
twice as long, so four days, as you mentioned, but
we're also offering more than double the number of deals
that are fifty percent off or more. That's striving engagement
across the millions of deals we have.
Speaker 6 (14:23):
In thirty five categories.
Speaker 8 (14:24):
So we're really pleased.
Speaker 6 (14:25):
But it's very early.
Speaker 3 (14:26):
Of course, Amazon Prime Vice President Jamil Ghani there weighing
in on the early numbers trickling in yesterday. Let's get
an update now, with Bluemberg Spenser Sopa on day three
of Prime Day. How hard is it to really discern
whether this is going well or not?
Speaker 2 (14:41):
Spencer, Yes, it's very tricky.
Speaker 9 (14:44):
You know, you're comparing a two day event last year
last year to a four day event this year. Amazon's
experiment is like, hey, what if we kind of slow
the sales cadence down a little bit but spread it
over a longer period of days, will that be.
Speaker 6 (14:56):
Better for us?
Speaker 9 (14:57):
And that's the experiment, So that's where these next two
days are really going to be critical. Can Amazon kind
of sustain this kind of more muted momentum over four
days or do people just trail off? So that's really
what we're what we're looking to see on days three
and four.
Speaker 3 (15:12):
I mean talking to momentum. Momentum Commerce was an advisor
that you went to on a data analysis service that
seems to be showing that the numbers were pretty poor
at the start of things.
Speaker 9 (15:24):
Yeah, and again let's emphasize that was the first day
of what is now a four day event as opposed
to the first day last year when it was a
two day event. So when you prolong an event, you
lose that sense of urgency. People aren't necessarily shopping, people
aren't necessarily compelled.
Speaker 6 (15:40):
To act immediately.
Speaker 9 (15:41):
And you know when you had the Amazon executive on yesterday,
the key word there was engagement.
Speaker 6 (15:46):
You know, they're pleased with engagement.
Speaker 9 (15:48):
That means people are kind of poking around, but they're
not really buying, so everything kind of tracks. It's just
a question of you know, are people going to keep
buying through days three and four? And if they do,
Momentum said, you know, the four day event should yield
a lift compared to the two day event last year.
Speaker 6 (16:04):
But that's the big question.
Speaker 9 (16:05):
We'll see if Amazon's experiment is working today and tomorrow.
Speaker 2 (16:09):
Big bet.
Speaker 3 (16:10):
Spenser, Sopa, you stay on it, Thank you so much.
Now coming up, Gecko Robotics says twenty twenty five is
the Year of the robot. We speak with a CEO
live from some Vallely, Idaho. Next this is Bluemberg Tech.
(16:35):
Let's get back out to sun Valley, Idaho, where Ed
Ludlow is standing by with.
Speaker 2 (16:40):
A special guest.
Speaker 10 (16:41):
Ed.
Speaker 4 (16:42):
Yeah, let's talk about robotics and we're not talking humanoid rebulics.
Actually really interesting. Gecko Robotics and his co founder CEO
Jake Lucerian, There's a lot of attention on you. And
one of the reasons why is this super interesting business model.
The robots basically can assess the structural integrity of various facilities,
cool energy infrastructure. But actually a big part of your
(17:04):
business is AI and data. You sell them as a package.
Give us the basics on Gego Robotics and we'll go
from that.
Speaker 11 (17:10):
Yeah, thanks, Ed, Well, it's the company was started out
with this mission of diagnosing the health of the built world,
and so I started out of a college dorm with
that with that ambition emission, and what we were able to
understand and determine is we start with the foundation of
understanding the health of built structures, whether it's a dam
or a bridge, a vessel, of oil and gas refinery.
Speaker 6 (17:31):
Whatever the physical structure is.
Speaker 11 (17:32):
If you understand the health of that, you begin to
get really good at understanding how these facilities work and function.
And then what we've done is we've used all that
information data and the processing data from the facilities and
also the smart centors that we've built to be able
to create an operating platform to help those facilities run
in some cases twenty percent less cost for CAPEX and
(17:53):
running those facilities, but also operate fifteen percent more efficient
in some cases. And so the applications of the fence
and commercial are enabled because.
Speaker 6 (18:02):
We have an unfair data advantage.
Speaker 4 (18:05):
Someone that's never seen a Geho robotics robot scaling a wall,
explain what it looks like, how many of them are
out there and deployed in the real world, and what
kinds of commercial businesses are using them right now.
Speaker 11 (18:17):
Yeah, So the robots feed into the software operating platform
called Candilever, and the robots can look like the size
of a briefcase, climbing up a wall, climbing around, piping.
We also have drones, were walking robots, or smart sensors.
But all these tools are gathering information and data and
plugging into that central software.
Speaker 6 (18:36):
And then what we've been able to.
Speaker 11 (18:37):
See is that customers are actually asking can we put
our operational or our IoT sensor data into that platform
as well to derive better results, outcomes or decisions on
the ground. So what we've done is we've begun to
enable the folks on the ground to become not just
blue collar workers, but engineers because they can use the
LLLM that we have inside of Candile Leaver to be
(18:57):
able to understand what sort of decisions to make because
they have the data that allows them.
Speaker 6 (19:02):
To be superpowered in that way.
Speaker 4 (19:03):
So the energy sector is somewhere where you're very focused.
Next week, in your hometown, home state, there is a
gathering of political gathering with the energy industry. You work
with Legacy Energy to basically say here's how you could
be better, just explain how the type works.
Speaker 11 (19:19):
Yeah, So they're right in Pittsburgh next week, the President
and put on by Senator McCormick, is bringing some of
the biggest names in energy and AI to the city
of Pittsburgh.
Speaker 6 (19:29):
Where Get Goobrobotics is based.
Speaker 11 (19:32):
And so what it's going to be all about is
not just what's the state of things and what the
infrastructure for AI look like, But more importantly, the message
that I'm delivering and I'm speaking on in particular with
some of those leaders is how we should begin to
judge AI's impact and energy by how much AI can
actually create energy. Example, one of the biggest customers that
(19:53):
we have that helps to run five percent of the
power in the US, the five facilities that we've been
able to deploy our technology stack and operating platform at
have been able to show a huge increase to the
heat rate of these natural gas power plants. The extrapolated
impact is about ten gigawatts of more power without having
to create a single new power point. This is a
(20:14):
huge outcome as it relates to how do you combat
this issue. We don't have enough energy to power the
US In particular, to be the leader in artificial intelligence,
energy is a constraint, and so AI should be able
to unlock.
Speaker 6 (20:27):
Energy, not always the other way around.
Speaker 4 (20:29):
And what's so interesting is we're talking about some carbon
based legacy source of energy.
Speaker 6 (20:33):
There's a lot of.
Speaker 4 (20:34):
Focus on you right now because last month you hit
a one point two five billion dollar valuation.
Speaker 6 (20:39):
You want to scale.
Speaker 4 (20:40):
I know that you're kind of very focused on just
keep continuing growing where we see a follow on round.
Speaker 6 (20:46):
Just to keep that on the hardware side going.
Speaker 11 (20:49):
We really so the way that customers buy and work
with get Go is they buy the operating platform and
all the robots and centsives just by the rover.
Speaker 6 (20:57):
They don't buy any of that.
Speaker 11 (20:58):
It all feeds into the outcomes that we are deliver
with the operating platform, and so the biggest oil and
gas companies in the world will be unveiling things in
a couple quarters. Same with manufacturing facilities that are making
steel or making copper. We go into these old sectors
as Silicon Valley has very much forgotten about, and what
that means for us is that the scale and the
(21:18):
impact of pragmatic artificial intelligence robotics.
Speaker 6 (21:21):
Is hitting this inflection point.
Speaker 11 (21:23):
And you think about physical intelligence, and Jensen talks a
lot about this physical intelligence in these built environments, whether
it be in the defense sector that we work in
or it's in the industrial there is very little physical
intelligence and so we are the leading company as it
relates to creating that which enables a moat for our
artificial intelligence to be exceptionally better than those that are
(21:43):
coming with a software only solution. Meaning to the answer
to your question, scaling and increasing the amounts of fuel
to fire to that fire is something that we're going
to need to do the next time we.
Speaker 4 (21:56):
Have fifteen seconds. But your relationship with Nvidia, you're happy
with them.
Speaker 6 (22:00):
Yeah, I'm really happy with them. There have been great but.
Speaker 11 (22:05):
Looking at physical intelligence for these industrial sectors, there's a
lot more to go and so hopefully we can help out.
Speaker 4 (22:12):
Jake Luceerrian is the Gecko Robotics co founder and CEO
out here in some valley, and a lot of people
were interested in talking to this guy, Karro.
Speaker 3 (22:19):
I bet it really does feel like the year of
the robot. Great conversation. I thank you. Coming up, our
analysts say Apple should consider replacing Tim Cook. More on
that note. Next, this is bluebg Tech. Welcome back to
(22:42):
Blue beg Tech. Let's take a quick check on these
markets as we are open for trade, and.
Speaker 2 (22:46):
Then as that one hundred has been flitting.
Speaker 3 (22:48):
Between gains and losses. Earlier in trade, we went down
in the red.
Speaker 2 (22:51):
We're off of our lowers. We're off about a quarter
of a percent.
Speaker 3 (22:54):
But when you've got in video in the red, of
course that ples us a little bit lower. But more generally,
we're waiting on a big bond orction. A lot of
the yield movements been dictating what's happening in the stock
market right now, but we are still near all time highs.
Let's move on and think about what is very close
to an all time high. We hit in excess of
one hundred and twelve thousand dollars for bitcoin yesterday. We're
(23:15):
at sixtens a percent on the interday basis, one hundred
and eleven thousand, ninety five is where we trade, and
it really does seem that in mid the trade anxiety
there still is a risk asset of choice.
Speaker 2 (23:25):
Of course, apparently next week is crypto week according to
the House.
Speaker 3 (23:30):
In terms of Washington, we'll see how those changes in
rules and regulations continue to transpire. But let's get back
to some of the big tech news here. When it
comes to AI, Google's Gemina Gemini AI gets a new feature.
The company says it's the latest AI update, can now
produce eight second videos out of ordinary still pictures if
you're a paid user.
Speaker 2 (23:50):
For all, Bloomberg's Dana Woman is here to join us.
Speaker 3 (23:53):
This is really AI ultra AI pro users of Google
being able to keep up with what perhaps you're already
getting on an open AI or with a runway for example.
Speaker 12 (24:03):
Yes, this is a feature that will be available to
paid Gemini users, and it is a technology that Google
first launched at its Io developer conference back in May.
But this makes it available to a much broader swath
of users, allowing it, as you said, to better keep
pace with rivals.
Speaker 2 (24:18):
And it does have many rivals in this space, it
does have rivals.
Speaker 3 (24:21):
It does also have a stop start feeling to Google
because many would say, look, there were the reason they're
underlying research is why we've got general to AI in
the first place with the transformer perspective. But they had
this sort of inevitable feeling that they were losing pace
with rivals. Then they had the issue about whether or
not there was some generative AI image features that change
(24:43):
people's race, or there was a sort of a concern there.
Speaker 2 (24:47):
This is being replicated again in the video or no.
Speaker 12 (24:51):
Yes, So Google had a while ago pause certain generative
AI features just for that reason. But here it there
are some guardrails in place that said. We tested it
ourselves and the results were both fascinating and somewhat mixed.
For instance, reporter Natalie Lung, who is Asian, she did
(25:12):
include a photo of herself and the AI did change
her race in one of the videos without her asking it.
To other goofs, we did ask for a video of
someone breakdancing, we did not see any break dancing. Sadly,
the AI did follow other instructions we gave faithfully, but
not that one. So the results were mixed and we
do expect the technology to improve.
Speaker 3 (25:34):
And people do seem to really love what's been coming
out in terms of image generation coming from Google and Gemini.
Speaker 2 (25:40):
Writ large of late.
Speaker 3 (25:41):
But meanwhile Apple everyone still thinks is behind the curve
when it comes to AI.
Speaker 2 (25:45):
Is interesting.
Speaker 3 (25:46):
We've got a note out today with analysts really talking
about how maybe it's time to replace Tim Cook even
we were hearing some view that really the AI focus
is what needs to win here from a product perspective,
rather than having someone in charge who.
Speaker 2 (26:00):
Owns the supply chain. But the fierce talent for AI.
Speaker 3 (26:03):
The wall going on is an amazing scoop that Mark
German's been bringing us all week. And now we understand
two hundred million might have been parted with to.
Speaker 2 (26:11):
Nab the head of AI.
Speaker 12 (26:13):
A pay package worth more than two hundred million. And
I should say that this package does include not just
a base salary and assigning bonus, but stock options. And
there are a few things tied to that. One, someone
would have to stay with the company for a certain
period of time, so they wouldn't get all that money
if they left early.
Speaker 2 (26:28):
And it also is tied.
Speaker 12 (26:30):
To the performance of the stock. So it does assume
that the stock would reach certain levels, so if it doesn't,
someone wouldn't receive that quite that large of a pay
package in the end.
Speaker 3 (26:39):
And that is, of course all as we see transfer
of talent going from Apple to Meta or everyone to Meta.
As Mark Zuckerberg is all in on winning the superintelligence race.
Where are we though, with the feeling around Apple right
now as to ultimately whether it's business model is really
necessary to be the generative AI developer of choice, So
whether it ultimately can be the generator of AI products
(27:01):
user of choice, we'd just go to it to use
other AI offronts.
Speaker 12 (27:05):
So just given its overall stack performance and its market cap,
I really hesitate to call it struggling, but I think
it struggles in AI specifically have been well documented.
Speaker 2 (27:14):
I think it's.
Speaker 12 (27:15):
Efforts to have a new blockbuster hardware hit have been
well documented. I think we know that the Vision pro
has not been a great seller, even though reviewers and
sort of tech enthusiasts have praised it for being innovative.
And yes, the company does have a way to go,
ways to go in AI, and certainly the poaching is
(27:36):
yet another setback in that space.
Speaker 3 (27:39):
And maybe that's why hardware focusing with Apple Senior VP
of Hardware Engineering, John Turnus's why he's being singled out
by the note coming from Lightshed analysts as to whether
he'd be the right replacement for Tim Cook Data Wollman.
Fantastic to have her on all things Google and Apple. Now,
let's return to that talent war that we were just
talking of. Gray Oxweed Hoffmann calls multimillion dollar offers for
(28:02):
a talent quote not crazy when such investments are expected
to generate billions for a company. Now, he weighed in
on the escalating AI talent war between Apple and Meta
their own ed Ludlow over in Sun Valley.
Speaker 13 (28:14):
Well, so, essentially, one of the things that's good about
having an open marketplace of anything, including labor, is that
you essentially get multiple bidders, and that's what sets a price.
And so the fact that someone goes, I'm willing to
spend one hundred million dollars in order to get this talent.
Speaker 6 (28:28):
You think that's crazy.
Speaker 13 (28:29):
Why would it possibly do? That's like higher than the
vast majority of CEOs the construct, Well, if that one,
what you're judging is that one talent will potentially create
billions of dollars of value for you, whether it's in
a social network, whether it's in you know, kind of
you know, a search engine, anything else as a way
of doing it. And so if you think that, then
(28:52):
that's just a bold bet.
Speaker 4 (28:54):
What Sam Oltman, the CEO of open ai, said was
that that particular figure which he had accused Meta of
offering to some open ai staff, what he claimed was
it was crazy. That it sounds like it's not crazy.
Speaker 13 (29:05):
Really, well, I don't think it's crazy when we know
that AI is going to transform all businesses in all industry,
in the entire tech industry, and these are trillion dollar
companies et cetera at the high end and the whole
stack all the way through undown in.
Speaker 6 (29:19):
A billion today.
Speaker 4 (29:20):
For a brief moment, one of them was a four
trillion dolesmpany exactly, please continue.
Speaker 13 (29:23):
Yes, Nvidia, and so the you know so saying hey,
this piece of talent is a risk adjusted bet. That
that makes a difference in that that's just a risk bet.
You might say it's a dumb risk bet, but it's
not done to do if those are the stakes you're
playing for.
Speaker 3 (29:44):
Gray Lock partner and LinkedIn co founder Reid Hoffman with
ed Ludlow. Coming up, we speak with the CEO and
co founders Varda Space the latest series funding.
Speaker 2 (29:54):
Again, we're live from some valley. Stick with us. This
is really bad time.
Speaker 3 (30:09):
Welcome back to Bloomberg Tech. Now, Varda Space Industries has
announced it's raised one hundred and eighty seven million dollars
in the Series C funding, bringing its total capital raise
to three undred and twenty five million dollars. Then the
company aims to produce medicines in low Earth orbit. And
(30:29):
Ludlow is standing by in Sun Valley with Delian Asparov is,
co founder of Varda, and we're also joined by will Brewy,
Varda CEO. And will I start with you the money,
what will you use it for?
Speaker 14 (30:43):
Oh, we have big plans. The Series C raise is
great because it allows us to open up our biologics
lab down the street, and the entire purpose of that
is to prepare biologics formulations, so medicines for flight in
outer space. And the entire purpose of this is micro
gravity allows us to create new drug formulations that otherwise
(31:03):
would be impossible to create on Earth. That's the whole
purpose of Varda. And so with this Series C funding
we're able to build out a large lab space to
vet those drug formulations before flying them and create an
entire pipeline on the way to space.
Speaker 4 (31:19):
Delian, I just want to point something out, which is
that you were also a partner at Founder's Fund and
this has worked exactly the same way it always works.
You recuse yourself from the round and you represent VARDA,
but they participated as depeated t as an individual. Just
expand a little bit in what we were saying. You know,
one of the big changes that you guys have made
over the last yearies to start manufacturing yourself away from
(31:41):
Rocket Lab who used to do that as kind of
the third party for you. Why is that significant and
how does that ramp look?
Speaker 10 (31:47):
Yeah, I mean they were you know, sort of phenomenal
partner and were deeply grateful you know too them and
there they got us off the ground, you know, very quickly.
But as we were looking to you know, sort of
scale up our cadence of flights and ability to truly
control our supply chain, we had.
Speaker 6 (31:58):
To go more full stacking.
Speaker 10 (32:00):
We see this in a variety of different aerospace companies,
where as they scale up, they tend to go more
full stack over time. So that's going only to allow
us to you know, fly four times next year, and
then after that ideally start doubling basically every single year
and be able to control that doubling entirely in house.
Speaker 6 (32:14):
And then it just expands the.
Speaker 10 (32:15):
Capabilities that we have on the spacecraft in terms of
you know, biopharmaceutical production. By owning more of it, we're
able to expand the capabilities of what we can have
on board.
Speaker 4 (32:23):
Well, Dell's talking about all of the rampant launch getting
to orbit. The big beautiful benefit is zero gravity right
in the pharmaceuticals context. But then you've got to get
back down to Earth and that's been one of the
really exciting parts to track with Vada. The video in
particular explain how that process works and also how it's
(32:44):
going to go going forward.
Speaker 14 (32:46):
Yeah, so you hit the nail on the head. The
entire reason for going to space is for micro gravity.
It's just a fundamental force of physics that you can't
change on Earth. You know, we don't want to go
to space to have to make these new drugs. You
just have to because of the laws of physics, and
so our spacecraft allows us to do that quickly and cheaply,
and a lot and a part of that is once
we make those drugs in orbit in microgravity so that
(33:09):
they're different than they would be made if they were
on Earth, we have to bring them back and that's
certainly the most exciting part for the aerospace engineers here.
You know, my background is also airspace engineering. Like you said,
the video is phenomenal. We hit the atmosphere going mock
twenty five, So that's twenty five times the speed of sound,
twenty seven times the speed that you go on a
commercial airliner, and so you rip apart the atmosphere and
(33:31):
you create this beautiful plasma of a shooting star for
over a minute across the sky. It's beautiful on the
way back.
Speaker 3 (33:38):
The passion is real, Delian, the demand is real as well.
I'm interested in the DoD and the government perspective here
because we have seen so much more focus coming on
space and defense more broadly.
Speaker 10 (33:50):
Yeah, you know, sort of new Secretary of Defense in
the early days of the Trump administration came out with
what his national security priorities were going to be a
top three where nuclear weapons modernization, the development of next
generation hypersonic capabilities, and shipbuilding and you know, while VARDA
doesn't really you know, sort of build chips, you know,
we do you help the DoD you know, sort of
(34:11):
with the other two and have a variety of different
you know, sort of collaborations across groups from the Air Force, Navy,
and NASA. What they're primarily interested in is as well
mentioned as we come into the atmosphere we're going mock
twenty five. It's regime that typically is very difficult to
study on the ground and really understand what the flight
dynamics will be like. And so you can think about
the value proposition for VARDA to the Defense Department as
(34:33):
effectively like a hypersonic wind tunnel. They can attach net
new materials, wing shapes, sensors onto the vehicle as we're
coming in as a way to validate the subsystems performance
before they go and take that and integrate that into
the ultimate you know, sort of production system. And so
by enabling faster testing, more frequent testing, cheaper testing, it
(34:53):
allows us to expand on our capabilities and ideally both
catch up to you and surpass some of our new
europeer adversaries like China in Russia that have demonstrated some
pretty phenomenal hypersonic boost glide capabilities.
Speaker 3 (35:04):
I also think that the competition is somewhat closer to home,
or maybe it's a frenemy situation. Will you have got
such a broad experience, You've been flying the Dragon yourself
when you were working at SpaceX. You've also worked in banking,
you've founded your own venture fund in many ways, But
will I ask you now, what is your biggest headache?
(35:26):
Is it talent? Is it the race for talent? How
do you expand?
Speaker 9 (35:31):
Well?
Speaker 14 (35:31):
Believe it or not, there's not that much competition in
re entry because we're so much different than other commercial
re entry vehicles like SpaceX's Dragon or Boeing Starliner. We
are much smaller, we're off access in almost every way.
We're much smaller, we're higher cadence, we're not human rated,
and we're lower cost. So it's more of a complementary situation.
As far as headaches go, you know, a lot of
(35:53):
them are gone now thanks to a working spacecraft and
funding to grow. It's more about enthusiasm and exciting here.
I would say the challenge over the next couple of
years is really to scale up that cadence because as
we scale cadence. We begin to look more and more
just like a zero gravity oven. To the pharmaceutical industry,
they they quite frankly, don't care that we're going to
(36:16):
outer space. They only care that we can deliver new
drug formulations, and we just happen to have to do
that by going the space and back. So the more
often and the more frequent we can do that, the
more easily it is for us to integrate into the
pharmaceutical supply chain. And so that's the big challenge over
the next couple of years is increase cadence, fly more
spacecraft and fly more, develop more drugs on board.
Speaker 4 (36:39):
Dellian, The misconception maybe or misunderstanding about what Will's talking
about ramping the cadence is the cost consideration flying SpaceX
on transport emissions. Actually, my understanding is it's not that
great a cost. But would you talk a bit about
the relationship and how it works hitching a ride on
Falcon nine.
Speaker 10 (36:59):
Yeah, I mean to give you a send. It's like,
of our total all in mission costs that like space X,
like launch costs represent about twenty twenty five percent of
our costs.
Speaker 6 (37:05):
So it's not like the.
Speaker 10 (37:07):
You know, majority or anything. It's also you know, definitely
one of the larger line on it was, but there's
a lot of other things that go into operating either
of our business. Our business could not exist without the
you know sort of Falcon nine being reusable and that
you know sort of transporter program. If you think about
even you know, sort of five six years ago, your
only real option to get up to orbit was having
to buy a full dedicated rocket or find some way
to squeeze yourself onto a large geosats you know, sort
(37:29):
of launch and you know be on board, but you know,
find your own way to orbit. Now, these transporter missions,
they're doing them, I think on the order of three
times a year. They start to add on you know,
different inclinations in orbit. They have these bandwagon bandwagon missions
that go to you know, sort of polar orbit, and
so it's been a really you know, sort of critical
you know, sort of part of our ability to actually succeed.
Speaker 6 (37:47):
As a business.
Speaker 10 (37:47):
I think there's just this interesting you know of pattern
and technological history where when certain infrastructure layers get built up,
it enables these net new you know sort of business
opportunities right in some ways for a business like Uber
Airbnb to get built, you kind of needed mobile to
be very distributed and have the app store, and that's
what really enabled those business models. We think of, you know,
sort of the access to space and what SpaceX provides
us in the same way. It's a fundamental infrastructure layer
(38:10):
that enables our business and it wouldn't have been possible,
you know, sort of without them, and so we're thrilled
to have them, you know, you know sort of partners
and you know, have lots and lots of you know,
rockets out you know, sort of bud with the SpaceX.
I think, you know, we've booked out all the way through.
I think like Q three twenty seven or something like that.
Speaker 4 (38:24):
You know, will we recognize here on Bloomberg Tech there
are many in the space industry who have looked to
the defense apparatus of this country and the defense use
case to kind of start to get off the ground
because of grants and because of the availability of freedom
to move. But long term, what is the business model?
Who will your customers be when you're able to bring
(38:46):
payload back down to Earth. Pharmaceuticals developed in that zero
gravity environment.
Speaker 14 (38:52):
Yeah, it's it is a new technology and what's cool
going to the defense aspect and Deli and hinted towards us,
is that because reusable andmercial launch is a unique American asset,
VARDA is also a unique American asset on top of that,
so it's almost a leap frog above everyone else. And
so we have this cool armada of small, low cost,
(39:13):
high cadence re entry vehicles that can provide a service
continuously for the Department of Defense as a hypersonic wind
tunnel like Dally mentioned. So to answer your question, what
that means is now the Department of Defense can test
things like sensors, new materials for heat shields, they can
test flight computers and RF or radio frequency systems navigation systems.
(39:36):
So very similar to like a wind tunnel where it's
acting as infrastructure so that the Department of Defense can
advance new technologies faster and cheaper.
Speaker 2 (39:46):
Fascinating conversation.
Speaker 3 (39:48):
Thank you for all joining us, Will Brewy, Delian, Asmare
Evada and of course Ed stick.
Speaker 2 (39:53):
With us out there in some valley. We thank you.
Speaker 3 (39:55):
Now, coming up, we're going to hear from s Pat
Yukaro Bain Capital and his outlook the AI sector. This
is Bloomberg Tech let's head back to some value, where
(40:18):
the future of AI is one of the hot topics
among tech leaders. Steve Baruca of main Capital sat down
with Ed Ludlow to discuss well why the AI revolution
reminds him of the early days of the Internet and
actually talk about the potential for AI to transform productivity.
Speaker 15 (40:33):
It reminds me a lot of the end of the
period of kind of ninety eight, ninety nine, two thousand
and two thousand.
Speaker 6 (40:38):
And one in the Internet. That's worrying. Well, no, no, no, no.
Speaker 15 (40:43):
There were some fits and starts, and there'll be some
fits and starts in AI, but I think we're past
the almost past the fits and start stage. You've got
kind of nine global foundation model companies the Internet out there,
perplexity of open AI. We actually funded a startup from
MIT that has modeled the new AI system on a
(41:04):
worm's brain.
Speaker 6 (41:05):
Believe it or not?
Speaker 15 (41:06):
Do you know do you know how many neurons and
worms brain hasn't it?
Speaker 6 (41:08):
Educate me?
Speaker 15 (41:09):
Three hundred and five?
Speaker 4 (41:10):
Okay?
Speaker 15 (41:10):
How many? How many neurons do you think a human
brain hasn't many?
Speaker 6 (41:13):
Many more?
Speaker 15 (41:14):
Eighty six billion?
Speaker 6 (41:15):
Right? And you're smarter than most. Maybe you have eighty
seven billions kind of you.
Speaker 15 (41:18):
But they they these wonderful scientists that a MIT came
up with this model called the neural network model. Of
those nine companies, they're one of those nine companies out
there that are state of the art. They actually because
they they've used the neural network model. They use eighty
percent less GPUs to build the model, and therefore they'll
(41:38):
be perfect for edge devices like phones, because you don't
need the power and so a chet GPT model costs
about five hundred million dollars of GPU to load it.
You can do that for eighty five or ninety percent less,
and this model can sell on phones. And they're they're
just they're they're signing deals with edge device companies, you
know as we speak. So it's very exciting, and all
the costs are going down for the chips, for the
(41:59):
for the for the model loading. And as opposed to
the Internet, which there were some very few big winners,
I think there will be a lot of winners here
because there's going to be vertical models and foundational models,
and the vertical models will have the chuck full of expertise.
First example, a medical diagnosis model that focuses on disease
states and giving you a diagnosis. And some of these
(42:19):
models are proving much better than humans because it's a
game of elimination. Once you put the right data in,
it gives you the diagnosis. The same thing reading x rays. Right,
the models AI models reading x rays are now beating
the top humans in the field. So really it's really
we're only still in the early stages. It's going to
revolutionize productivity. It's going to have the same impact that
(42:40):
the Internet had twenty years ago, and it's going to
be a twenty year run. And that's why Navidia's values
so high, because they're in the cat bird seed of
the chips needed to power that revolution.
Speaker 3 (42:50):
Steve Padiuca a baying capital and the Boston Celtics. Now
that does it for this edition of Bloomberg Tech. Next week,
bloombag Tech will be live from Panasics new EV battery
plant in Kansas. Tune in for an exclusive interview with
the Panasonic North America CEO on Monday.
Speaker 2 (43:08):
And meanwhile, len't forget to check out our podcast.
Speaker 3 (43:10):
We'll find an on the terminal as well as online
on Apple, Spotify, and iHeart from the Sun Valley, Idaho,
or from New York.
Speaker 2 (43:17):
This is Bloomberg Tech
Popular Podcasts
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by audiochuck Media Company.
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.