June 20, 2025 • 42 mins
Bloomberg’s Caroline Hyde discusses SoftBank founder Masayoshi Son’s pitch to build a trillion-dollar AI industrial complex with TSMC in Arizona. Plus, the global robotaxi market heats up as Baidu looks to expand internationally and Tesla prepares to launch its service in Austin. And Meta rolls out new smart-glasses in a partnership with Oakley.
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 (00:00):
Bloomberg Audio Studios, Podcasts, radio news. Bloomberg Tech is live
from coast to coast with Caroline Hyde in New York
and ever Though in San Francisco.
Speaker 2 (00:21):
This is Bloomberg Tech coming up one trillion dollars for
a USAI hub, the latest big bet being pitched by SoftBank,
band of Masayosti's son. We dig into the reality of
the numbers. Plus Meta launches three hundred and ninety nine
dollars oakly AI glasses aimed at athletes and thrill seekers.
We discussed the surprise hit smart glasses and China's baid Eyes,
(00:42):
a Robotaxi expansion in Asia. We discuss the next steps
for a the adoption. But we're also eyeing what's happening
in terms of Middle East tensions. I want to move
into those Middle Eastern tensions as well, because look, more broadly,
we are seeing Israeli officials warning that Iran is tapping
into private, securely focused cameras to gather some information, particular
(01:05):
intelligence against adversaries.
Speaker 3 (01:08):
Now it's going to be a.
Speaker 2 (01:09):
Key focus that we've seen repeated with hamas with Russia
in other conflicts. Bloomberg's Mike Shephard joins US now on
what is some deep investigation, mic.
Speaker 4 (01:21):
Cara.
Speaker 5 (01:21):
This is something that highlights a longstanding vulnerability with surveillance cameras,
which have become ubiquitous around the world. When we look
in our homes, even you see those doorbell cameras, and
you go to any office or department store, there are
cameras everywhere, even on the streets. Yet the data that
(01:42):
is being gathered, the images, the footage is not necessarily
as secure. We've seen instances several years ago of hacks
of startups that had just reams of data from tens
of thousands of surveillance cameras. But there is also a
geopolitical concern as well, and we see the highlighted with
Israel warning citizens to either turn the cameras off or
(02:04):
at the very least update the passwords.
Speaker 3 (02:07):
Okay.
Speaker 2 (02:08):
So a key call to their own citizens at the
moment and one that will follow throughout. But from real
war tensions to the global trade war once more, Mike,
there's some really interesting news breaking very recently across the
BlueBag terminal coming from the Wall Street Journal talking about
how the US is preparing action targeting Allies chip plants
(02:28):
in China. In particular, there's talk of ending waivers for
the likes of TSMCSK, Heinek, Samsung that actually make chips
in China. Right, it seems to be trying to make
it harder for global chip makers basically to operate in China.
Speaker 5 (02:41):
Well to a degree, yes, and the idea from the
Trump administration is to what they see it as according
to the journals reporting, level the playing field when it
comes to requirements for rare earths to be shipped to
the US. In other words, the same degree of licensing
required for that would soon be perhaps required for bringing
(03:01):
American technology, American chip making equipment into China or at
least updated or repaired in the Chinese market. Remember that TSMCSK, Heinix,
and Samsung produce a lot of chips for the American
market in China, but what they use is US equipment
and equipment from overseas. Use equipment would now need a license.
(03:25):
Rather than just getting a blanket waiver where you can
bring it in without having to check every time with
the US government, you would need that level of permission. Now,
this is not pertaining to the most advanced, the highest
end AI chips, for instance, this is a lot of
the legacy level semiconductors, the older generation chips that go
into automobiles, consumer electronics, and other lower end things. So
(03:47):
it's not exactly targeting the upper end that we've seen
the US government do, for instance by barring AMD in
video of course, from selling their most advanced AI chips. Instead,
this is taking aim in a different end of production
in China of certain kinds of chips.
Speaker 2 (04:04):
Mike Sheffin breaking it all down for us around the world,
thank you so much. Now, look, there is another top
story that we're watching. SoftBank founder Massio Shei's son is
planning a trillion dollar industrial complex in Arizona. It's to
build robots. Further, it's AI endeavors. Look TSMC potentially playing
a key role here. Lumberg's Peter Elstrom joins us to
break it down.
Speaker 3 (04:23):
One trillion is a lot, One trillion is a lot, Caroline.
Speaker 6 (04:28):
Yeah, So Masio Sheisana has another big bowld vision. What
he's pitching companies like TSMC and Samsung on now is
building out these industrial parks and places like Arizona in
the United States. The ideal is to kind of appeal
to the Trump administration and try to bring some advanced
manufacturing back to the United States. Now. There are a
few related projects going on right now, and it's not
(04:50):
as crazy as it may sound at first. Blush TSMC,
for example, is trying to bring in a lot of
its suppliers into Arizona into the United States so that
it can produce more of the advanced chips within the country.
Samsung is doing a very similar thing in Texas. And
remember soft Bank spent years investing in startups and they
have hundreds of startups that would be able to put
(05:12):
facilities there. And the focus here is going to be
on AI and robotics. Those are two big areas for
their investments. So you could also see some soft Bank
backed companies move into these kinds of ventures.
Speaker 2 (05:23):
Okay, so it doesn't sound as crazy as it seems
as how you articulate it, Peter, But we know that
actually promising money and deploying it at two very different things.
And actually Stargate, for example, already what massiochisn has been
wooing President Trump with and in line with Oracle and
Open Ai to build infrastructure in the US that's been
hard to deploy right.
Speaker 6 (05:45):
Right, that's only five hundred billion in Carolina.
Speaker 3 (05:47):
That's much more reasonable.
Speaker 6 (05:49):
Yeah, they talked about the Stargate venture with open AI
in particular, where they're going to build out data centers
in the United States. That began as one hundred billion
dollar venture and then they said, hey, this could get
as big as five hundred billion dollars. But as we've reported,
they ran into some problems there in terms of the financing.
Some of the terriff issues that the Trump administration was
talking about ran into their ability to be able to
(06:11):
get banks on board, to be able to get the
money to be able to get some of these projects underway.
So that's going a little bit slow at this point.
And you're exactly right when it comes to this adventure
that we're talking about now, the industrial parks that also
would need financing. It's not clear how quickly they'd be
able to pull that together.
Speaker 5 (06:26):
Now.
Speaker 6 (06:26):
Selt Bank has had preliminary conversations with state officials, Howard
Lutnick's office at the Commerce Department in particular, so they
have begun some of these conversations. It's not clear exactly
how far along they are. It will depend on whether
they get some of the tax incentives that they want.
Some of the other corporate support to determine whether they're
going to be able to proceed.
Speaker 2 (06:44):
Peter astrom Me, thank you for bringing us the story.
Let's dig in now with Jordan Klin. Mazuoko America's managing director,
actually wrote about this soft bank promise in his morning note.
Speaker 3 (06:54):
You're joining us.
Speaker 2 (06:55):
Now so much to cover you think this is perhaps
more of a marketing employee.
Speaker 7 (07:00):
Yeah, that's that's where the you know, the skeptic in
me thinks it's it's a little bit more of that.
I mean, you never want to underestimate soft bank. And
when it comes to AI and robotics, I think there's
a lot of interest, particularly from the capital markets, but
you know, the size and the scale seem a little
bit trumped up, so to speak, I think, and the
(07:20):
timing I would question as well. Like and that's why
I don't think you're seeing the stocks move at all,
because investors are just kind of, you know, somewhat immune
to these sort of numbers that keep you know, going
up and up, but they're not necessarily you know, believing
that that's going to impact the stocks anytime soon.
Speaker 2 (07:36):
I love that you say we're becoming immune to these
sorts of numbers. That sort of number was actually articulated
by Samumon a little bit ago with our own memory.
Speaker 3 (07:43):
Chang, just take and listen to Sam's point here.
Speaker 8 (07:47):
The interesting question is if we had if we knew
how to get a trillion dollars right now, if we don't,
would we be able to deploy that profitably in the
next few years? And I'm not sure about that, but
I feel confident we can like make five hundred million
value back.
Speaker 2 (08:02):
Samonman is saying he's confident he can get half of
your money back if you give him a trillion. So, ultimately,
are these the right bets to be betting on as
an investor in the likely winners? Would you still allocate
to TSMC if indeed a trillion is talked about in
Arizona for example?
Speaker 7 (08:21):
Well, I love TSM personally, and I really like a
lot of these AI compute names that are going to
facilitate a lot of this expansion into gen Ai. But honestly,
I don't think you need the trillion dollar headlines to
make these stocks look attractive or compelling. Everything in the
near term sounds like the demand for AI tokens and
(08:44):
inferencing is rising and Video is doing a great job
of improving the production and output of their new Blackwell
server racks and the cloud Hyperscalers who buy all this
equipment all seem keen to increase investment as opposed to
limited or low it down. So yeah, it's great when
soft Bank and others talk big picture plans down the road,
(09:07):
but I think there's just so much demand and interest
near term that investors should be focused, you know, more
on that and less on you know, big scale projects
that may or may not come about because as your
you know, prior analysts mentioned before this piece, you know,
there's a lot of regulation in government involvement that would
(09:29):
need to come to pass even before you can even entertain.
Speaker 3 (09:33):
Some of this talking of government involvement.
Speaker 2 (09:35):
That's perhaps what's in the eye of the storm for
investors right now.
Speaker 3 (09:38):
I'm seeing nvidios of.
Speaker 2 (09:39):
Near session lows at the moment and others that perhaps
make things in China, particularly chips, because there's Wall Street
Journal reporting saying that the US might make it even
harder to use US equipment in China, the likes of
tsmcsk Henich, Samsung, for example. How much of a headwind
is this for the overall semiconductor space.
Speaker 7 (09:58):
Well, it's a great question, and I think what you're
seeing here is these reports get leaked or information from
the administration gets leaked to the press. They put out
stories that talk about potential restrictions that the US could
put on our intellectual property, our leading edge equipment and
technology that would go either to our allies or anyone
(10:21):
related to China. And in my opinion, this is similar
to the reports you heard on EDA software being blocked
or restricted from China when we were trying to get
those rare earth and magnet imports turned back on. So
what I think is happening is you have two big
Korean memory companies that make memory inside China, and the
(10:45):
US wants to get leverage in the current ongoing trade
negotiations with South Korea or maybe other Asian nations who
would be exposed to these rules, and so they throw
this out as well, we might restrict the ability to
get those leading edge tools, which would basically make those
fabs inefficient. It wouldn't shut them down, but if you
(11:07):
can't get the best latest equipment, it basically puts you
at a competitive disadvantage versus companies like Micron, So I
think it's more posturing and leverage, but it would be
a real It definitely doesn't help the sentiment across SEMIS.
And that's to answer your question, because anytime you bring
up government restrictions in China, it just puts some downward
(11:30):
pressure on SEMIS. But again, I don't necessarily think this
is what is going to happen anytime soon.
Speaker 3 (11:35):
It's just a risk, just a risk, but people act
on those risks.
Speaker 2 (11:38):
And as we see the socks Semi conduct To index
now trading off by one point seven percent, and it's
kla it's land research in particular in the headwinds. Can
you give us a sense of who in despite these
sorts of environment you just have to allocate to because well,
the positives just so outweigh. Does it still remain in
video all cases TSMC at all cases? Are they're more
(12:00):
nuanced plays to be had, because it feels as though
we've been ringing that one for a while.
Speaker 7 (12:06):
Well, you mentioned the names that would be most at
risk would be the equipment company's kla ASML, potentially AMAT
and LAMB, and those stocks have been going up in
the face of a lot of concern that the US
could put further restrictions on those equipment and that those
tools going into China. So a lot of the people
(12:27):
I speak to are already cautious on the equipment stocks.
But what I actually think it definitely really won't impact
Nvidia in my opinion TSM. Honestly, again, demand exceed supply.
They're going to do fine. And the other name that
I would look at, as you mentioned, like who could
benefit is Micron. So Micron is the other large dram
(12:50):
nand flash memory supplier, and they don't have big factories
inside China. They're the only US memory company, so they're
like our leading champ and we want them to survive
and do well. And obviously they're going to try to
move more of their manufacturing into the United States, but
they would be at an advantage if the US was
(13:10):
to restrict Samsung and Heinez from getting those needed tools
in those Chinese facilities. And that's the one that's reporting
next week. That's the one I think that's going to
sound super bullish on demand for memory in AI, and
that's the one I think that could really be the
surprise upside winner.
Speaker 2 (13:28):
Upside already up twenty eight percent so far. This month
they came out with real vision for allocating to R
and D here in America. But last week we'll wait
for their earnings a little bit later. As you say,
thank you so much Jordan Klin. Always great to have
him on from Mazuha coming up as Tesla prepares to
laun James robotaxi service. Well, the global competition in autonomous
(13:51):
cab space is heating.
Speaker 3 (13:52):
Up A's next as Bluebertech.
Speaker 2 (14:03):
Chinese technology giant Baidu well it is planning to expand
its robo taxi service in Singapore and Malaysia as as
early as this year, according to people familiar with the matter.
Now the company is in discussions with potential partners in
the two markets. By news Apollogo, it's about one thousand
self driving vehicles worldwide, mostly in China.
Speaker 9 (14:21):
Now.
Speaker 2 (14:22):
This newsicalse comes as Elon Musks Tesla prepares to launch
its robotaxi services in Austin this weekend. Let's get more
on the autonomous market. Karako comments here with us. She's
a professor of Civil and Transport engineering over at the
University of Texas at Austin.
Speaker 3 (14:35):
Her work includes research on the.
Speaker 2 (14:36):
Impact of bringing self driving cars to the roads and
kara for anyone who is worried about the safety involved here,
your data shows, look, we should be pursuing this path.
Speaker 7 (14:47):
Yeah.
Speaker 10 (14:47):
So, in an early paper, just looking at the kinds
of technologies that auto manufacturers already have, we expected about
an eighty five percent crash reduction and fatality reduction and
crash cost reduction, which is a lot of money because
Americans already lose about three thousand dollars a year per
capita every year to crash costs. And what we've already
(15:10):
seen is Weimo is hitting that target on the streets
of San Francisco.
Speaker 11 (15:15):
Phoenix, and an La, where.
Speaker 10 (15:17):
It's been comparing itself to the other drivers that are
on those exact same streets.
Speaker 2 (15:22):
Of course, cost to vehicle owners, cost of life as well,
car is something that just everyone is so focused on,
but there has to be regulation that keeps up. China
is going at this full speed, so is Asian. We're
just talking about how bydo is rolling out across Singapore Malaysia.
Are we keeping pace with how quickly the innovation could
change here in the US?
Speaker 10 (15:41):
Well, I think so thanks to Tesla, but of course
they also have a gigafactory in China, and so China
may have access to all of that intellectual property and
be able to mimic it quickly to catch up with
Tesla with a Chinese based company.
Speaker 2 (15:57):
Okay, so thanks to Tesla, thanks to Weimo, thanks Zooks,
who we had on earlier in the week. There is
plenty of companies that are showing that it can be done.
Just remind us of also not the cost implications from
an insurance perspective, but more broadly, how much easier, how
much more cost efficient will it be to make these vehicles,
Because look, zuks is saying it's going to make ten
thousand in the next few years or so per year
(16:20):
in their new facility. But all of these means they're
pretty expensive pieces of equipment as it stands.
Speaker 10 (16:26):
Right so people are estimating that the Waimo vehicles right now,
which are on a very expensive Jaguar Ipace, which is
a suggested retail price of seventy five thousand, I think,
so add another seventy five thousand to get all of
the lighter and everything else installed. But if you can
install it all directly on the assembly line in one
coherent piece, that the cost comes down a bit, and
(16:48):
of course weimo's not in the business of actually manufacturing
the vehicles and Tesla is and if you go to
the Tesla website, I think they're offering the technology for
eight thousand dollars up front, plus another one hundred dollars
a month, and so that starts to sound pretty reasonable.
And we do expect those costs to come down long
term to maybe under five thousand dollars off the top.
(17:09):
And if you have to have a monthly check fee
or something, so they're updating, you know, to new standards
each month, and there might be another fifty dollars fee
on top of that.
Speaker 2 (17:18):
I suppose talk to us therefore about how FSD seems
to work at the moment, because we hear on both
sides the equation. Many love it, Many feel that it
is incredibly efficient. Others say there are still so many holes,
so many ways in which that it underperforms. We cannot
roll it out yet in the United States.
Speaker 11 (17:38):
Well, it's already out there.
Speaker 10 (17:39):
So a lot of us have friends and neighbors and
people just alongside us in the traffic stream and you
may not notice because there is somebody in that driver's seat,
but they are in FSD mode, and so I believe
they've driven billions, billions of miles already, whereas the Waymouth
fleet has gone about sixty million miles of rider only.
Speaker 11 (18:01):
So nobody in the driver's seat.
Speaker 10 (18:02):
And that's the big jump here, is that they're they're
tweaking the algorithms so that the car will be able
to move without somebody in that driver's seat and will
do a lot of the parking, which they're not doing
right now for the owners of FSD who paid that
eight thousand dollars plus one hundred dollars a month, and
of course have to be in that driver's seat legally,
(18:24):
have to be a licensed driver and have to keep
their eyes open, you know, much of the time. But
when I was in that driver's seat, I did nothing.
It was very precise, It was very impressive on very
complex network here, the routes that we were taking, but
it was broad daylight. And the one thing I noticed is,
you know, it does go above the posted speed limit
(18:45):
to keep up with the traffic around it, which the
Waymo vehicles don't do.
Speaker 11 (18:49):
And it also came, you know, a lot.
Speaker 10 (18:51):
It was much more confident around pedestrians and scooters and
cyclists than I would normally be as a driver myself.
Speaker 11 (18:58):
So that's where I'm a little bit nervous. If somebody
turns around mid stream.
Speaker 10 (19:02):
You know, you've seen the images of scooters falling in
front of.
Speaker 11 (19:06):
A way MO.
Speaker 10 (19:07):
That would be tough, So there will be some really
hard breaking or maneuvering if they're going to avoid those
kinds of crashes.
Speaker 11 (19:14):
So that's that's something I'd be watching out for.
Speaker 2 (19:16):
And when it comes to well keeping in the speed limit,
there are some settings you can adjustice there. Kara, We
thank you Kara Kocheman. She's the professor at the University
of Texas At Austin. Meta is rolling out new aiglasses
with Oakley that are targeted to athletes. The new versions,
(19:36):
which start at three hundred and ninety nine dollars, include
improved video quality features like water resistance. Man Neat Singh
are in house athlete and Bloomberg intelligence.
Speaker 3 (19:46):
Writer talks to us a little bit.
Speaker 2 (19:47):
Look it's funny the way the story is written by
Mark German. He says people are using them also for
thrill seeking. They're getting on roller coasters and using them.
Why are these Oakley versions going to be better?
Speaker 12 (19:58):
I mean, it just turns out to be a better
phone factor. When you think about, you know, the ray
Bang glasses compared to the VR headset, I would say
it's a much easier phone factor to use, you know,
as an AI glass. But at the same time, I
think they also realized that the VR headset weren't really
(20:20):
as impressive to use from a gross margin perspective, So
I think the ray Band glasses and the oak Ley
glasses probably are better from a gross margin standpoint. And look,
in the case of Meta, we know they'll be losing
about twenty billion dollars this year on reality labs, and
there is no line of sight in terms of when
(20:42):
this is going to get profitable or you know, when
the economies of scale will kick in. So from that
standpoint as well, I think it's a better pivot for
them to focus on their glasses compared to the VR headset,
where they were doing all the manufacturing themselves from the
ground up.
Speaker 2 (21:00):
They stick to display lists, or at least they enhance
the offering on display lists. How many of these do
you think they're actually going to be able to sell overall?
Because you run the numbers with your latest big Yes, I.
Speaker 12 (21:10):
Have So look, when it comes to Apple Watch, they
are about forty million units sold. Air pods are about
you know, er take seventy to eighty million in a year.
Raybang glasses were about one million, and we are headsets
were about seven million last year. So adding Oakley, which
has the same parent company we're talking about, maybe it
(21:32):
scales from one to two million this year, but it
still doesn't match the scale of an Apple simply because
the ecosystem that Apple has with the content and the
app store, it's so hard to replicate for a meta,
and that's been their challenge, even though they have invested
so much. So to my mind, ecosystem is where a
meta lacks, and it'll be a hard thing to scale
(21:54):
up even with the Oakly glasses.
Speaker 3 (21:56):
And I'm saying, I'm Brett meg Intelligence. We thank you.
Welcome back to Blueberg Tech and Karen Hydeen, New York,
not in London. Let's get a pict check on these
markets because as.
Speaker 2 (22:11):
We stand, we are currently down about one point three
percent and we've got anxiety building, whether or not it's
about Middle East tensions, whether or not it's about yet
further restrictions on international chip makers making chips with US
equipment in China. That was a key story coming from
the Wall Street Journal put some pressure on the chip
stocks in particular or off by one and a quarter percent,
as we're also just embracing ourselves with some.
Speaker 3 (22:32):
Volatility into the end of.
Speaker 2 (22:33):
The day's close that of course, as we anticipate triple witching.
Speaker 3 (22:37):
But DOLLI what resolutely remains high AMD.
Speaker 2 (22:40):
We're up one and a half percent on the day,
we were up ten percent on the week AMD. Lisa
sou that discussion about where they go next with their
own generative AI chips really ringing through when she presented
last week. We've continued on the higher side for this
particular stock meta though off by one point four percent.
Speaker 3 (22:55):
Interesting as they say, look.
Speaker 2 (22:57):
We've got the latest and greatest Oakley AI glasses. Market
less convinced as it stams. But let's talk about another
key top corporate story, Microsoft planning to cut thousands of jobs,
particularly in sales. The company tries to trim its workforce,
I mean heavy spending on artificial intelligence that's bringing bluebags
proty Ford. This has been sort of a story that
we get a drip feed on who is it hitting
(23:17):
this time?
Speaker 9 (23:19):
A lot of people see this, and they get confused because
they say, Microsoft world's most valuable company. Most days, they're
making all this money, why do they keep firing people?
A big part of it is because, you know, you
got to start spending eighty billion dollars a year on
AI data centers, and you don't want your margins to
move so much. And now we're seeing that there's this
kind of recurring rolling layoff. They laid off about six
(23:41):
thousand people a month or two ago. Now it's the
end of the fiscal year. And now who turns it
to get fired salespeople? And so it's something has happened
a bit in the social contract of these large companies
where these rolling layoffs have become much more common. And
it's not just Microsoft.
Speaker 2 (23:56):
I'm talking of contracts. There's a contract between open ai
and Microsoft, and that seems to be getting ever more
contentious at the moment. Bring us the latest prodi on
just how open ai is competing against the hand that feeds.
Speaker 9 (24:08):
Yeah, that was a nice transition. I mean, you know,
Microsoft inked this contract with open ai back in twenty
nineteen at the point a kind of you know, fledgling
AI startup, which is now one of the most powerful
companies in the world with a lot of leverage, and
they probably want a lot of things that are not
in that contours or that initial contract. And so you know,
essentially open ai wants to restructure itself. Who needs to
(24:31):
sign off on that. It's largest shareholder Microsoft. So there's
been a kind of drip drip of information around where
these disagreements are, where Microsoft wants to hold control, where
open ai wants to you know, it's it's the teenager
that's angry at their parent and wants to move out
of their house, but they got to figure out the
details still.
Speaker 2 (24:50):
I love the analogies, and they're moving out of the
house of Microsoft to get compute from others and also
get money from others, largely Massouchy Son, who we know
has got big visions when it comes to Arizona today, Brody,
but just talk to us about why the tensions are
really hitting a focus point now.
Speaker 9 (25:05):
The focus point now is because open ai wants to restructure,
and it's we've been hearing for a long time now
about negotiations between them and Microsoft. And you know, Microsoft,
as you pointed out, they made some changes that Okay,
you want so much compute. Maybe we don't want to
give all that compute, so we're going to allow you
to go seek it from others. Seek it from Oracle,
(25:27):
seek it from you know, they're getting some from core
Weave now. And so we are seeing little bits and
changes in the contract. I mean, you know Microsoft put
in I think it was fourteen billion dollars. They are
a savvy shop. They're going to make sure they get
their full ROI here, lu.
Speaker 3 (25:44):
Meg's Brody Ford. We appreciate it.
Speaker 2 (25:45):
Thank you so much as shift gears talk quantum now
because IBM has announced the latest milestone in its roadmap
to bring quantum computing into the future, planning to build
the world's first fault.
Speaker 3 (25:56):
Tolerant quantum computer by twenty twenty nine.
Speaker 2 (25:59):
It's called them Starling and expected to perform twenty thousand
times more operations than today's current quantum computers. Joining us
now Jerry Chow, IBM, Fellow Director of Quantum Systems. And
this is a new error correcting code. You say, basically
it cracks the code on science challenges of actually making
quantum of reality.
Speaker 13 (26:18):
Absolutely.
Speaker 4 (26:19):
I mean the whole here, the whole thing here is
that we crack the code in actually building something that's
realistic and planning for this with the full architecture, which
gives us the roadmap to go year of the year
of new technologies to hit that milestone in twenty twenty
nine with IBM Starling, can you.
Speaker 2 (26:33):
Take us from here in twenty twenty five to twenty
twenty nine what.
Speaker 3 (26:36):
Needs to be achieved?
Speaker 2 (26:38):
Because there is so much hype, we kind of want
to disentangle what actually becomes reality.
Speaker 13 (26:42):
Yeah, no, absolutely, there's a great question.
Speaker 4 (26:44):
There's a lot of work that's been going on within
small demonstrations of error correction and the concepts that make
us believe that air creation can work. But what reve
defined is a technology roadmap that takes us from an
error correct did memory and the technologies that go into that,
then into operation across multiple memories, and then into a
large a larger scale system that brings in the operations
(27:06):
that are needed for a universal control to do any
kind of algorithm that gets us the Starling.
Speaker 13 (27:11):
So a year after year of new technologies in our roadmap.
Speaker 3 (27:13):
Okay, people are excited.
Speaker 2 (27:14):
IBM shares are at a record high, and I'm sure
there's no surprise that has come in tank At the
same time as some of your announcements, can you talk
to us about them different ways that different companies are
deploying this, because we've had, you know, a year or
so ago, what was it, the Heron chip from you
guys really set the pet bar. But we've also had
the Willow chip coming from the likes of Google. You've
(27:35):
had as a lot out of Amazon, and there are
different ways of doing quantum, that's right. How is what
you're doing different from some of the other competitors.
Speaker 4 (27:42):
Yeah, so I think we've really taken approach of getting
this out there and really building the whole ecosystem around us.
We've actually been putting these out on the clouds in
twenty sixteen, first just a five cubet machine back then, right,
and we've been successfully scaling since then, and we have
multiple one hundred cube machines the Heroin processors that.
Speaker 13 (27:58):
Are available to our cloud users today.
Speaker 4 (28:00):
So there's an ecosystem of many users in companies that
are leveraging our machines to actually find and discover new
algorithms and applications.
Speaker 3 (28:09):
When you see many users, what are we talking about.
Speaker 13 (28:12):
We have over five hundred thousand users.
Speaker 4 (28:13):
We have a business ecosystem, our IBM Quantum network of
nearly three hundred institutions, companies, startups, national laboratories, academia.
Speaker 2 (28:22):
And the reason everyone gets so excited when we hear
of the latest greatest chip, when we hear of a
roadmap to twenty twenty nine is because of the things
it could do. Absolutely, what are they doing it with
these five thousand people that are already out there.
Speaker 4 (28:33):
Yeah, So, like there's a lot of there's a lot
of work that's getting ready for the larger systems, but
also we're in this phase where we're going to hit
quantum advantage in the.
Speaker 13 (28:40):
Next few years. Quantum it's when we can.
Speaker 4 (28:43):
Do something that's cheaper, faster, or more accurately using a
quantum computer plus classical computer than just a classical computer alone.
Speaker 2 (28:50):
It's like AGI, but for quantum versus versus normal computers.
Speaker 4 (28:54):
And you're already starting to see the intrusion of using
it with high performance supercomputers to look at things like
lecular structure, which is a very hard problem that you
tend to only use high performance computers for.
Speaker 2 (29:06):
And that is why everyone thinks medical applications are gonna
be so important. But everyone's also got very excited about
the fact that quantum's gonna help with generas of AI.
Speaker 3 (29:14):
How is that feeding into you right now?
Speaker 4 (29:16):
Yeah, it's a really really good point because I think,
you know, AI is the buzz of Obviously quantum is
also coming right right right at the same time, I
think the way to think about it is it's additive. Right,
there are certain algorithms in the AI space that quantum
can help with. But also what we see here is
really that we're defining the future of computing. It's really bits, neurons,
and cubits, right, they all need to be working together
(29:37):
in an architecture we call quantum centric supercomputing. So you're
gonna have your CPUs, you're gonna have your GPUs, and
we're going to have your QPUs as well, all working
together to solve problems that best leverage each of those
modalities for parts of the problem and giving you overall
more power.
Speaker 2 (29:53):
By putting it in the hands of people. Power probably
is one of the key words you're thinking of. Is
that going to be a bottleneck? We thin thinking about
how power and infrastructure is the limiting factor when it
comes to AI right now.
Speaker 4 (30:04):
I mean, certainly one of the key selling points for
quantum is actually how efficient it can be for the
types of problems that you're looking to solve which require large,
high performance supercomputers. We're doing it a fraction of the
power despite even the infrastructure that we're building around some
of our super connecting Cuban architectures.
Speaker 2 (30:20):
Can you talk about the architecture of the ecosystem for.
Speaker 3 (30:23):
Us as well?
Speaker 2 (30:24):
Because Jensen Woan suddenly kind of corrected himself on what
he thinks is going to take until twenty twenty twenty
twenty years from now he thought would be when quantum
computers become really useful, and then he backtracked.
Speaker 3 (30:36):
Ultimately, he's probably.
Speaker 2 (30:37):
Backtracking because he sees this as a business line and
an opportunity for him. Where does IBM fit in terms
of who it sells to and who it works alongside.
Speaker 4 (30:45):
So our strategies really provide the differentiating computing platform in
terms of the hardware and the software at the bottom,
and then building this ecosystem all around it. A big
part of our strategy has been having an open source
software development Kisskit, which really appeals to a broader set
of users so that different startups users can all be
plugged into using this in addition to their other computing resources.
Speaker 3 (31:08):
So taught me through.
Speaker 2 (31:09):
The supply chain here, because at the moment you are
going to create the actual hardware. You're doing it in
Upstate New York, You're doing it probably across the United States,
and you've been focused on at home manufacturing in many
ways for that. But you're also going to be making
this software. You want to be a one stop shop.
But who do you have to rely on?
Speaker 13 (31:25):
Well, actually, the software pieces that open source.
Speaker 4 (31:28):
We have a lot of integration partners, third party of
startups are actually building functions within our ecosystem so that
they can drive verticals in optimization or in chemistry or
in machine learning that plug right in and leverage other
forms of compute in parallel with our machines.
Speaker 2 (31:44):
I mean, ask a bit of a loaded question, Jerry,
But we keep on comparing US versus China when it
comes to AI.
Speaker 3 (31:51):
What are you.
Speaker 2 (31:52):
Actually seeing international adoption and quantum? Does everyone work harmoniously
together when it comes to quantum at the moment, because
it's still very much that early stage or you.
Speaker 4 (31:59):
Race, We certainly see a lot of adoption throughout the world,
but certainly it is a race right, there is definitely
a technological race between the US and China. There's been
a lot of government funding in the US space in
this area, but I think we are also looking to really
make some bold moves here and really sort of cement
technologies a roadmap towards a quantum centric supercomputer that has
(32:20):
bold vision.
Speaker 2 (32:21):
Well, thanks for talking us through that bold vision today,
Jerry Choo of IBM. Fascinating to dig in on quantum. Meanwhile,
coming up, check out these last mile delivery robots.
Speaker 3 (32:30):
A company behind them.
Speaker 2 (32:31):
It is closed a new funding round backing from Open Aiyes,
Sam Altman and his brother will have the details next.
Speaker 3 (32:37):
This is a bluebg tech. They're coming behind last mild.
Speaker 2 (32:49):
Delivery robots Coco Robotics. Well, it has just raised eighty
million dollars in its latest funding round. Startups backers include
open Aiyes, Sam Altman, and SNR Zach Rash Coco Robotics
c young founder joins us.
Speaker 3 (33:01):
Now the money, what was it for, Zach?
Speaker 14 (33:04):
Well, we really want to expand the service to more areas,
So we're building the best automas vehicle for our urban environment.
So we're operating in Los Angeles, Chicago, Miami, Helsinki, Finland.
Speaker 15 (33:17):
We have a lot more cities coming up soon.
Speaker 14 (33:19):
But you know, we want to build the best vehicle
to move goods around cities, to make it much cheaper
for our customers and commercials.
Speaker 3 (33:26):
Okay, so is it all about US markets? Is there
an international focus as well here?
Speaker 15 (33:31):
Yeah?
Speaker 14 (33:31):
So in the US we've been partnering with sorda Ash
and uber Eats, and then we recently partnered with waltz
over in Helsinki. So we're live in Helsinki, Finland and
we have a lot more European markets coming up soon.
Speaker 2 (33:41):
And what are some of the bottlenecks that you've seen
thus far rolling these things out? I've seen them at
plenty of conferences. I'm yet to get the deliveries here
in New York.
Speaker 3 (33:49):
So what are is it? Government relations?
Speaker 2 (33:52):
Is it ultimately just making sure that they can they're
fit for purpose in some of these more urban spaces.
Speaker 15 (33:57):
Yeah.
Speaker 14 (33:57):
So at first we had to get the technology to
work well enough where we can deliver goods to people
at high quality. We're moving a lot of food and
groceries today, right, so you're very time sensitive. It needs
to get there on time or people won't use the robots,
and the second thing we.
Speaker 15 (34:11):
Need to do is get the cost to be lower.
Speaker 14 (34:14):
Than the cost of a human driver or doesn't make
any sense, right, So we got the technology to that place.
We did that first in Los Angeles. Now we're scaling
up as fast as we can to get this service
to more and more people. The government relations side of
this has actually been going really well. We have a
really good relationship with the cities. We work with half
the states in the country of past legislation to allow
(34:34):
this sort of vehicle to operate, and I think we're
seeing a big trend in Europe to.
Speaker 15 (34:39):
Adopt this as well.
Speaker 14 (34:41):
There's typically not enough drivers to do the deliveries that
exist today, and it's incredibly expensive. So we're able to
make it more available to people, make it much more affordable,
and that usually aligns with the goals of our cities
and just.
Speaker 2 (34:54):
Walk us through the process that those who haven't been
lucky enough to have their ice cream delivered via Coco Robotic,
what is it? How do they get contacted when they're
upping their high rise?
Speaker 14 (35:04):
Yeah, so if you order on uberate's your DoorDash, you'll
have the you know, the option to opt in to
an atomous vehicle. If you opt in, it's usually a
lower lower cost to you. There's no tipping involved, so
most people do opt in. If you're in a high rise,
you can choose. I don't want a robot. I want
someone to come and bring it up to my door.
But I think we're seeing that happen less and less,
(35:25):
and it's a I think the human delivery experience can
also be a bit inconsistent. Maybe they'll bring it up
to your door, maybe they'll leave it the lobby, maybe
they'll leave it on the curb side. So one of
the benefits of using our Times vehicles is they will
be parked in front of your store. They'll be waiting
patiently or and sorry, in front of your building. They'll
be waiting patiently for you to come down and retrieve
you order. It's locked, it's insulated, the food is packaged
(35:47):
and preserved in the best possible way for you, and
you will wait for you to be ready.
Speaker 15 (35:51):
So once you're ready, you come down, you click a.
Speaker 14 (35:53):
Button in the app, a little open and you get
your goods.
Speaker 2 (35:56):
What's interesting is, sadly we've heard over the course of
some moments of protest people taking out their anger on
driverless cars, for example, over in LA. Have you seen
any implication on your own robotics? How do people treat
them out there in the world.
Speaker 15 (36:12):
Yeah, it's well, it's interesting. We were we were I
just saw this video that was.
Speaker 14 (36:17):
Going around TikTok where the robot was driving through a protest.
We know, we worked at the city to geofence off
some areas where we shouldn't operate because there's active's protest,
but you know, they're popping up and moving all around
the city, so it's hard to be super precise. But
what ended up happening is everyone made a like a
tunnel a path and started chanting Coco and let the
(36:37):
robot through, which I thought was was was quite nice
and somewhat unexpected. But I think people have a lot
of affinity towards these I mean, we called it Coco
on purpose. It's it's you know, we looked up a
list of cute dog names. It was number one on
the list. So I think on one end, I think
people have some affinity towards them, and on the other hand,
people realize it's delivering their local businesses and their neighbors,
(37:00):
you know, dinner for their family. So I do think
people are actually quite respectful of the vehicle and making
sure that they're not disrupting, you know, their neighbors are
really the community product, Zach.
Speaker 2 (37:11):
At the moment, the competition, I'm sure is a gig
worker and a car and a delivery, But how many
robotics competitors are starting to pop up?
Speaker 3 (37:20):
And is that a good thing for the ecosystem?
Speaker 14 (37:23):
Yeah, I think the timing is really the timing is
really interesting right now. Right we've been doing this for
about five years, A few things have happened in the world.
You have AI clearly has become extremely capable.
Speaker 15 (37:37):
You first had chats to b T.
Speaker 14 (37:38):
Now you have kind of AI everywhere and all of
your products, all big techs talking about it. You have
waymos you see on the road all the time now.
So I think a time as vehicles and AI and
how that goes into the physical world has become very
real for businesses, for investors, for you know, analysts covering
some of these stocks, and for consumers. So consumers are
starting to demand these services be integrated more and more.
(38:01):
At the same time, the cost of delivery has continued
to go up and up and up, so you have
kind of this macro where these brands and businesses are
concerned about you know, consumer spending and you know how
much how much more cost they can tolerate on delivery.
And then you have this other thing of AI that's
seeming to clearly work and do useful things. So I
(38:23):
think that's caused this to be the right time. And
we have, you know, seemingly unlimited demand for the product,
and we're just trying to ramp up production as fast
as possible.
Speaker 2 (38:30):
And get this very briefly, Zach, you must have unlimited
demand for data to feed into the underlying lognacle's model
or whatever you're using to power this.
Speaker 3 (38:40):
Where are you getting it from?
Speaker 15 (38:42):
So we've done millions of miles of deliveries.
Speaker 14 (38:45):
We've done over half a million deliveries, and those are
deliveries happening in the most chaotic parts of the world.
Speaker 12 (38:51):
Right.
Speaker 14 (38:51):
So this is Chicago, this is downtown LA and Hollywood,
this is Helsinki covered and covered in snow.
Speaker 15 (38:56):
It's Miami with you know, hurricane winds going and rain.
Speaker 14 (39:01):
And not only do we have you know, video data
of the robots navigating these situations, you know, we have
our remote operators who can come in during a delivery
take over for the autonomy, take over for the AI
and actually correct it and explain for it what to do.
So that's a very rich data set to train the system,
and it has to work really well because again, we're
moving food. So if you have autonomy that works kind
(39:24):
of well, but it delays the trip by twenty minutes,
it's not useful.
Speaker 3 (39:27):
Zach, It's been great having you.
Speaker 2 (39:29):
Zach Rash, co founder and CEO of Coco Robotics. Bloomberg's
Emily Chang sat down with Melinda French Gates in Seattle
to discuss her next chapter as a philanthropic powerhouse now.
They discussed her early career at Microsoft, but also her
outlook on the Trump administration and big text influence in Washington.
Speaker 16 (39:51):
How is President Trump and these actions? How is this
impacting the philanthropic world?
Speaker 17 (39:56):
The cuts in USA ideal alone, sixteen point nine million
women will not get maternal health services. People are going
hungry because of those cuts. As Americans? Is that what
we want? Do we want to create more crises around
the world, or do we want peace and prosperity for
others so they can live where they are? Philanthropy can
(40:19):
take a risk that we wouldn't want government to take
with our taxpayer funding, but it can prove things out
at scale, and then governments can come in to scale
that up.
Speaker 16 (40:29):
Alon Musk and Doge, how do you feel about one man,
a billionaire, having that much independent power within the federal government.
Speaker 17 (40:38):
Well, it's unprecedented to have anyone come in and say,
you know, we're just going to wipe it out like that.
Speaker 3 (40:46):
That makes no sense to me.
Speaker 16 (40:48):
What do you make of the tech billionaires lining up
behind President Trump and Silicon Valley's right turn.
Speaker 17 (40:54):
What I've seen in the last sixty months to a
year is many people who use to say one thing
have absolutely shifted over here. Look, a democracy is made
up by our beliefs, and our investments and our values,
and we of all times right now should be living
those values out, not pivoting to what some calms person
(41:17):
tells us is.
Speaker 15 (41:17):
The right thing to do.
Speaker 13 (41:19):
Right That would be ridiculous.
Speaker 2 (41:20):
Emine Chang's full conversation with Melinda French Gates. It's on
the circuit on bloombgg dot com and on Bloomberg Originals.
Now it's time for talking tech, and first up, were Australia,
it's moving closer to a ban on social media for
users under the age of sixteen. Key trial found that
checking a user's age is technologically possible and can be
integrated into existing services now.
Speaker 3 (41:41):
The government backed findings.
Speaker 2 (41:42):
Helped to clear the way for new laws that would
find digital platforms responsible for enforcing the age limit plus
TENUS US sales they decline and it deepens, with weekly
sales dropping more than twenty five percent.
Speaker 3 (41:54):
Compared to a year ago. Am mid President Trump's.
Speaker 2 (41:55):
Taris now the sales route comes alongside the company's cuts
in advertising spends. The commercial output now only in the
dozens and Emil musks thirty three billion dollars buyout of
x by Xai, but it's facing new scrutiny from EU regulators.
According to sources, the European Commission says that the deal
could influence the scale of potential fines that's under the
(42:16):
Digital Services Act. The Commission is also considering whether sales
from Musks other private businesses should be included to determine
potential fines against x Now that does it For this
edition of Bloomberg Tech, do not forget to check out
the podcast Tune in this weekend. You can find it
on the terminal as well as online on Apple, Spotify,
an iHeart.
Speaker 3 (42:37):
This is Bloomberg Tech
Bloomberg Audio Studios, Podcasts, radio news. Bloomberg Tech is live
from coast to coast with Caroline Hyde in New York
and ever Though in San Francisco.
Speaker 2 (00:21):
This is Bloomberg Tech coming up one trillion dollars for
a USAI hub, the latest big bet being pitched by SoftBank,
band of Masayosti's son. We dig into the reality of
the numbers. Plus Meta launches three hundred and ninety nine
dollars oakly AI glasses aimed at athletes and thrill seekers.
We discussed the surprise hit smart glasses and China's baid Eyes,
(00:42):
a Robotaxi expansion in Asia. We discuss the next steps
for a the adoption. But we're also eyeing what's happening
in terms of Middle East tensions. I want to move
into those Middle Eastern tensions as well, because look, more broadly,
we are seeing Israeli officials warning that Iran is tapping
into private, securely focused cameras to gather some information, particular
(01:05):
intelligence against adversaries.
Speaker 3 (01:08):
Now it's going to be a.
Speaker 2 (01:09):
Key focus that we've seen repeated with hamas with Russia
in other conflicts. Bloomberg's Mike Shephard joins US now on
what is some deep investigation, mic.
Speaker 4 (01:21):
Cara.
Speaker 5 (01:21):
This is something that highlights a longstanding vulnerability with surveillance cameras,
which have become ubiquitous around the world. When we look
in our homes, even you see those doorbell cameras, and
you go to any office or department store, there are
cameras everywhere, even on the streets. Yet the data that
(01:42):
is being gathered, the images, the footage is not necessarily
as secure. We've seen instances several years ago of hacks
of startups that had just reams of data from tens
of thousands of surveillance cameras. But there is also a
geopolitical concern as well, and we see the highlighted with
Israel warning citizens to either turn the cameras off or
(02:04):
at the very least update the passwords.
Speaker 3 (02:07):
Okay.
Speaker 2 (02:08):
So a key call to their own citizens at the
moment and one that will follow throughout. But from real
war tensions to the global trade war once more, Mike,
there's some really interesting news breaking very recently across the
BlueBag terminal coming from the Wall Street Journal talking about
how the US is preparing action targeting Allies chip plants
(02:28):
in China. In particular, there's talk of ending waivers for
the likes of TSMCSK, Heinek, Samsung that actually make chips
in China. Right, it seems to be trying to make
it harder for global chip makers basically to operate in China.
Speaker 5 (02:41):
Well to a degree, yes, and the idea from the
Trump administration is to what they see it as according
to the journals reporting, level the playing field when it
comes to requirements for rare earths to be shipped to
the US. In other words, the same degree of licensing
required for that would soon be perhaps required for bringing
(03:01):
American technology, American chip making equipment into China or at
least updated or repaired in the Chinese market. Remember that TSMCSK, Heinix,
and Samsung produce a lot of chips for the American
market in China, but what they use is US equipment
and equipment from overseas. Use equipment would now need a license.
(03:25):
Rather than just getting a blanket waiver where you can
bring it in without having to check every time with
the US government, you would need that level of permission. Now,
this is not pertaining to the most advanced, the highest
end AI chips, for instance, this is a lot of
the legacy level semiconductors, the older generation chips that go
into automobiles, consumer electronics, and other lower end things. So
(03:47):
it's not exactly targeting the upper end that we've seen
the US government do, for instance by barring AMD in
video of course, from selling their most advanced AI chips. Instead,
this is taking aim in a different end of production
in China of certain kinds of chips.
Speaker 2 (04:04):
Mike Sheffin breaking it all down for us around the world,
thank you so much. Now, look, there is another top
story that we're watching. SoftBank founder Massio Shei's son is
planning a trillion dollar industrial complex in Arizona. It's to
build robots. Further, it's AI endeavors. Look TSMC potentially playing
a key role here. Lumberg's Peter Elstrom joins us to
break it down.
Speaker 3 (04:23):
One trillion is a lot, One trillion is a lot, Caroline.
Speaker 6 (04:28):
Yeah, So Masio Sheisana has another big bowld vision. What
he's pitching companies like TSMC and Samsung on now is
building out these industrial parks and places like Arizona in
the United States. The ideal is to kind of appeal
to the Trump administration and try to bring some advanced
manufacturing back to the United States. Now. There are a
few related projects going on right now, and it's not
(04:50):
as crazy as it may sound at first. Blush TSMC,
for example, is trying to bring in a lot of
its suppliers into Arizona into the United States so that
it can produce more of the advanced chips within the country.
Samsung is doing a very similar thing in Texas. And
remember soft Bank spent years investing in startups and they
have hundreds of startups that would be able to put
(05:12):
facilities there. And the focus here is going to be
on AI and robotics. Those are two big areas for
their investments. So you could also see some soft Bank
backed companies move into these kinds of ventures.
Speaker 2 (05:23):
Okay, so it doesn't sound as crazy as it seems
as how you articulate it, Peter, But we know that
actually promising money and deploying it at two very different things.
And actually Stargate, for example, already what massiochisn has been
wooing President Trump with and in line with Oracle and
Open Ai to build infrastructure in the US that's been
hard to deploy right.
Speaker 6 (05:45):
Right, that's only five hundred billion in Carolina.
Speaker 3 (05:47):
That's much more reasonable.
Speaker 6 (05:49):
Yeah, they talked about the Stargate venture with open AI
in particular, where they're going to build out data centers
in the United States. That began as one hundred billion
dollar venture and then they said, hey, this could get
as big as five hundred billion dollars. But as we've reported,
they ran into some problems there in terms of the financing.
Some of the terriff issues that the Trump administration was
talking about ran into their ability to be able to
(06:11):
get banks on board, to be able to get the
money to be able to get some of these projects underway.
So that's going a little bit slow at this point.
And you're exactly right when it comes to this adventure
that we're talking about now, the industrial parks that also
would need financing. It's not clear how quickly they'd be
able to pull that together.
Speaker 5 (06:26):
Now.
Speaker 6 (06:26):
Selt Bank has had preliminary conversations with state officials, Howard
Lutnick's office at the Commerce Department in particular, so they
have begun some of these conversations. It's not clear exactly
how far along they are. It will depend on whether
they get some of the tax incentives that they want.
Some of the other corporate support to determine whether they're
going to be able to proceed.
Speaker 2 (06:44):
Peter astrom Me, thank you for bringing us the story.
Let's dig in now with Jordan Klin. Mazuoko America's managing director,
actually wrote about this soft bank promise in his morning note.
Speaker 3 (06:54):
You're joining us.
Speaker 2 (06:55):
Now so much to cover you think this is perhaps
more of a marketing employee.
Speaker 7 (07:00):
Yeah, that's that's where the you know, the skeptic in
me thinks it's it's a little bit more of that.
I mean, you never want to underestimate soft bank. And
when it comes to AI and robotics, I think there's
a lot of interest, particularly from the capital markets, but
you know, the size and the scale seem a little
bit trumped up, so to speak, I think, and the
(07:20):
timing I would question as well. Like and that's why
I don't think you're seeing the stocks move at all,
because investors are just kind of, you know, somewhat immune
to these sort of numbers that keep you know, going
up and up, but they're not necessarily you know, believing
that that's going to impact the stocks anytime soon.
Speaker 2 (07:36):
I love that you say we're becoming immune to these
sorts of numbers. That sort of number was actually articulated
by Samumon a little bit ago with our own memory.
Speaker 3 (07:43):
Chang, just take and listen to Sam's point here.
Speaker 8 (07:47):
The interesting question is if we had if we knew
how to get a trillion dollars right now, if we don't,
would we be able to deploy that profitably in the
next few years? And I'm not sure about that, but
I feel confident we can like make five hundred million
value back.
Speaker 2 (08:02):
Samonman is saying he's confident he can get half of
your money back if you give him a trillion. So, ultimately,
are these the right bets to be betting on as
an investor in the likely winners? Would you still allocate
to TSMC if indeed a trillion is talked about in
Arizona for example?
Speaker 7 (08:21):
Well, I love TSM personally, and I really like a
lot of these AI compute names that are going to
facilitate a lot of this expansion into gen Ai. But honestly,
I don't think you need the trillion dollar headlines to
make these stocks look attractive or compelling. Everything in the
near term sounds like the demand for AI tokens and
(08:44):
inferencing is rising and Video is doing a great job
of improving the production and output of their new Blackwell
server racks and the cloud Hyperscalers who buy all this
equipment all seem keen to increase investment as opposed to
limited or low it down. So yeah, it's great when
soft Bank and others talk big picture plans down the road,
(09:07):
but I think there's just so much demand and interest
near term that investors should be focused, you know, more
on that and less on you know, big scale projects
that may or may not come about because as your
you know, prior analysts mentioned before this piece, you know,
there's a lot of regulation in government involvement that would
(09:29):
need to come to pass even before you can even entertain.
Speaker 3 (09:33):
Some of this talking of government involvement.
Speaker 2 (09:35):
That's perhaps what's in the eye of the storm for
investors right now.
Speaker 3 (09:38):
I'm seeing nvidios of.
Speaker 2 (09:39):
Near session lows at the moment and others that perhaps
make things in China, particularly chips, because there's Wall Street
Journal reporting saying that the US might make it even
harder to use US equipment in China, the likes of
tsmcsk Henich, Samsung, for example. How much of a headwind
is this for the overall semiconductor space.
Speaker 7 (09:58):
Well, it's a great question, and I think what you're
seeing here is these reports get leaked or information from
the administration gets leaked to the press. They put out
stories that talk about potential restrictions that the US could
put on our intellectual property, our leading edge equipment and
technology that would go either to our allies or anyone
(10:21):
related to China. And in my opinion, this is similar
to the reports you heard on EDA software being blocked
or restricted from China when we were trying to get
those rare earth and magnet imports turned back on. So
what I think is happening is you have two big
Korean memory companies that make memory inside China, and the
(10:45):
US wants to get leverage in the current ongoing trade
negotiations with South Korea or maybe other Asian nations who
would be exposed to these rules, and so they throw
this out as well, we might restrict the ability to
get those leading edge tools, which would basically make those
fabs inefficient. It wouldn't shut them down, but if you
(11:07):
can't get the best latest equipment, it basically puts you
at a competitive disadvantage versus companies like Micron, So I
think it's more posturing and leverage, but it would be
a real It definitely doesn't help the sentiment across SEMIS.
And that's to answer your question, because anytime you bring
up government restrictions in China, it just puts some downward
(11:30):
pressure on SEMIS. But again, I don't necessarily think this
is what is going to happen anytime soon.
Speaker 3 (11:35):
It's just a risk, just a risk, but people act
on those risks.
Speaker 2 (11:38):
And as we see the socks Semi conduct To index
now trading off by one point seven percent, and it's
kla it's land research in particular in the headwinds. Can
you give us a sense of who in despite these
sorts of environment you just have to allocate to because well,
the positives just so outweigh. Does it still remain in
video all cases TSMC at all cases? Are they're more
(12:00):
nuanced plays to be had, because it feels as though
we've been ringing that one for a while.
Speaker 7 (12:06):
Well, you mentioned the names that would be most at
risk would be the equipment company's kla ASML, potentially AMAT
and LAMB, and those stocks have been going up in
the face of a lot of concern that the US
could put further restrictions on those equipment and that those
tools going into China. So a lot of the people
(12:27):
I speak to are already cautious on the equipment stocks.
But what I actually think it definitely really won't impact
Nvidia in my opinion TSM. Honestly, again, demand exceed supply.
They're going to do fine. And the other name that
I would look at, as you mentioned, like who could
benefit is Micron. So Micron is the other large dram
(12:50):
nand flash memory supplier, and they don't have big factories
inside China. They're the only US memory company, so they're
like our leading champ and we want them to survive
and do well. And obviously they're going to try to
move more of their manufacturing into the United States, but
they would be at an advantage if the US was
(13:10):
to restrict Samsung and Heinez from getting those needed tools
in those Chinese facilities. And that's the one that's reporting
next week. That's the one I think that's going to
sound super bullish on demand for memory in AI, and
that's the one I think that could really be the
surprise upside winner.
Speaker 2 (13:28):
Upside already up twenty eight percent so far. This month
they came out with real vision for allocating to R
and D here in America. But last week we'll wait
for their earnings a little bit later. As you say,
thank you so much Jordan Klin. Always great to have
him on from Mazuha coming up as Tesla prepares to
laun James robotaxi service. Well, the global competition in autonomous
(13:51):
cab space is heating.
Speaker 3 (13:52):
Up A's next as Bluebertech.
Speaker 2 (14:03):
Chinese technology giant Baidu well it is planning to expand
its robo taxi service in Singapore and Malaysia as as
early as this year, according to people familiar with the matter.
Now the company is in discussions with potential partners in
the two markets. By news Apollogo, it's about one thousand
self driving vehicles worldwide, mostly in China.
Speaker 9 (14:21):
Now.
Speaker 2 (14:22):
This newsicalse comes as Elon Musks Tesla prepares to launch
its robotaxi services in Austin this weekend. Let's get more
on the autonomous market. Karako comments here with us. She's
a professor of Civil and Transport engineering over at the
University of Texas at Austin.
Speaker 3 (14:35):
Her work includes research on the.
Speaker 2 (14:36):
Impact of bringing self driving cars to the roads and
kara for anyone who is worried about the safety involved here,
your data shows, look, we should be pursuing this path.
Speaker 7 (14:47):
Yeah.
Speaker 10 (14:47):
So, in an early paper, just looking at the kinds
of technologies that auto manufacturers already have, we expected about
an eighty five percent crash reduction and fatality reduction and
crash cost reduction, which is a lot of money because
Americans already lose about three thousand dollars a year per
capita every year to crash costs. And what we've already
(15:10):
seen is Weimo is hitting that target on the streets
of San Francisco.
Speaker 11 (15:15):
Phoenix, and an La, where.
Speaker 10 (15:17):
It's been comparing itself to the other drivers that are
on those exact same streets.
Speaker 2 (15:22):
Of course, cost to vehicle owners, cost of life as well,
car is something that just everyone is so focused on,
but there has to be regulation that keeps up. China
is going at this full speed, so is Asian. We're
just talking about how bydo is rolling out across Singapore Malaysia.
Are we keeping pace with how quickly the innovation could
change here in the US?
Speaker 10 (15:41):
Well, I think so thanks to Tesla, but of course
they also have a gigafactory in China, and so China
may have access to all of that intellectual property and
be able to mimic it quickly to catch up with
Tesla with a Chinese based company.
Speaker 2 (15:57):
Okay, so thanks to Tesla, thanks to Weimo, thanks Zooks,
who we had on earlier in the week. There is
plenty of companies that are showing that it can be done.
Just remind us of also not the cost implications from
an insurance perspective, but more broadly, how much easier, how
much more cost efficient will it be to make these vehicles,
Because look, zuks is saying it's going to make ten
thousand in the next few years or so per year
(16:20):
in their new facility. But all of these means they're
pretty expensive pieces of equipment as it stands.
Speaker 10 (16:26):
Right so people are estimating that the Waimo vehicles right now,
which are on a very expensive Jaguar Ipace, which is
a suggested retail price of seventy five thousand, I think,
so add another seventy five thousand to get all of
the lighter and everything else installed. But if you can
install it all directly on the assembly line in one
coherent piece, that the cost comes down a bit, and
(16:48):
of course weimo's not in the business of actually manufacturing
the vehicles and Tesla is and if you go to
the Tesla website, I think they're offering the technology for
eight thousand dollars up front, plus another one hundred dollars
a month, and so that starts to sound pretty reasonable.
And we do expect those costs to come down long
term to maybe under five thousand dollars off the top.
(17:09):
And if you have to have a monthly check fee
or something, so they're updating, you know, to new standards
each month, and there might be another fifty dollars fee
on top of that.
Speaker 2 (17:18):
I suppose talk to us therefore about how FSD seems
to work at the moment, because we hear on both
sides the equation. Many love it, Many feel that it
is incredibly efficient. Others say there are still so many holes,
so many ways in which that it underperforms. We cannot
roll it out yet in the United States.
Speaker 11 (17:38):
Well, it's already out there.
Speaker 10 (17:39):
So a lot of us have friends and neighbors and
people just alongside us in the traffic stream and you
may not notice because there is somebody in that driver's seat,
but they are in FSD mode, and so I believe
they've driven billions, billions of miles already, whereas the Waymouth
fleet has gone about sixty million miles of rider only.
Speaker 11 (18:01):
So nobody in the driver's seat.
Speaker 10 (18:02):
And that's the big jump here, is that they're they're
tweaking the algorithms so that the car will be able
to move without somebody in that driver's seat and will
do a lot of the parking, which they're not doing
right now for the owners of FSD who paid that
eight thousand dollars plus one hundred dollars a month, and
of course have to be in that driver's seat legally,
(18:24):
have to be a licensed driver and have to keep
their eyes open, you know, much of the time. But
when I was in that driver's seat, I did nothing.
It was very precise, It was very impressive on very
complex network here, the routes that we were taking, but
it was broad daylight. And the one thing I noticed is,
you know, it does go above the posted speed limit
(18:45):
to keep up with the traffic around it, which the
Waymo vehicles don't do.
Speaker 11 (18:49):
And it also came, you know, a lot.
Speaker 10 (18:51):
It was much more confident around pedestrians and scooters and
cyclists than I would normally be as a driver myself.
Speaker 11 (18:58):
So that's where I'm a little bit nervous. If somebody
turns around mid stream.
Speaker 10 (19:02):
You know, you've seen the images of scooters falling in
front of.
Speaker 11 (19:06):
A way MO.
Speaker 10 (19:07):
That would be tough, So there will be some really
hard breaking or maneuvering if they're going to avoid those
kinds of crashes.
Speaker 11 (19:14):
So that's that's something I'd be watching out for.
Speaker 2 (19:16):
And when it comes to well keeping in the speed limit,
there are some settings you can adjustice there. Kara, We
thank you Kara Kocheman. She's the professor at the University
of Texas At Austin. Meta is rolling out new aiglasses
with Oakley that are targeted to athletes. The new versions,
(19:36):
which start at three hundred and ninety nine dollars, include
improved video quality features like water resistance. Man Neat Singh
are in house athlete and Bloomberg intelligence.
Speaker 3 (19:46):
Writer talks to us a little bit.
Speaker 2 (19:47):
Look it's funny the way the story is written by
Mark German. He says people are using them also for
thrill seeking. They're getting on roller coasters and using them.
Why are these Oakley versions going to be better?
Speaker 12 (19:58):
I mean, it just turns out to be a better
phone factor. When you think about, you know, the ray
Bang glasses compared to the VR headset, I would say
it's a much easier phone factor to use, you know,
as an AI glass. But at the same time, I
think they also realized that the VR headset weren't really
(20:20):
as impressive to use from a gross margin perspective, So
I think the ray Band glasses and the oak Ley
glasses probably are better from a gross margin standpoint. And look,
in the case of Meta, we know they'll be losing
about twenty billion dollars this year on reality labs, and
there is no line of sight in terms of when
(20:42):
this is going to get profitable or you know, when
the economies of scale will kick in. So from that
standpoint as well, I think it's a better pivot for
them to focus on their glasses compared to the VR headset,
where they were doing all the manufacturing themselves from the
ground up.
Speaker 2 (21:00):
They stick to display lists, or at least they enhance
the offering on display lists. How many of these do
you think they're actually going to be able to sell overall?
Because you run the numbers with your latest big Yes, I.
Speaker 12 (21:10):
Have So look, when it comes to Apple Watch, they
are about forty million units sold. Air pods are about
you know, er take seventy to eighty million in a year.
Raybang glasses were about one million, and we are headsets
were about seven million last year. So adding Oakley, which
has the same parent company we're talking about, maybe it
(21:32):
scales from one to two million this year, but it
still doesn't match the scale of an Apple simply because
the ecosystem that Apple has with the content and the
app store, it's so hard to replicate for a meta,
and that's been their challenge, even though they have invested
so much. So to my mind, ecosystem is where a
meta lacks, and it'll be a hard thing to scale
(21:54):
up even with the Oakly glasses.
Speaker 3 (21:56):
And I'm saying, I'm Brett meg Intelligence. We thank you.
Welcome back to Blueberg Tech and Karen Hydeen, New York,
not in London. Let's get a pict check on these
markets because as.
Speaker 2 (22:11):
We stand, we are currently down about one point three
percent and we've got anxiety building, whether or not it's
about Middle East tensions, whether or not it's about yet
further restrictions on international chip makers making chips with US
equipment in China. That was a key story coming from
the Wall Street Journal put some pressure on the chip
stocks in particular or off by one and a quarter percent,
as we're also just embracing ourselves with some.
Speaker 3 (22:32):
Volatility into the end of.
Speaker 2 (22:33):
The day's close that of course, as we anticipate triple witching.
Speaker 3 (22:37):
But DOLLI what resolutely remains high AMD.
Speaker 2 (22:40):
We're up one and a half percent on the day,
we were up ten percent on the week AMD. Lisa
sou that discussion about where they go next with their
own generative AI chips really ringing through when she presented
last week. We've continued on the higher side for this
particular stock meta though off by one point four percent.
Speaker 3 (22:55):
Interesting as they say, look.
Speaker 2 (22:57):
We've got the latest and greatest Oakley AI glasses. Market
less convinced as it stams. But let's talk about another
key top corporate story, Microsoft planning to cut thousands of jobs,
particularly in sales. The company tries to trim its workforce,
I mean heavy spending on artificial intelligence that's bringing bluebags
proty Ford. This has been sort of a story that
we get a drip feed on who is it hitting
(23:17):
this time?
Speaker 9 (23:19):
A lot of people see this, and they get confused because
they say, Microsoft world's most valuable company. Most days, they're
making all this money, why do they keep firing people?
A big part of it is because, you know, you
got to start spending eighty billion dollars a year on
AI data centers, and you don't want your margins to
move so much. And now we're seeing that there's this
kind of recurring rolling layoff. They laid off about six
(23:41):
thousand people a month or two ago. Now it's the
end of the fiscal year. And now who turns it
to get fired salespeople? And so it's something has happened
a bit in the social contract of these large companies
where these rolling layoffs have become much more common. And
it's not just Microsoft.
Speaker 2 (23:56):
I'm talking of contracts. There's a contract between open ai
and Microsoft, and that seems to be getting ever more
contentious at the moment. Bring us the latest prodi on
just how open ai is competing against the hand that feeds.
Speaker 9 (24:08):
Yeah, that was a nice transition. I mean, you know,
Microsoft inked this contract with open ai back in twenty
nineteen at the point a kind of you know, fledgling
AI startup, which is now one of the most powerful
companies in the world with a lot of leverage, and
they probably want a lot of things that are not
in that contours or that initial contract. And so you know,
essentially open ai wants to restructure itself. Who needs to
(24:31):
sign off on that. It's largest shareholder Microsoft. So there's
been a kind of drip drip of information around where
these disagreements are, where Microsoft wants to hold control, where
open ai wants to you know, it's it's the teenager
that's angry at their parent and wants to move out
of their house, but they got to figure out the
details still.
Speaker 2 (24:50):
I love the analogies, and they're moving out of the
house of Microsoft to get compute from others and also
get money from others, largely Massouchy Son, who we know
has got big visions when it comes to Arizona today, Brody,
but just talk to us about why the tensions are
really hitting a focus point now.
Speaker 9 (25:05):
The focus point now is because open ai wants to restructure,
and it's we've been hearing for a long time now
about negotiations between them and Microsoft. And you know, Microsoft,
as you pointed out, they made some changes that Okay,
you want so much compute. Maybe we don't want to
give all that compute, so we're going to allow you
to go seek it from others. Seek it from Oracle,
(25:27):
seek it from you know, they're getting some from core
Weave now. And so we are seeing little bits and
changes in the contract. I mean, you know Microsoft put
in I think it was fourteen billion dollars. They are
a savvy shop. They're going to make sure they get
their full ROI here, lu.
Speaker 3 (25:44):
Meg's Brody Ford. We appreciate it.
Speaker 2 (25:45):
Thank you so much as shift gears talk quantum now
because IBM has announced the latest milestone in its roadmap
to bring quantum computing into the future, planning to build
the world's first fault.
Speaker 3 (25:56):
Tolerant quantum computer by twenty twenty nine.
Speaker 2 (25:59):
It's called them Starling and expected to perform twenty thousand
times more operations than today's current quantum computers. Joining us
now Jerry Chow, IBM, Fellow Director of Quantum Systems. And
this is a new error correcting code. You say, basically
it cracks the code on science challenges of actually making
quantum of reality.
Speaker 13 (26:18):
Absolutely.
Speaker 4 (26:19):
I mean the whole here, the whole thing here is
that we crack the code in actually building something that's
realistic and planning for this with the full architecture, which
gives us the roadmap to go year of the year
of new technologies to hit that milestone in twenty twenty
nine with IBM Starling, can you.
Speaker 2 (26:33):
Take us from here in twenty twenty five to twenty
twenty nine what.
Speaker 3 (26:36):
Needs to be achieved?
Speaker 2 (26:38):
Because there is so much hype, we kind of want
to disentangle what actually becomes reality.
Speaker 13 (26:42):
Yeah, no, absolutely, there's a great question.
Speaker 4 (26:44):
There's a lot of work that's been going on within
small demonstrations of error correction and the concepts that make
us believe that air creation can work. But what reve
defined is a technology roadmap that takes us from an
error correct did memory and the technologies that go into that,
then into operation across multiple memories, and then into a
large a larger scale system that brings in the operations
(27:06):
that are needed for a universal control to do any
kind of algorithm that gets us the Starling.
Speaker 13 (27:11):
So a year after year of new technologies in our roadmap.
Speaker 3 (27:13):
Okay, people are excited.
Speaker 2 (27:14):
IBM shares are at a record high, and I'm sure
there's no surprise that has come in tank At the
same time as some of your announcements, can you talk
to us about them different ways that different companies are
deploying this, because we've had, you know, a year or
so ago, what was it, the Heron chip from you
guys really set the pet bar. But we've also had
the Willow chip coming from the likes of Google. You've
(27:35):
had as a lot out of Amazon, and there are
different ways of doing quantum, that's right. How is what
you're doing different from some of the other competitors.
Speaker 4 (27:42):
Yeah, so I think we've really taken approach of getting
this out there and really building the whole ecosystem around us.
We've actually been putting these out on the clouds in
twenty sixteen, first just a five cubet machine back then, right,
and we've been successfully scaling since then, and we have
multiple one hundred cube machines the Heroin processors that.
Speaker 13 (27:58):
Are available to our cloud users today.
Speaker 4 (28:00):
So there's an ecosystem of many users in companies that
are leveraging our machines to actually find and discover new
algorithms and applications.
Speaker 3 (28:09):
When you see many users, what are we talking about.
Speaker 13 (28:12):
We have over five hundred thousand users.
Speaker 4 (28:13):
We have a business ecosystem, our IBM Quantum network of
nearly three hundred institutions, companies, startups, national laboratories, academia.
Speaker 2 (28:22):
And the reason everyone gets so excited when we hear
of the latest greatest chip, when we hear of a
roadmap to twenty twenty nine is because of the things
it could do. Absolutely, what are they doing it with
these five thousand people that are already out there.
Speaker 4 (28:33):
Yeah, So, like there's a lot of there's a lot
of work that's getting ready for the larger systems, but
also we're in this phase where we're going to hit
quantum advantage in the.
Speaker 13 (28:40):
Next few years. Quantum it's when we can.
Speaker 4 (28:43):
Do something that's cheaper, faster, or more accurately using a
quantum computer plus classical computer than just a classical computer alone.
Speaker 2 (28:50):
It's like AGI, but for quantum versus versus normal computers.
Speaker 4 (28:54):
And you're already starting to see the intrusion of using
it with high performance supercomputers to look at things like
lecular structure, which is a very hard problem that you
tend to only use high performance computers for.
Speaker 2 (29:06):
And that is why everyone thinks medical applications are gonna
be so important. But everyone's also got very excited about
the fact that quantum's gonna help with generas of AI.
Speaker 3 (29:14):
How is that feeding into you right now?
Speaker 4 (29:16):
Yeah, it's a really really good point because I think,
you know, AI is the buzz of Obviously quantum is
also coming right right right at the same time, I
think the way to think about it is it's additive. Right,
there are certain algorithms in the AI space that quantum
can help with. But also what we see here is
really that we're defining the future of computing. It's really bits, neurons,
and cubits, right, they all need to be working together
(29:37):
in an architecture we call quantum centric supercomputing. So you're
gonna have your CPUs, you're gonna have your GPUs, and
we're going to have your QPUs as well, all working
together to solve problems that best leverage each of those
modalities for parts of the problem and giving you overall
more power.
Speaker 2 (29:53):
By putting it in the hands of people. Power probably
is one of the key words you're thinking of. Is
that going to be a bottleneck? We thin thinking about
how power and infrastructure is the limiting factor when it
comes to AI right now.
Speaker 4 (30:04):
I mean, certainly one of the key selling points for
quantum is actually how efficient it can be for the
types of problems that you're looking to solve which require large,
high performance supercomputers. We're doing it a fraction of the
power despite even the infrastructure that we're building around some
of our super connecting Cuban architectures.
Speaker 2 (30:20):
Can you talk about the architecture of the ecosystem for.
Speaker 3 (30:23):
Us as well?
Speaker 2 (30:24):
Because Jensen Woan suddenly kind of corrected himself on what
he thinks is going to take until twenty twenty twenty
twenty years from now he thought would be when quantum
computers become really useful, and then he backtracked.
Speaker 3 (30:36):
Ultimately, he's probably.
Speaker 2 (30:37):
Backtracking because he sees this as a business line and
an opportunity for him. Where does IBM fit in terms
of who it sells to and who it works alongside.
Speaker 4 (30:45):
So our strategies really provide the differentiating computing platform in
terms of the hardware and the software at the bottom,
and then building this ecosystem all around it. A big
part of our strategy has been having an open source
software development Kisskit, which really appeals to a broader set
of users so that different startups users can all be
plugged into using this in addition to their other computing resources.
Speaker 3 (31:08):
So taught me through.
Speaker 2 (31:09):
The supply chain here, because at the moment you are
going to create the actual hardware. You're doing it in
Upstate New York, You're doing it probably across the United States,
and you've been focused on at home manufacturing in many
ways for that. But you're also going to be making
this software. You want to be a one stop shop.
But who do you have to rely on?
Speaker 13 (31:25):
Well, actually, the software pieces that open source.
Speaker 4 (31:28):
We have a lot of integration partners, third party of
startups are actually building functions within our ecosystem so that
they can drive verticals in optimization or in chemistry or
in machine learning that plug right in and leverage other
forms of compute in parallel with our machines.
Speaker 2 (31:44):
I mean, ask a bit of a loaded question, Jerry,
But we keep on comparing US versus China when it
comes to AI.
Speaker 3 (31:51):
What are you.
Speaker 2 (31:52):
Actually seeing international adoption and quantum? Does everyone work harmoniously
together when it comes to quantum at the moment, because
it's still very much that early stage or you.
Speaker 4 (31:59):
Race, We certainly see a lot of adoption throughout the world,
but certainly it is a race right, there is definitely
a technological race between the US and China. There's been
a lot of government funding in the US space in
this area, but I think we are also looking to really
make some bold moves here and really sort of cement
technologies a roadmap towards a quantum centric supercomputer that has
(32:20):
bold vision.
Speaker 2 (32:21):
Well, thanks for talking us through that bold vision today,
Jerry Choo of IBM. Fascinating to dig in on quantum. Meanwhile,
coming up, check out these last mile delivery robots.
Speaker 3 (32:30):
A company behind them.
Speaker 2 (32:31):
It is closed a new funding round backing from Open Aiyes,
Sam Altman and his brother will have the details next.
Speaker 3 (32:37):
This is a bluebg tech. They're coming behind last mild.
Speaker 2 (32:49):
Delivery robots Coco Robotics. Well, it has just raised eighty
million dollars in its latest funding round. Startups backers include
open Aiyes, Sam Altman, and SNR Zach Rash Coco Robotics
c young founder joins us.
Speaker 3 (33:01):
Now the money, what was it for, Zach?
Speaker 14 (33:04):
Well, we really want to expand the service to more areas,
So we're building the best automas vehicle for our urban environment.
So we're operating in Los Angeles, Chicago, Miami, Helsinki, Finland.
Speaker 15 (33:17):
We have a lot more cities coming up soon.
Speaker 14 (33:19):
But you know, we want to build the best vehicle
to move goods around cities, to make it much cheaper
for our customers and commercials.
Speaker 3 (33:26):
Okay, so is it all about US markets? Is there
an international focus as well here?
Speaker 15 (33:31):
Yeah?
Speaker 14 (33:31):
So in the US we've been partnering with sorda Ash
and uber Eats, and then we recently partnered with waltz
over in Helsinki. So we're live in Helsinki, Finland and
we have a lot more European markets coming up soon.
Speaker 2 (33:41):
And what are some of the bottlenecks that you've seen
thus far rolling these things out? I've seen them at
plenty of conferences. I'm yet to get the deliveries here
in New York.
Speaker 3 (33:49):
So what are is it? Government relations?
Speaker 2 (33:52):
Is it ultimately just making sure that they can they're
fit for purpose in some of these more urban spaces.
Speaker 15 (33:57):
Yeah.
Speaker 14 (33:57):
So at first we had to get the technology to
work well enough where we can deliver goods to people
at high quality. We're moving a lot of food and
groceries today, right, so you're very time sensitive. It needs
to get there on time or people won't use the robots,
and the second thing we.
Speaker 15 (34:11):
Need to do is get the cost to be lower.
Speaker 14 (34:14):
Than the cost of a human driver or doesn't make
any sense, right, So we got the technology to that place.
We did that first in Los Angeles. Now we're scaling
up as fast as we can to get this service
to more and more people. The government relations side of
this has actually been going really well. We have a
really good relationship with the cities. We work with half
the states in the country of past legislation to allow
(34:34):
this sort of vehicle to operate, and I think we're
seeing a big trend in Europe to.
Speaker 15 (34:39):
Adopt this as well.
Speaker 14 (34:41):
There's typically not enough drivers to do the deliveries that
exist today, and it's incredibly expensive. So we're able to
make it more available to people, make it much more affordable,
and that usually aligns with the goals of our cities
and just.
Speaker 2 (34:54):
Walk us through the process that those who haven't been
lucky enough to have their ice cream delivered via Coco Robotic,
what is it? How do they get contacted when they're
upping their high rise?
Speaker 14 (35:04):
Yeah, so if you order on uberate's your DoorDash, you'll
have the you know, the option to opt in to
an atomous vehicle. If you opt in, it's usually a
lower lower cost to you. There's no tipping involved, so
most people do opt in. If you're in a high rise,
you can choose. I don't want a robot. I want
someone to come and bring it up to my door.
But I think we're seeing that happen less and less,
(35:25):
and it's a I think the human delivery experience can
also be a bit inconsistent. Maybe they'll bring it up
to your door, maybe they'll leave it the lobby, maybe
they'll leave it on the curb side. So one of
the benefits of using our Times vehicles is they will
be parked in front of your store. They'll be waiting
patiently or and sorry, in front of your building. They'll
be waiting patiently for you to come down and retrieve
you order. It's locked, it's insulated, the food is packaged
(35:47):
and preserved in the best possible way for you, and
you will wait for you to be ready.
Speaker 15 (35:51):
So once you're ready, you come down, you click a.
Speaker 14 (35:53):
Button in the app, a little open and you get
your goods.
Speaker 2 (35:56):
What's interesting is, sadly we've heard over the course of
some moments of protest people taking out their anger on
driverless cars, for example, over in LA. Have you seen
any implication on your own robotics? How do people treat
them out there in the world.
Speaker 15 (36:12):
Yeah, it's well, it's interesting. We were we were I
just saw this video that was.
Speaker 14 (36:17):
Going around TikTok where the robot was driving through a protest.
We know, we worked at the city to geofence off
some areas where we shouldn't operate because there's active's protest,
but you know, they're popping up and moving all around
the city, so it's hard to be super precise. But
what ended up happening is everyone made a like a
tunnel a path and started chanting Coco and let the
(36:37):
robot through, which I thought was was was quite nice
and somewhat unexpected. But I think people have a lot
of affinity towards these I mean, we called it Coco
on purpose. It's it's you know, we looked up a
list of cute dog names. It was number one on
the list. So I think on one end, I think
people have some affinity towards them, and on the other hand,
people realize it's delivering their local businesses and their neighbors,
(37:00):
you know, dinner for their family. So I do think
people are actually quite respectful of the vehicle and making
sure that they're not disrupting, you know, their neighbors are
really the community product, Zach.
Speaker 2 (37:11):
At the moment, the competition, I'm sure is a gig
worker and a car and a delivery, But how many
robotics competitors are starting to pop up?
Speaker 3 (37:20):
And is that a good thing for the ecosystem?
Speaker 14 (37:23):
Yeah, I think the timing is really the timing is
really interesting right now. Right we've been doing this for
about five years, A few things have happened in the world.
You have AI clearly has become extremely capable.
Speaker 15 (37:37):
You first had chats to b T.
Speaker 14 (37:38):
Now you have kind of AI everywhere and all of
your products, all big techs talking about it. You have
waymos you see on the road all the time now.
So I think a time as vehicles and AI and
how that goes into the physical world has become very
real for businesses, for investors, for you know, analysts covering
some of these stocks, and for consumers. So consumers are
starting to demand these services be integrated more and more.
(38:01):
At the same time, the cost of delivery has continued
to go up and up and up, so you have
kind of this macro where these brands and businesses are
concerned about you know, consumer spending and you know how
much how much more cost they can tolerate on delivery.
And then you have this other thing of AI that's
seeming to clearly work and do useful things. So I
(38:23):
think that's caused this to be the right time. And
we have, you know, seemingly unlimited demand for the product,
and we're just trying to ramp up production as fast
as possible.
Speaker 2 (38:30):
And get this very briefly, Zach, you must have unlimited
demand for data to feed into the underlying lognacle's model
or whatever you're using to power this.
Speaker 3 (38:40):
Where are you getting it from?
Speaker 15 (38:42):
So we've done millions of miles of deliveries.
Speaker 14 (38:45):
We've done over half a million deliveries, and those are
deliveries happening in the most chaotic parts of the world.
Speaker 12 (38:51):
Right.
Speaker 14 (38:51):
So this is Chicago, this is downtown LA and Hollywood,
this is Helsinki covered and covered in snow.
Speaker 15 (38:56):
It's Miami with you know, hurricane winds going and rain.
Speaker 14 (39:01):
And not only do we have you know, video data
of the robots navigating these situations, you know, we have
our remote operators who can come in during a delivery
take over for the autonomy, take over for the AI
and actually correct it and explain for it what to do.
So that's a very rich data set to train the system,
and it has to work really well because again, we're
moving food. So if you have autonomy that works kind
(39:24):
of well, but it delays the trip by twenty minutes,
it's not useful.
Speaker 3 (39:27):
Zach, It's been great having you.
Speaker 2 (39:29):
Zach Rash, co founder and CEO of Coco Robotics. Bloomberg's
Emily Chang sat down with Melinda French Gates in Seattle
to discuss her next chapter as a philanthropic powerhouse now.
They discussed her early career at Microsoft, but also her
outlook on the Trump administration and big text influence in Washington.
Speaker 16 (39:51):
How is President Trump and these actions? How is this
impacting the philanthropic world?
Speaker 17 (39:56):
The cuts in USA ideal alone, sixteen point nine million
women will not get maternal health services. People are going
hungry because of those cuts. As Americans? Is that what
we want? Do we want to create more crises around
the world, or do we want peace and prosperity for
others so they can live where they are? Philanthropy can
(40:19):
take a risk that we wouldn't want government to take
with our taxpayer funding, but it can prove things out
at scale, and then governments can come in to scale
that up.
Speaker 16 (40:29):
Alon Musk and Doge, how do you feel about one man,
a billionaire, having that much independent power within the federal government.
Speaker 17 (40:38):
Well, it's unprecedented to have anyone come in and say,
you know, we're just going to wipe it out like that.
Speaker 3 (40:46):
That makes no sense to me.
Speaker 16 (40:48):
What do you make of the tech billionaires lining up
behind President Trump and Silicon Valley's right turn.
Speaker 17 (40:54):
What I've seen in the last sixty months to a
year is many people who use to say one thing
have absolutely shifted over here. Look, a democracy is made
up by our beliefs, and our investments and our values,
and we of all times right now should be living
those values out, not pivoting to what some calms person
(41:17):
tells us is.
Speaker 15 (41:17):
The right thing to do.
Speaker 13 (41:19):
Right That would be ridiculous.
Speaker 2 (41:20):
Emine Chang's full conversation with Melinda French Gates. It's on
the circuit on bloombgg dot com and on Bloomberg Originals.
Now it's time for talking tech, and first up, were Australia,
it's moving closer to a ban on social media for
users under the age of sixteen. Key trial found that
checking a user's age is technologically possible and can be
integrated into existing services now.
Speaker 3 (41:41):
The government backed findings.
Speaker 2 (41:42):
Helped to clear the way for new laws that would
find digital platforms responsible for enforcing the age limit plus
TENUS US sales they decline and it deepens, with weekly
sales dropping more than twenty five percent.
Speaker 3 (41:54):
Compared to a year ago. Am mid President Trump's.
Speaker 2 (41:55):
Taris now the sales route comes alongside the company's cuts
in advertising spends. The commercial output now only in the
dozens and Emil musks thirty three billion dollars buyout of
x by Xai, but it's facing new scrutiny from EU regulators.
According to sources, the European Commission says that the deal
could influence the scale of potential fines that's under the
(42:16):
Digital Services Act. The Commission is also considering whether sales
from Musks other private businesses should be included to determine
potential fines against x Now that does it For this
edition of Bloomberg Tech, do not forget to check out
the podcast Tune in this weekend. You can find it
on the terminal as well as online on Apple, Spotify,
an iHeart.
Speaker 3 (42:37):
This is Bloomberg Tech
Popular Podcasts
Bookmarked by Reese's Book Club
Welcome to Bookmarked by Reese’s Book Club — the podcast where great stories, bold women, and irresistible conversations collide! Hosted by award-winning journalist Danielle Robay, each week new episodes balance thoughtful literary insight with the fervor of buzzy book trends, pop culture and more. Bookmarked brings together celebrities, tastemakers, influencers and authors from Reese's Book Club and beyond to share stories that transcend the page. Pull up a chair. You’re not just listening — you’re part of the conversation.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!