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 live
from coast to coast with Caroline Hide in New York
and Vla Loow in San Francisco.
Speaker 2 (00:22):
This is Bloomberg Tech.
Speaker 3 (00:23):
Coming up, Open Ai releases GPT five. We'll discuss how
the rollout is going in the enterprise opportunity with Open
Ai COO Brad Lightcap, plus the race for AI data centers.
It heats up with Softbak making moves in Ohio, while
Meto announces financing for its expansion in Louisiana. And we
keep on these tech earnings. We'll speak with CEOs and Twilio,
(00:44):
Chime and Acami. The first we return to the public markets.
We're a new record high for the Nasdaq one hundred,
so we managed to push on higher, but I'm looking
at individual stocks that help contribute to the moves. Tesla
actually a decent points perspective higher.
Speaker 2 (00:58):
We're up three point three percent.
Speaker 3 (01:00):
Breaking news come Foring one and only Ed Ludlow about
Dojo the future of the supercomputer. Shares actually managed to
take high, even though we see that they're wrapping that
down focusing more on using third parties. We'll get to
that story a little bit later. I'm looking at Intel
off by two ten percent. Have been trading higher on
the pushback from the CEO saying he has committed the
US and National Security. Will get to that in a
(01:21):
moment as well, But first we want to really dive
into the world.
Speaker 2 (01:25):
Of data centers.
Speaker 3 (01:26):
SoftBank taking ownership of fox Conn's EV plant in Ohio
in order to kick SAT, the Japanese company's five hundred
billion dollar Stargate data center project with open Ai and Oracle.
Bloomberg's Peter Elstrom joins us for Ashleson reporting that Ed
Ludlow was really driving before he got on his flight
to the United Kingdom. Peter, it's an interesting deal. Basically
(01:48):
soft Bank kind of wanting to get fox Conn more
integrated with its manufacturing part of this.
Speaker 4 (01:55):
Yeah, this is kind of a mysterious deal. So fox
Cohn made this announcement a couple days ago. They said
they were selling this facility to an entity called Crescent Dune,
that nobody knew who it was. So we've discovered that
in fact, at SoftBank they had been partners in the past.
Soft Bank is going to own the facility, and then
soft Bank and fox Con together are going to work
(02:15):
on these AI developments. They're probably going to build AI
servers in that plant to be able to use throughout
the Stargate venture that they're supposed to be building in
the US. As we've talked about a few times before,
open Ai and soft Bank announced that Stargate venture with
President Trump alongside them, talking about spending one hundred billion
dollars immediately at the beginning of the year or five
(02:35):
hundred billion dollars over the long term. But it's been
kind of slow to roll out in a number of
different cases. So this deal is aimed at giving them
a bit of a kickstart so they can move ahead
with some of those projects, be able to build the
AI servers, and then actually be able to build the
data centers over the long term.
Speaker 3 (02:51):
It's interesting this pivot from a US macro perspective from
evs into AI data centers. But open ai has really
been pushing had almost on its own hair. It's been
going into Norway for its infrastructure, has been going to
the Middle East and in the United States with Oracle
and almost leaving well it felt like soft Bank behind
in EMO right.
Speaker 4 (03:12):
Open Ai has a lot of power here to be
able to pick their partners, as we've talked about before. Also,
they are building this big data center in Avelene, Texas
that's been going quite well, quite aggressively. Open Ai course
was very closely affiliated with Microsoft in the early days.
It put in a bunch of money early on. It
had been kind of their cloud computing supportive partner for
(03:34):
a long period of time. Now they've also talked about
working with soft Bank Oracles. They're hurt too, so they're
sort of choosing their dance partners as it suits their need.
At this point, soft Bank, of course wants to be
able to move more aggressively into the AI field. Soft
Bank shares are actually on a tear right now. They're
an all time high. Masieo she Son is certainly getting
a bit of a tailwind because they put that money
(03:56):
into Open Ai at a lower valuation. Now they've been
able to ride that up quite a bit. So I
think there is some optimism that SoftBank is getting its
footing here, is being able to make some progress in
the AI feel bit Masio Sisnis wanted for.
Speaker 5 (04:08):
A long time.
Speaker 3 (04:08):
Great context. Peter Elstrom, we thank you so much. And look,
speaking of Stargate tune in later this hour for an
interview with the Open AI COO Brad Lightcap on GPT.
Speaker 2 (04:18):
Five and much more.
Speaker 3 (04:19):
But in more data center news, it keeps on coming.
Meta has announced financing for a twenty nine billion dollar
data center expansion in rural Louisiana. Now it's actually selected
Pimco and Blue Out Capital to lead theb in equity
financing here.
Speaker 2 (04:32):
That's all according to sources.
Speaker 3 (04:34):
Representatives of these three companies actually declined to comment. Let's
get back to also elsewhere in chips Intel, the saga continues.
After the US President, of course, called for at CEO's
resignation over conflicts of interest, Intel CEO lit Bhutan replied
in a letter to staff posted on the company's website,
where he says he got the full backing of the
board and what they are working on with the administration
(04:57):
to quote ensure they have the facts. Mags in King
rings us with more facts. What does he think there
is misinformation about Ian?
Speaker 6 (05:06):
Yeah, I mean we saw that the Trump social media posting,
and this came on the back of what a senator
had done and sending a letter to Intel's chairman saying, hey, look,
all of these things have happened in Bhutan's past. Are
we sure he's the right person to be running Intel?
And there was a lot of sort of cluging of
(05:29):
facts together and in accusations made, which the President then
responded to, and finally late last night we had Lib
Bhutan directly addressing that and saying, look, there's a lot
of misinformation out there. You know, we look at the
facts and don't worry. I'm still going to be the
CEO of Intel, and I've got the boards backing.
Speaker 3 (05:49):
Yeah, he's concerned about his past roles at Walden International
venture capital firm that does make bets in China and
based in Singapore. Is a founder of it. Also Kina's
Design Systems, which was CEO.
Speaker 2 (05:59):
Of This sort of should have been known.
Speaker 3 (06:01):
He was on the board of Intel for a long
time himself well known. What do you think this means
for their manufacturing presence in the United States? Because I
almost got a hint that maybe he is reing committing
to that here.
Speaker 6 (06:15):
Yeah, I mean, this is arguably I mean, first of all,
there was nothing new about these accusations against him. Twenty
twenty three we had the House Committee looking into this,
so nothing really new here, But You're right. Perhaps the
trigger point had come a week earlier during earnings when
they discussed the possibility of not moving forward with the
(06:35):
latest manufacturing technology, of cutting back on spending, maybe not
building Ohio, maybe not building other plants. And this is
not what the Trump administration wants to hear when they
want to be talking about what we're bringing to America
and what we're doing as soft Bank has just played
too with the news that we broke there, So it
(06:56):
was a kind of counternews cycle there that they didn't
really think they were more perhaps talking about what investors
wanted to hear, which is we're going to get back
to profitability quicker, we're going to spend less, and that
kind of unfortunately appears to have blown up on them
in the political arena and they had to sort of
address that late last night in king.
Speaker 3 (07:15):
I'm sure the story will continue, Thank you very much. Now,
let's bring in an investor perspective. Kim Forrest going to
break this all down. Your chief investment officer are Boca
Capital Partners.
Speaker 2 (07:24):
You do have exposure to Intel.
Speaker 3 (07:26):
I'm interested in your perspective on what they now need
to do in terms of turning the narrative.
Speaker 7 (07:33):
Sure, well, this is just kind of a distraction the
whole CEO thing he's in. I'm sure the board thoroughly
looked at him, so I'm going to move on from that.
But he is there because the last several CEOs have
just kind of worn down the bright shining starre that
is Intel, and he is trying to find his way
(07:55):
and trying to find where Intel should play. And I
think it's pretty clear that he's looking at what they
can do right now. And let's just say they're not
up to the task of making a Blackwell competitor, but
they do have some good technologies that are used in AI,
mostly around packaging. And if I can feel your eyes
glazing over, don't do that, because packaging is the thing
(08:17):
that makes really low, you know, small scale chips work
because you have to make sure the electrons are staying
where they're supposed to be going.
Speaker 8 (08:27):
So packaging is a big thing.
Speaker 7 (08:29):
But he has a very large task ahead of him
of making Intel relevant because let's say it's been drifting
off for the past ten years.
Speaker 3 (08:38):
Our eyes glazing over about parts of ship making never Kim,
I am interested though more broadly about Therefore, this overwhelming
drive towards data center capacity and what that does for
tailwinds for the chip makers that you look at that
we had a This week has been a long one
in the world of chips. You've got AMD numbers which
are still showing growth, it's not quite at the pace
(09:00):
that people wanted to see. You've also had, of cause,
this threat of one hundred percent tariffs on semiconductors coming
into the United States, but then huge carve outs. Where
do we focus attention when you want exposure to the sector?
Speaker 7 (09:14):
Well, I would have to unbelievably agree that AI is
the thing that's driving it forward now, and it's a
very narrow slice of AI. We're talking about the data center.
What I'm really interested in is, yes, we're going to
have large language models, large reasoning models. I'm much more
interested in smaller models that companies are going to actually
(09:35):
use to become more productive. The large language models are
getting all the love, but I'm not sure that they
generate quite enough interest in direct measurable productivity.
Speaker 5 (09:47):
Right.
Speaker 7 (09:47):
They're being played around with at this point, But that's
where all the money's going into, is building data centers
that support large language and large reasoning models.
Speaker 5 (09:59):
Will that go on for ever?
Speaker 9 (10:00):
Know.
Speaker 7 (10:01):
I think what's more interesting is the further we get
away from the data centers out to where the information
is going to be used. That's when more companies are
going to be involved in tel may have its place
out there, because what we really want are not necessarily
walking around robots, but robots that do things that are
(10:22):
under the control of the data center kind of computing.
So that is a very complicated world, and more semiconductors
are going to be needed.
Speaker 2 (10:32):
Okay, so spread that love a little bit.
Speaker 3 (10:34):
It's so interesting that this comes hot on the heels
of the scoop from Ludlow talking.
Speaker 2 (10:38):
About how Teser itself is riding down.
Speaker 3 (10:41):
It's focused on hardware and it's own chip making with
Dojo Supercomputer and looking to others to provide the hardware
going forward, more reliant.
Speaker 2 (10:49):
On video and on Samsung kim.
Speaker 3 (10:51):
Where else do we like then, other than all eyes
on in video and MD to a certain extent.
Speaker 7 (10:57):
Well, I really love Micron, and because for many reasons,
Micron has the hot dram that goes on the Blackwell
chip and probably on AMD's chip, and that's memory that's
going to move data in and out of a chip, right,
So that's thing one. But thing two is we are
generating so much information and it's not on spinning discs anymore.
(11:22):
It's more on nand chips. So I love that area.
I think it's an unloved area and really unexplored. So
I think we could never underestimate the amount of storage
that AI uses and how much we're going to have
to store just to you know, get that data that
we want to have looked at by AI. Has to
(11:44):
live somewhere. I'm thinking against nand devices.
Speaker 3 (11:47):
Micron currently up about five percent on the day. Also,
of course t SMC we're putting that twenty six percent
growth of a month of July. We're just on a
tear in many areas. Kim Forest of Boka Capital Partners.
It's always great to check in with you. Happy weekend.
Now coming out, we're going to take a look at Twilio.
Look at the shares currently off by eighteen percent. They
reported earnings yesterday. We go deep with the CEO.
Speaker 2 (12:07):
This is Bloomberg Tech.
Speaker 3 (12:15):
Let's talk about cloud communications platform Twilio reporting earnings after
the market closed yesterday. The report showed an acceleration in
revenue growth that the company's gross margin earnings for share
do seem to be a concern to some analysts. Let's
bring in Twilio CEO Couzema, Ship Chandler joining us. Now
I'm looking at your shares, they are down a lot.
Just give your perspective on the gross margin.
Speaker 10 (12:38):
Well, I would say, in general, we delivered a solid quarter.
I mean it was a quarter of accelerated growth, record
non gap income from operations, free cash flow. And I
think what we're finding is is that at a time
when AI is kind of this generational opportunity, we're finding
some of those tailwinds get ascribed to Twilio. I understand
(12:58):
the gross margin dynamic. It's not necessarily a new phenomenon
for Twilio, but it hasn't impeded our ability to get
after operating profits and generate free cash. Importantly, on free
cash we raised our guidance for the year and so
we feel pretty good about the outlook.
Speaker 3 (13:14):
So for now, is the opportunity more on growth and
revenue and the AI tailwinds.
Speaker 2 (13:19):
As you call it.
Speaker 3 (13:19):
Because Jefferies, for example, putting out saying there are big
hopes you're going to be a big beneficiary of AI, adoption.
How is that playing out for you, because many now
potentially question it.
Speaker 10 (13:31):
Yeah, I do think we'll be a big beneficiary of AI.
What I don't want folks to take away from the
print is that somehow we're returning to kind of growth
at all costs dynamic. I think quite the contrary. We're
running the company with as much focus and discipline as
we ever have. We happen to be at a time
in which, number one, we're able to take share for
(13:51):
a number of our competitors, and we're doing that in
a price disciplined way. And I think the second thing
is is that given that the AI is a generation opportunity,
we do see an opportunity as well to make some
short term investments that we think will drive some medium
term benefits. And I think that's the right thing to do,
and so I'm not fixated on a particular quarter. I
(14:12):
think over the course of time that'll play out in
our favor and certainly be a good thing for investors.
Speaker 3 (14:19):
What is that investment? Then, you detailed or platform vision
almost in signal your event you're thinking about audio in particular,
I mean you are the way in which I communicate
with LIFT or with any app and many as ways
that's going to move to voice.
Speaker 2 (14:33):
Do you make acquisitions therefore, I.
Speaker 10 (14:35):
Think acquisitions are something that we think about. I wouldn't
necessarily read anything into that. I mean, we obviously have
some flexibility in terms of what we can do with
our capital. But I think more importantly the investment in
the short term is around voice AI for example. I
mean voice AI is where so many different AI startups
are really launching right. I mean, a lot of these
(14:56):
use cases that you see are voice based. We have
the benefit of delivering incredible voice infrastructure that startups can
get going with almost immediately. That's kind of on the
low end, and then on the high end, you know,
there are fully orchestrated experiences that we drive, combining our
communications capabilities across a number of channels, utilizing data as
(15:20):
well as utilizing AI, plugging into all manner of different lms,
and I think that's going to be exciting for us
from a growth standpoint, and it's going to drive profitability
and cash over time.
Speaker 3 (15:32):
I'm looking at your one year share price move, it's
up sixty three percent, so that puts what today's move
is in perspective. Because aim but you talk about making
perhaps your acquisition yourself, or having the flexibility to do so.
Someone as are saying that maybe you're a target. Have
you had any inbound would you ever consider that?
Speaker 10 (15:54):
I mean, I wouldn't comment on on M and A
one way or the other. I think we're always going
to do things that are in the best interests of
our share owners. I can tell you with conviction that
we're very focused on running a great business. We think
that we have all of the necessary ingredients to create
tremendous value for our customers. And as you alluded to
(16:15):
a moment ago, over the last year or so, we've
shown that we can demonstrate focus, discipline, operating rigor in
the way that we're running the company. Investors have benefited
from those attributes. And again I don't get fixated on
a particular quarter. I think over periods of time this
is going to play out very much in investors favor,
(16:36):
and I think they're going to read really substantial benefits.
Speaker 3 (16:38):
And where are your opportunities for new wins? In particular,
when you're getting new customers on because of the voice
offerings and the communication offerings, is it geographical in nature
that you look around for expansion.
Speaker 10 (16:50):
I think there are international opportunities, no doubt, but I
would say in the near term. You know, what's really
starting to happen is we're We've always kind of been
like a product led growth company, and so the vast
majority of our customers, they're actually self starters. They come
to our console, they get going, they love the usability
and the features that are our platform gives them, and
(17:13):
so that's really the way that people get started. So
we've been putting investment into that making as simple as
possible for our customers to get going, and right now,
at least in the immediate term, that's really played out
with respect to voice AI. Again, so many different voice
AI companies that are launching they're choosing Twilio. We're the
best known brand in the space for super simple and
(17:35):
easy to use, and there's a ton of value that
our customers can get out of that opportunity. On the
enterprise side, I do think what we're finding is is
that they want the kind of broader offering. Right they
want our communications capabilities, but they also want to use
data and they want to plug in the LMS that
they're excited about to be able to drive both cost
(17:57):
and revenue opportunities.
Speaker 3 (17:58):
Because aymanship Chandla Vilios a year. We thank you so
much for your time today after the earnings and coming
up more of them more earnings. Chime releases this in
orgeral reporters. A public company and investors a little disappointed.
Speaker 2 (18:10):
This is a blue bag tech. Chime It put out.
Speaker 3 (18:18):
Its first earning report since going public two months ago.
Revenue for the fintech company was strong, but customer growth
it failed to reach the top end of expectations. Let's
talk it through. Chime CEO Chris Spritt is with us. Chris,
solid thirty percent revenue growth. You've got gross margin of
eighty seven percent, but it seems to be this growth
in active members twenty three percent that just failed to
(18:39):
meet the bar for some What do you make of it?
Speaker 8 (18:42):
Well, thanks for having me. It's great to be here.
Speaker 9 (18:44):
I'm so proud to be reporting on an awesome Q
one that we've had our Q two in our first
quarter as a public company, and being on the show
to share more. We had a great quarter. We like
you said, we beat consensus on that top line. We
grew thirty seven percent. We saw outsized results exceeding expectation
(19:05):
across the board on growth of on the revenue side,
active member growth, and our success on adjusted EBITA, and
we raised our expectations for the year as well on
the call as it relates to the active member growth.
You know, it was a great quarter for us. I
think naturally as a new company, new issuance, I think
(19:28):
investors are just starting to understand the nuances of our
business a bit, and we do have a seasonal business,
so on a quarter to quarter basis, sequentially, the growth
for people that maybe don't understand the seasonality the business
might have been disappointing. But the way that you really
need to look at the business is how we're growing
on a year over year basis. So in Q one
(19:48):
we grew actives twenty three percent. In Q two again
we grew actives twenty three percent year over year, accelerating
that rate.
Speaker 8 (19:56):
So we actually had an outstanding quarter.
Speaker 9 (19:58):
And I think it's just natural now that need to
make sure that we educate investors on that some of
these details of the business, such as the seasonal elements
of it.
Speaker 3 (20:06):
And maybe investors were looking to competitors in the space,
not direct competitors, but certainly in the fintech space so
far just had such strength. Do you think perhaps it
was what others are doing in the market, and how
can you show that you're winning as much and building
as much.
Speaker 9 (20:21):
Well, you know, I think we have a different business
model than a lot of the other FinTechs out there
in that we're really focused on serving everyday consumers and
banking them in a primary account capacity. In other words,
people signing up for Chime and using us as their
primary direct deposit account. And we're really, at the day,
at the end of the day, a payments driven business,
(20:43):
not a lending business. And I think that's a real
differentiation that's important for investors to understand about the core
of our business model. You know, credit and lending is
sort of a mid teens percentage of our revenue. Were
predominantly a payment driven business, which I think is quite
a bit different. And you know, we love these relationships
because it allows us to develop long term partnerships with
(21:06):
our members and help them in the areas that matter
most around avoiding fees, around building credit, getting some short
term liquidity, around the edges, but also just helping them
get into a healthy rhythm of savings and moving their
financial progress.
Speaker 3 (21:20):
Ahead, and many of them want to do that far
more efficiently using generative AI.
Speaker 2 (21:24):
That's something you've been into weaving, you've.
Speaker 3 (21:26):
Got the voice, what for example, Chris, why more can
you lean into that?
Speaker 9 (21:30):
We're so excited about the opportunity in AI, and I
just think, you know, what, are an exciting time to
run a technology company. We feel like we're just in
the early stages of a new industrial revolution. If you
think about, you know, the original Industrial Revolution, it was
all about new sources of energy that created it and
unlocked so many new opportunities, new industries and so forth.
(21:52):
I kind of think of llms as the same sort
of new form of energy that. Sure, the LLLM companies
are incredible, amazing businesusinesses. We're all using them every day.
But the opportunities for companies like Chime to leverage AI
and move it into the product experience, the operations is
just enormous. We're already serving over seventy percent of our
(22:13):
customer interactions with generative AI, and not only is the
cost saving exciting, but the actual outcomes are even more
exciting because we're seeing a significant increase in the satisfaction
levels of consumers getting the answers they want.
Speaker 8 (22:27):
Quickly through an AI.
Speaker 9 (22:29):
Of course, you can always talk to a live customer
service rep. But we're really excited about the opportunities not
just on the service side, but even using AI as
a think of like a digital partner that kind of
helping you manage your data day, financial life.
Speaker 3 (22:44):
I have to leave it there, fascinating. Chris bret Chime, CEO.
We thank you for your time coming out next one
of those lms, Open AI. The CEO joins us, this
is Broomberg Tech. We want to welcome our global audience
across Bloomberg Radio and television Open AI. It's released GPT five.
(23:04):
It's most advanced model yet. The company says it offers
key improvements in major areas like reliability, accuracy, and there's
the strongest generator of AI model yet incoding and writing
and health.
Speaker 2 (23:15):
For more.
Speaker 3 (23:15):
We bring in Brad Light Open AI COO and this
feels so primed for enterprise adoption. Brad, when I think
of writing, when I think of coding, what is the
opportunity there?
Speaker 5 (23:27):
Yeah, well, good morning, thanks for having me.
Speaker 11 (23:30):
GVT five is a significant step forward in a few
different domains.
Speaker 5 (23:34):
So you mentioned coding.
Speaker 11 (23:35):
You mentioned writing in health for consumers and medical professionals,
and we think that opportunity unlocks an amazing set of
things in the enterprise that now become possible.
Speaker 5 (23:46):
It's a much more reliable model.
Speaker 11 (23:47):
So it's better at things like calling tools, it's better
at things like structured thinking and reasoning, problem solving. And
what we see in the enterprise is when you make
these core capabilities better, the number of use cases enterprises
can adopt these models for increases and so coding being significant.
Speaker 5 (24:04):
It really is the language of.
Speaker 11 (24:05):
Computers, and that was a significant area of demand for
us when we were talking to customers about what they
wanted in this model.
Speaker 3 (24:12):
I mean, PhD level is what many are calling it,
well what Sam is calling it, and I'm sure yourself.
What's interesting is when you've got seven hundred million weekly
users of chatchipt, how much is that a funnel a
read across into enterprise. How much you could get the
inbound because ultimately the people in the old workforce are
already using it.
Speaker 11 (24:30):
Well, really early on when we launched chat Gipt, what
we found is I think, you know, a number of
months after we launched it, I think something like ninety
two percent of the Fortune five hundred we're actively using chatchipt,
or people at ninety two percent of the Fortune five
hundred were actively using it, and so it was very
obvious for us we needed to go build a work
product because I think chatchapt as a product is as
(24:52):
useful in an enterprise environment, in a work environment as
it is in your personal life. It's an amazing companion
if you do anything from marketing to software engineering, to
data analysis and research. And I think there was a
lot of organic adoption when people discovered the tool, realizing
that it could make people much better at their jobs
(25:12):
and able to do more. And so we really leaned
in with that and we're trying to build the best
product we can.
Speaker 3 (25:17):
It's like one tenth of the planet using chat gpt Brad.
But I'm interested in some analysis that Meno Ventures has done.
They've analyzed the LLLM market, particularly in the enterprise space,
and they've just tried to push back saying, look, you
lost market share.
Speaker 2 (25:30):
Open Ai went from fifty percent in the.
Speaker 3 (25:32):
Enterprise market share down to twenty five percent, and funnily enough,
the company that they back, which is Anthropic, took the lead.
What do you say to those of statistics, is it
something you're seeing within your own numbers.
Speaker 5 (25:44):
It's hard to measure these things.
Speaker 11 (25:45):
You know, you can find you can find measurements that
say the opposite. But what we really focus on is
value for customers, Like we've got to deliver the absolute
best models and then the absolute best products for developers,
for startups, for enterprises large and small, and that's that's
been our focus. I think you know, our API is
a good example of where we've really invested lately. We've
got over four million developers now actively using the API
(26:09):
every day to build new products. We support thousands and
thousands of startups that are building with us that we
work deeply with on trying to improve our product so
that they can ultimately build a better product. And in
the enterprise, I think you know, the demand that we
see there is really unabated. We grew chat Gypt enterprise
seats from three million seats to five million seats in
(26:29):
a matter of two months, and so that growth is
accelerating and we're just starting to scratch the surface, I
think on the impact that we can have both for
developers and for enterprises. So we see it as a
long game, and you know we're here to just do
our best for customers.
Speaker 2 (26:44):
For that, BOYD, you need infrastructure.
Speaker 3 (26:47):
Tell us a little bit about the costs of training
this model and how you're looking to expand with Stargate,
the project we've just been talking about how SoftBank's been
teaming up with Foxcon, for example, to take over an
ohio ev plant. How is that continuing to meet your
demands or not meet them as the case may be.
Speaker 11 (27:05):
Well, yeah, we've seen demand for AI increase at just
a torrid pace. Obviously, at the root of that is
getting right the infrastructure equation, and we think ultimately that's
going to be a critical input to the US and
its allies being competitive in this area. And so Stargate
for US was a five hundred billion dollar project to
(27:26):
invest here in the United States to build AI infrastructure
for open AI and ultimately for the country. We're working
with a lot of great partners on that project and
hope to bring in more and I think, you know,
that's just the beginning. We're going to continue it to
invest aggressively. We've always found ourselves somewhat on the wrong
side of the demand curve for AI, you know, despite
(27:48):
the investment, the significant investment to date, and so for
as long as we see demand, we're going to continue
to invest aggressively. AI is interesting in that the more
you invest and the more you make it available, the
more you make it, you know, cost cost approachable for
enterprises and for consumers, the more people want to use it.
So it's an amazing trend and we'll continue to invest
behind it.
Speaker 2 (28:07):
We're speaking with Brad Lightcap of open Ai.
Speaker 3 (28:10):
The CEO for our radio and TV audience is Brad,
how has the rollout ultimately been. Do you think because
so many people wanted to use the app that maybe
we hit limits quicker than some anticipated.
Speaker 5 (28:22):
Well, we're trying our best to keep up with demand.
Speaker 11 (28:25):
Serving infrastructure at scale at seven hundred million users and
then millions of developers and many billions of tokens per
minute that we process is not for the faint of heart. Thankfully,
I don't have to do that part of it. But
we're doing our best to make sure the rollout is
successful and we're hoping that by the end of the
week here everyone gets access.
Speaker 3 (28:44):
Your job description is more about building the enterprise relationships
and partnerships. I also think about the partnership you have
with Microsoft and Sati and adelays out there really talking
about integrating already and how excited he was for the product.
Speaker 2 (28:55):
But there's a tension in the relationship there.
Speaker 3 (28:58):
I'm interested in what you think the progress being made.
Speaker 2 (29:01):
Sam Alman was on Networks.
Speaker 3 (29:03):
Talking about progress being made with a relationship going forward. Microsoft,
can you give us a timeline about when you think
a deal will be done in the future of how
they interact with your product and more broadly, how much
ownership they continue to have in the business as a
full profit one.
Speaker 11 (29:18):
Yeah, Well, we feel really positive about the relationship with
Microsoft and they've they've been a great partner throughout the
history of open Ai. They've been with us from the
beginning really since before chatch Ept obviously have been a
huge infrastructure partner for us with Azure.
Speaker 5 (29:33):
And so we continue expect that to continue. We see
we see no future.
Speaker 11 (29:37):
That you know of open Ai that doesn't include Microsoft
in a significant way. We've got to work on what
that future looks like together. We're in that process with
them right now. We feel very good about it, but
we think also ultimately there's you know, the world is
really big and the demand for these systems and these
models in the enterprise and consumer is significant. And so
they represent not only an infrastructure partner for us, but
(29:59):
a a partner to be able to help distribute and
bring the benefits of the technology to the world given
the size of their footprint, so you know, more to
work through there.
Speaker 5 (30:07):
But we feel good about it.
Speaker 11 (30:08):
And like I said, I think you know, when we're
standing at the finish line of all this, they'll be
there with us.
Speaker 3 (30:13):
Meanwhile, the stuff of sulimanover at Microsoft is busy in
the tussle for talent, so to plenty of other rivals
that we understand, Mark Zuckerberg busy. And what's interesting is
while you is one of those long tenured employees and
staff over at open Ai remain committed because of innovations
such as this. But what about the liquidity that we're
talking about bringing to some of your well people that
(30:36):
you work alongside.
Speaker 2 (30:37):
How is that going?
Speaker 3 (30:38):
We understand they might be even a five hundred billion
dollar valuation involved in what is a secondary sale of
your shares to the likes of Thrive Capital.
Speaker 11 (30:47):
Yeah, well, nothing there to share here, but we continue
to see very healthy demand on the investors side for
wanting to be I think part of the Open Eye
journey and mission, and we're very grateful for that. And
on the talent side, look, I think you know, Opening
Eye was founded as a nonprofit. It was founded as
a mission driven company to be able to build you know,
general intelligence that's beneficial for all of humanity, and we
(31:09):
haven't strayed from that mission. I think ultimately that's what
attracts talent is people want to work for a project
that's bigger than themselves and something that's going to be
impactful for us and for you know, for for for
humanity and our species. And so I think that's the
thing that ultimately attracts people to where they work. Obviously,
like it's a competitive market and we continue to compete.
But at the end of the day, when we ask
(31:30):
people what it is that keeps them at Opening It
it's the mission.
Speaker 2 (31:33):
The mission.
Speaker 3 (31:34):
At the moment, you've got something that's generally intelligence, but
it's not artificial general intelligence.
Speaker 2 (31:39):
Brad, when do you get there?
Speaker 11 (31:41):
Briefly, you know, I've sworn off making predictions in AI.
It's too hard, the curves are too steep. But I think,
you know, we feel really good about the rate of progress.
GBD five is a great representation of how we start
to make progress on little things, you know, things like,
for example, being able to have the model dynamically reason
and decide how much it wants to think about the
(32:03):
problem that you ask it to solve. That's something that
we do natively as humans that previously our models couldn't do.
So it's these little steps forward that we think accumulated
and ultimately get us to something that will be truly remarkable.
Speaker 3 (32:15):
Radi cap talking about the latest GPT five. We thank
you so much, Coeo of open Ai. Now let's check
back on these markets, because we are up three point
three percent over the course of five days, and then
that's that one hundred where are a new record high.
Speaker 2 (32:30):
Folks. We have had one of.
Speaker 3 (32:32):
The longest weeks and tech in terms of newsflow, whether
it's been about tariffs on chips, whether it's.
Speaker 2 (32:36):
Been the onslaught of earnings, which.
Speaker 3 (32:38):
We're going to be talking about time and time again.
Currently up three point three percent. Want to look about
what's happening in the rest of the industry and indeed
the market more broadly. Bitcoins down by percentage point, got
a bit of a risk off tone.
Speaker 2 (32:49):
You have to keep an eye what's happening with.
Speaker 3 (32:50):
Gold spot though, because talking of tariffs. While it's not
just semiconductors that's been in the eye the storm or pharmaceuticals,
but gold bars as well, huge volatility on the course
of the day. Meanwhile, coming up, we're going to be
speaking with akam I CEO to Tom Layton, I'm going
to go back to other earnings within this trade desk
is off by thirty nine percent. It's earnings underwhelmed significantly.
Amazon competition Pinterest off after its earnings, but Insta can't
(33:13):
trading higher Exipedia after Airbnb managed to underwhelm Expedia up
more than four percent. We can bring you Akami next though.
Tom Layton joins us the CEO, this is Blue Megtech.
(33:37):
Keep an eye on Akamai shares today, somewhat volatile after
posting second quarter results.
Speaker 2 (33:42):
That across the board in general beat estimates.
Speaker 3 (33:45):
But you have to read between the lines as to
where there was perhaps a bit of a pullback or
a miss when it came to Compute, for example, but
therefore you forecast that was raised. Tom Layton, Akami CEO
joins us now to demystify for us, Tom, because it's
interesting you have Hybisan really talking about lots of moving
parts and ultimately they see fairly cloudy still of what
(34:06):
the ultimate positioning is of the business.
Speaker 2 (34:08):
Can you tell it.
Speaker 3 (34:09):
You're really going in on well, cloud infrastructure here, not
just content delivery, not just security.
Speaker 12 (34:16):
Oh yeah, security majority of our revenue Cloud infrastructure services.
Speaker 5 (34:22):
My goodness.
Speaker 12 (34:23):
We are up thirty percent year every year, and we
think that's going to accelerate. So by the end of
this year will be up forty percent plus year over year.
Tremendous market opportunity. There, strong tailwinds of course from jen Ai.
So I think very exciting results for Akamai. You know,
a strong beat on the quarter, very profitable earnings per
(34:45):
share up nine percent and a dollar and seventy three cents,
and I think a very good outlook, raising guidance for
the year.
Speaker 3 (34:52):
So maybe just a little bit of a pause for investors.
I mean, the shows are still down over the year
or twelve months.
Speaker 2 (34:59):
Tom. I'm sort of interested therefore on what they're not getting.
Speaker 3 (35:02):
What is it about the opportunity in, for example, cloud infrastructure.
Why take on the hyperscalers, Well, it's.
Speaker 12 (35:09):
A tremendous market opportunity. Our customers have asked us to
do that. You know, we with our distributed platform, which
is unique in the marketplace. We're in seven hundred and
fifty cities. With our points a presence, we can get
enterprise compute instances. They're containers closer to the end users,
and that gives you lower latency and better performance and
(35:32):
for a lot of applications that matters. Also, we could
do it at a lower price point. In fact, one
of the hyperscalers is an early adopter of our new
managed container service. They want to have their compute instances
and hundreds of cities around the world, which is something
all the Akamai can do.
Speaker 2 (35:49):
Interesting.
Speaker 3 (35:49):
I mean, maybe people are looking at what drove some
of the revenue beat and they're looking at FX, but.
Speaker 2 (35:54):
They're also looking at TikTok inclusion. Have you decided that.
Speaker 3 (35:57):
The risk of being managed just off the table and
you're going to really be startling talking about this client
once again?
Speaker 8 (36:03):
Oh, there's always risk.
Speaker 12 (36:06):
Though, it seems that from what we're hearing from public
sources that you know, a deal has basically been structured
and you know there's obviously caught up in trade negotiations,
but that at this point it seems like the ban
has been postponed several times, and so we're not seeing
anything that suggests that that that's going to change in
(36:28):
the near future.
Speaker 2 (36:29):
You do that content delivery for them.
Speaker 3 (36:31):
It's interesting that security is where a lot of your growth,
and as you say, the most part of your revenue
has been coming from. What insecurity is driving that growth?
Where are people coming to you?
Speaker 12 (36:42):
Yeah, we have the market leading solutions for web app firewall,
for bought management, for API security, for stopping ransomware with segmentation,
and of course the attack rates have been going way up,
the penetrations have been going way up, and we have
a solution.
Speaker 8 (37:01):
It stops the damage caused by those attacks.
Speaker 12 (37:04):
So if the ransomware gets in, we identify it quickly,
we don't let it spread, and that protects our customers
so they don't, you know, have these huge costs associated
with being shut down by ransomware.
Speaker 3 (37:16):
Of course, general to AI in many ways, making the
field of security ever more evolving and nimble is what's crucial. Tom.
What I'm also so interested in is the way in
which you're thinking about education here in the United States.
Skills in particular and upskilling. This is something that the
US government is calling for a lot, and I'm particularly
in the era of general to AI. How are you
seeing the skill set and the pool of talent? Is
(37:39):
it good enough for you?
Speaker 8 (37:41):
It's going to change with Jenai.
Speaker 12 (37:43):
We're seeing tremendous efficiencies across most job types at Okhami,
so that our employees can do more in the same
amount of time by leveraging Jenai. And I think you'll
see training. We're doing training ourselves so that employees will
change maybe how they do certain things and they'll be
(38:06):
more efficient by learning how to use the Genai tools.
I think it's really very exciting and it probably is
going to be a sea change over time and education.
Speaker 3 (38:14):
Tominayton Avakami was great to have you on after your numbers.
Thank you very much for joining the show.
Speaker 2 (38:19):
Stay well.
Speaker 3 (38:20):
Meanwhile, coming up Tesla, well, it is disbanding the team
that was meant to give it the computing muscle in
the AI race. As an Ed Ludlow scoop, we'll get
behind it with you next.
Speaker 2 (38:30):
This is bluembg Tech.
Speaker 3 (38:43):
For decades, Samsung was the world's leading memory chip makup
but the tech chant was recently surpassed by smaller rival
sk Heinez, So how quickly can it reassert itself in
the AI race of Blue Beg Original's team to a
deep dive into this.
Speaker 13 (38:57):
When you're out shopping for seven megawatts offshore winds turbin,
you think Samsung, right, Well, probably not. The brand is
much more famous for its phones and flat screens, but
arguably the most important drivers of South Korea's biggest family
business are these things. And for the first time in decades,
the company's got a massive problem here.
Speaker 9 (39:16):
Samsung is under assault from all sides and they are
getting stretched.
Speaker 6 (39:21):
Then Samsung is in this position of playing catchup, and
that's difficult to do if you don't have growing revenue,
if you don't have expanding profitability.
Speaker 14 (39:29):
And unfortunately the ship profits tumble, big disappointment. The semiconducted
division reporting operating profit at four hundred billion one totally
missed analyst projections for two point seventy three trillion.
Speaker 13 (39:42):
Samson shares of whipsword in value since their twenty twenty
peak and tanked in twenty twenty four, while long standing
rival sk Heinez searched the crisis is easy to miss
amid all this.
Speaker 5 (39:57):
The Ultra experience is ready two.
Speaker 13 (40:01):
A four, But behind the cats and choreography, Samsung's hidden
struggle threatens its national identity amongst a population that relies
on it like no other.
Speaker 2 (40:13):
It is probably the most influential company in.
Speaker 13 (40:15):
Korea because of the impuortance to the economy, and maintaining
that influence may hinge on a piece of technology you
probably own but have never even seen.
Speaker 3 (40:30):
Catch the full episode at Bloomberg dot com, on YouTube
or on the terminal. Now, one company depending more on
Samsung for its future AI chips is Tesla, and we've
got some updated reporting from our own ed Ludlow. Tesla
is disbanding its Dojo supercomputer unit and its leader, Peter Bannon,
is leaving the company as it depends more on third
party hardware suppliers. It's all according to sources. Now, the team,
(40:51):
as we told you yesterday, had lost workers to Density AI.
It's a competing startup founded by former Dojo executives.
Speaker 2 (40:57):
Let's get more of.
Speaker 3 (40:58):
Bloomberg's Max Chaffkin, busy on his travels and Max, we
turned to you as elon inc Helmer, what do you
make of this move away? From making its own hardware
in the AI chip space.
Speaker 15 (41:12):
I mean, it's definitely a setback because Tesla, of course,
if somewhat famously, likes to make everything that's kind.
Speaker 8 (41:20):
Of Elon Musk's whole thing.
Speaker 15 (41:21):
He's tried to make a vertically integrated auto manufacturer. That
makes Tesla very unusual, and also has been trying, you know,
obviously desperately to pivot towards AI. So this is a setback.
I will say it's a setback that I think people
saw coming to some extent. In twenty twenty four, during
an Ernie's call, Musk sort of started to tamp down
(41:44):
expectations said they were going to rely on two paths
for training its AI models.
Speaker 8 (41:50):
One would be this dojo thing.
Speaker 15 (41:52):
The other would would rely on Nvidia chips, the same
very expensive chips that everyone is essentially using to train
their models. So it is it's a setback, but maybe
but I don't think it's devastating. I think what is
maybe more troubling is just losing talent. Right we're in
the middle of this AI talent war. Tesla historically has
(42:16):
been very competitive. Of course, the stock you know, for
for a while was just exploding and it has been
fairly stagnant over the last couple of years.
Speaker 8 (42:25):
And I think you're, you know, you're seeing that, right.
Speaker 15 (42:27):
There are startups out there and these kind of very
large privately held companies like open Ai and Thenthropic that
are just throwing huge sums of money at developers, and you're,
you know, you're seeing Tesla clearly like struggle a little
bit to keep up and retain.
Speaker 3 (42:44):
And does CI now forming as a new startup to
take on the world of hardware and AI max Briefly,
I mean Elon responded saying that it doesn't make sense
have two different avenues for chips, and AC.
Speaker 2 (42:55):
Fifteen and A sixteen they're going to continue with.
Speaker 3 (42:58):
But more broadly, this was thinking of the way it
set upself apart for the future of supercomputing being vertically integrated.
Speaker 2 (43:04):
Maugazani thought would be five hundred million dollars added to
the market cap.
Speaker 15 (43:08):
Yeah, and all those estimates, of course look look silly,
and I think you have to ask yourself there are
lots of estimates like that around Tesla and it's AI ambitions,
few of which are really realized yet, and it does
make you wonder like, should we revisit some of those
other assumptions, you know, people talking about valuations in the trillions.
Speaker 8 (43:31):
You know. Look, Elon Musk said this was a moonshot.
Speaker 15 (43:34):
I think it was kind of a moonshot because the
truth is Nvidia dominates this market, the market for training chips,
and Elon Musk took a crack but wasn't able to
get there.
Speaker 3 (43:45):
Bloomberg's Max Chafkin. Tune in to Elon ink It's weekly.
We thank you so much. Now that does it for
this edition of Bloomberg Tech.
Speaker 2 (43:51):
Quick check in on the markets for you.
Speaker 3 (43:53):
We're off by eight ten percent on bitcoin, digital gold,
but real gold flat. But boy, we can potentially see
some tariffs on all bars affecting Switzerland in particular.
Speaker 2 (44:02):
Keep an eye on that story.
Speaker 3 (44:02):
Nasat one hundred are by eight ten percent, a new
record high.
Speaker 2 (44:06):
Folks, check out our podcast. You can find one in
the terminal. This is Bluemberg Tech