[music]

Bloomberg Audio Studios, podcasts,

radio, news.

[music]

Bloomberg Tech is live from coast to

coast with Caroline Hyde in New York and

Ed Lelo in San Francisco. [music]

This is Bloomberg Tech. Coming up,

Nvidia CEO Jensen Wong makes a big

forecast for the AI buildout.

>> [music]

>> Right here where I stand,

I see [music] through 2027

at least $1 trillion. [music]

>> Plus, we discussed the AI and M&A

landscape with the CEO of IBM hot off

the company's own GTC announcement. and

Gecko Robotics works to deploy its AI

powered robots to assess the condition

and readiness of the US Navy's warships.

But first, we turn our attention to

well, once again, a market that is

focused on geopolitics on war, but also

on the eve of the Federal Reserve rate

decision. We're currently seeing some

buying of stocks. In fact, 89 of these

100 names are in the green for the

NASDAQ 100. We're up for a second

straight day. We're up 8/10% and we're

seeing a little bit more of a risk on

attitude. dead and you're going to dig

into well perhaps one of the catalysts.

>> As I stand here right now, Nvidia shares

are completely flat in the session, but

they've been on a roller coaster across

two sessions. Late Monday, when Jensen

had his keynote, you can see the big

spike almost 5% in the session and what

was extended visibility into the demand

for Nvidia's products. Here's what he

said. Right here where I stand, [music]

I see through 2027 [music]

at least $1 trillion. A trillion dollars

is an enormous amount of infrastructure.

That infrastructure investment you could

make on Nvidia, you could make with

complete confidence.

We have now proven that.

>> Let's get right into it with Bloomberg's

equities reporter Ryan Bastella. I mean,

you've summarized what the sell size

reaction has been to that $1 trillion

number. It's mixed, but on the whole

bullish. What are you seeing?

>> Yeah, absolutely. I would say people are

in general pretty positive on this

target. It really speaks to how much

visibility uh Nvidia has going forward.

It really speaks to how they expect all

this AI infrastructure spending to

remain pretty durable, which is

something that has come under question

this year and led to a lot of volatility

across the tech space. However, I think

it also just speaks to just a certain

amount of uncertainty and a certain

amount of hesitancy when it comes to

this company and this stock. Now, we

wrote about this earlier this week, but

Nvidia has been stuck in a pretty tight

trading range for months. It's been

unable to break out even on the back of

strong earnings, even on the back of uh

yesterday's announcement. We did see a

brief spike, but then as people sort of

dissected the the target, it came back

down again. So, I think there's still a

lot of skepticism out there. Now, I

would flag that it is hosting an analyst

Q&A. I think it starts in about an hour.

That is something I think people are

going to be really drilling into this

one trillion number trying to understand

it and uh maybe that's something that

will spur a little bit more gaining. But

so far, it seems like there's been

nothing that has been able to lift the

stock.

>> Yeah. Because of those analysts who are

going to be tuning in, 75 say buy the

stock. Only one as ever says Stell sell.

And look, they see more broadly a price

target in the next 12 months of $269,

well above where we currently trade,

Ryan. So, they're still just waiting on

some other catalyst.

>> Yeah, I it's hard to know what the

catalyst is going to be at this point.

They've really talked about their

product lineup. They've talked about uh

you know, their growth expectations. The

multiple has really come in. It's I

think one of the cheapest stocks in the

MAG 7. Uh there is still a lot of

questions about what is growth going to

look like you know years out. What kind

of deceleration could we see? and people

are factoring that in. But certainly

when you talk to a lot of people,

everybody still likes Nvidia. Everybody

still thinks it's a very well-run stock.

Uh they don't really see demand drying

up anytime soon. But it does seem like

there's a certain amount of hesitancy.

Uh you know, it's almost like they built

up a tolerance for the kinds of gains

that we've seen in quarters past where

they would give some kind of big target

and we'd see the stock jump 20%. It's a

lot harder to imagine anything like that

happening today.

>> Ryan, what else are you seeing on your

desk this morning as it relates to GTC?

Like is there a broader market digestion

of the 2 and 1/2 hour speech that Jensen

Wong gave yesterday?

>> Yeah, 2 and a half I did see a lot of

people saying you know this could have

been a lot shorter. So that was kind of

funny. But beyond that we are seeing a

lot of people just digesting this. We

saw some moves yesterday in uh in IBM in

optical companies in Intel. So there are

a lot of companies like it hasn't quite

been the same way that it was in years

past where they would mention a company

and we'd see that stock rally.

Everything is a bit more muted right

now. People are really trying to

understand this and trying to understand

where this trade goes from here and

we're not seeing the sort of excitement

that we saw really at the onset of the

AI era.

>> We ran Vicelica with all the technicals

on trading this stock. Now let's get you

the fundamentals, the market perspective

with Daniel Pilling. He's senior

research analyst and portfolio manager

over at Sans Capital. Daniel, I mean,

let's talk about the demand for Nvidia's

products. You still see them as

insatiable. We got GPU, CPU, LPU. What

drove your optimism from yesterday's

speech?

Um yes so I think uh the proof is in the

pudding I might say. So when we look at

sort of the the the the labs and their

revenue growth this year to date has

been massively accelerating right so

anthropics public's numbers went from 14

to 19 billion uh literally within a few

weeks and I think the crux of what is

going on here is that aentic AI is here

it is as viral as zoom was at the

beginning of co and what I mean with

that is that once somebody starts using

it and starts doing amazing things every

colleague will follow over time and we

just don't have enough compute to really

satiate this and this is likely a

multi-year process of sort of everybody

starting to use agentic AI within their

own workflows. We have a billion people

as knowledge workers and we maybe have a

few million agents as of today. So to me

and to us this feels like sort of the

iPhone moment of 2007 and 8 where

everybody will want to buy an iPhone and

everybody will want to run an agent

which means the numbers will likely

continue to be much much much bigger

over time. And Jensen wants to be the

infrastructure for that, but also not

just the chips, but the operating system

that that goes with. Talk to us about

open core and how they're developing

sort of their own operating system for

that and for Aentki.

>> Yeah. So, it's very interesting. So, so,

so Nvidia created this security

framework to wrap these open-source

agents within an enterprise, right? And

why do they do that, right? because an

an agent will be able to access your

calendar, your payment mechanisms, your

documents. So, it's got to be very very

safe and secure. So, this is one of the

key bottlenecks that people have to

overcome to install agents within the

enterprise space. And then, as Nvidia

always does, they sort of create their

own markets. So they created this

framework called Nemo Claw which means

that everybody every enterprise in the

world now can implement an agent locally

on their devices which means that this

growth can really start to accelerate

within the enterprise because the

security is there.

>> Daniel, my Bloomberg terminal tells me

Sans Capital has about 20 million Nvidia

shares. Is that right?

>> Uh yes ballpark.

Let's go back to the number. 2025

through the end of calendar 2026, five

fiscal quarters was $500 billion of

demand just for Blackwell and Reuben. I

think that Bernstein were able to

confirm with Nvidia CFO that the 1

trillion is also Blackwell and Reuben

including associated networking. But

there was a bit that that Jensen said in

passing right after that sentence which

was we will be short. In other words, I

think they're still massively supply

constrained relative to demand. How are

you interpreting that?

>> Yeah. So, I would I would agree. I think

there's two things we need to see here.

One, uh the hyperscaler revenue growth

has to accelerate to get, you know, to

basically maybe get this market away

from focusing on some sort of peak cycle

concern and back to focusing on sort of

this growth and tokens that Jensen is

talking about. So once we start seeing

hyperscalers really reacelerate because

they're buying all these chips,

>> I think that concern goes away. And then

the second piece is the demand is truly

viral, right? So everybody in the world

I think will start seeing these agents

as a massive productivity tool. If your

colleagues are five to 10 times more

productive than you, then you have to

implement it as well. And we think the

penetration of agents today is within

single digits percentages, right? So as

we go from single digits to 100% it's a

billion people that means he needs so

much more capitals. [snorts]

>> The associated data point with with the

one trillion bar chart was the pie chart

that 60% of that demand is still coming

from the hyperscalers. It wasn't that

long ago that we were very focused on

how much Nvidia diversified away from

the hyperscalers. It seems like that

hyperscala portion is just going up.

>> Uh yes and no. Right. So hypers scalers

are by far the biggest purchases of

these things but there's also the

neoclouds which are investing heavily

and actually also very interesting

enough governments will be able to

invest more going forward as well and

and Nvidia is playing a key role in that

right they they are the one of the

leading open- source providers of these

models which means as a government you

can take the open source neotron model

from Nvidia implement it locally you can

buy all the chips from Nvidia at the

same time and sort of run your own local

uh cloud within your country. So I think

those three will be quite important but

hyperscalers will always be very

important in all of this.

>> Daniel, it's interesting that really

where Jensen always tries to lean in is

is differentiation, the fact that he's

got the right chip for the right

workload at the right time and that's

why they started to talk a lot more

about inference and grot for example,

but I'm interested about he usually

pushes us forward a little bit more.

Yes, we've already actually heard a lot

about Vera Rubin. We know a lot about

Blackwell, but why not even push us into

Feainman more? Why not hear about the

next iteration? because it's like every

single year we've got a new architecture

coming for us.

>> Yeah. So I I would actually say what

happened yesterday is quite

revolutionary. Yeah. So Nvidia is the

first company that will have two

different types of chips for inference.

So for calling on the AI models to get

an answer. There's a new LPU chip which

is super fast. Nobody else has that. And

that's combined together with sort of

the old sort of the generalized GPUs

that they always had in the past. And

and then he he he did give an

astonishing quote um that LPU or the

fast chips, they will be 35 times more

performant on a performance per watt

basis than anything that was there

before. 35 times. So I do think those

jumps are just as good, maybe even

better than in the past. Daniel, you you

clearly have a command of the spec,

right? Generation to generation on on

the on the chips. I take it back to the

stock. You know, Cara outlined where the

sell side stands on this name. The

question for you and for Sans Capital,

do you buy more based on what you heard

on stage in San Jose?

>> Yes. So, so I I think this event itself

has been very re reaffirming of the

future, but but the entire debate about

the stock I think will be resolved truly

by this. The market is looking at Nvidia

as a business that is at peak revenue

and peak earnings. We disagree and the

reason why we disagree is everything we

discussed before. Agentic AI is

exploding in terms of demand. Now this

all will get resolved once the free cash

flow and the revenue growth of the

buyers of these chips that 60% of the

hyperskllers for example will start

growing again. And when whenever you

start seeing this this entire debate

will be resolved because then the market

will say we're going from peak concerns

to actually this is a very structural

durable growth uh trajectory over the

next 5 to 10 years. I don't know

precisely when this will be solved, but

my guess is that we're going to see a

significant acceleration hypers skiers

scale and revenue growth within the next

one or two years because they're buying

hundreds of billions of dollars of these

chips. And then maybe equally important,

the return or the payback periods of

these chips is actually improved in the

past 12 months because we're completely

sold out. So basically, you're buying a

GPU and you're getting your money back

in record time, which again then leads

to the point of the hyperskllers will

start seeing more growth, better free

cash flow growth, which should translate

then hopefully to a positive outcome for

Nvidia.

>> Daniel Pilling from Sans Capital. Thank

you very much.

[music]

Let's check in on the shares of Uber and

Lyft right now in the green as you'll

see significantly and that's after both

companies announced deepening

partnerships with Nvidia which is

currently down 310 a percent. Look Uber

says it plans to roll out a global fleet

of NVIDIA powered self-driving vehicles

across 28 cities by 2028. Lyft meanwhile

will use Nvidia's AI to strengthen

machine learning systems across its

operations. For more consumer apps and

gig economy reporter Natalie Lang joins

us now. What's interesting with Uber is

that we' had some big bold ambitions

articulated back in February, was it?

And 100,000 cars are going to be on the

road in this partnership with Nvidia.

But what's the timeline now looking

like?

>> So the updated timeline is that these

vehicles will be scaled across 28 cities

by 2028. And that will really start in

earnest in 2027 in Los Angeles and San

Francisco.

I mean, Uber's about a percentage point

of having its its best day since June of

last year. So, like at first you're

like, this is a name check that's

driving the the stock, but really it's

it's more updates to an existing

partnership. What about Lyft then? I

mean, this really is by name association

because it's more about internal use.

>> Yeah, it's about internal use. uh but

part of the release also mentions future

possible deployments uh of Nvidia

powered vehicles on the lift platform

but we don't have a lot of details on

the manufacturer uh partner yet um

whereas for Uber they have a couple of

partnerships uh already announced such

as Lucid and Neuro vehicles uh wave

vehicles with Nissan that will be that

will be powered by Nvidia uh chips and

technology

>> I think it sort of goes to what RBC's

analyst Brad Ericen was spelling out

that all of This news vindicates perhaps

both Uber and Lyft's role as platforms

in the AV era. We'd all been worried

about competition, but actually they're

the ones that can bring it all together,

right? Yeah. And and this further shows

it's not just a demand generation

platform. It's not just going to be an

app that will um allow people to hail a

robo taxi. Uh but here we can see that

Uber wants to be the fleet partner. It

wants to support some remote assistance

operations for these fleets um that they

they run themselves with partners as

well.

Blue Ni lung across all the movers in

the gig economy space after GTC. Thank

you. Another deal announced at Nvidia's

GTC conference. IBM will collaborate

with the chipmaker on an open-source

project aimed to help enterprises use AI

at scale. We sat down with IBM CEO Arvin

Krishna for more on the deal and on his

take about the wider M&A landscape.

Listen to this.

>> I think the regulatory environment is

definitely friendlier where we got this

done in just uh under 4 months whereas

it used to take a lot longer uh a few

years back.

>> If regulatory environment is friendlier,

should you be doing more of it? Should

there be more M&A particularly with some

beaten up overall valuations of software

companies at the moment?

>> I'd just say watch the space.

>> Oh, watch the space. Okay. But where

would you want to add on in this moment?

I mean what would make sense to be

adding to your portfolio?

>> So we're very focused hybrid cloud and

AI and the intersection. The work we're

doing together with NVIDIA was a five

times speed up. So five times not 5% not

a little amount but uh five times. So

there we began to leverage the Nvidia

GPUs together with some of their CUDF

software combining it with our Watsonx.

data and the example we used was our

client Nestle where together we managed

to get that speed up across their

massive amounts of data and that really

is important in that case combining some

of the technologies we work on also in

open source with the Presto Presto data

engine uh Nvidia and the example at

Nestle but then we're very excited we're

going to do more work on that and then

take it into the market and take it out

to hundreds of clients uh from

That was IBM CEO Arvin Krishna. Karen,

many more headlines out there.

>> There is. It's time to talk talking

tech. First up, the UK is ramping up its

push into quantum computing, committing

more than $1.3 billion over the next

four years. It's a major bet on a

technology increasingly seen as critical

to national security, future economic

competitiveness. Plus, [music] don't

expect memory chip shortages to ease

anytime soon. SK Highix, well, it says

that the crunch could last another four

to 5 years. SK Group chairman that's

Anthony Chay notes that while chipmakers

have already ramped up capacity may not

be enough to satisfy demand until around

2030. And Samsung well is already

pulling back on its Galaxy Z Trifold

just 3 months after the launch. Now the

South Korean company plans to halt sales

and the nearly $3,000 device in its home

market and the US discontinuation

expected to follow it. [music]

Okay, coming up, amid the chaos of the

Iran conflict, Bitcoin is emerging as an

unlikely oasis [music] for some

investors. We have more on that next.

This is Bloomberg Tech.

[music]

One of the world's best known investors

in Nvidia, ARK CEO Kathy Woods says

she's optimistic about the return on

investment from Frontier AI model

providers. She spoke with Bloomberg's

Anna Edwards in London. Take a listen.

>> We are seeing uh revenue generation

exploding from uh the frontier model

providers. Uh Anthropics annualized

revenue run rate. So ARR uh went from 9

billion in December to 19 billion today.

So annualizing revenue at 19 bill that's

that's astonishing growth. Open AI from

20 to 25 billion. uh the productivity

that uh we are enjoying from these large

language models is astonishing even

within our own firm. Uh I'm even former

skeptics uh that this was going to

amount to very much you know very New

York skepticism

>> [laughter]

>> uh they're they're blown away by what

they can do.

>> Just to take a detour to geopolitics

because it's become such a dominant

market driver and I wonder how it

influences your your thinking. We are

week three of a war that's taking place

in the Middle East. Many tech businesses

of course probably quite insulated from

everything that is happening there. But

I wonder does it what what uh rethinking

does it prompt at ARC? What um what uh

shift in in focus or shift in strategy

if any does does this kind of thing

prompt?

>> Yes. Uh well of course it depends how

long-term this is. Um and we are

thinking it will be shortterm. We do

have midterm elections this year and uh

other considerations. Uh but of course

energy prices going up any supply shock

to the extent it slows unit growth down

um it will slow down the learning curves

associated with various technologies. If

this is you know a a month or two months

uh it's not going to have a big impact

at all. If this is extended co, we did

not understand that the supply shock

would reverberate for 3 years and that

inflation would take off and that

monetary policy would accommodate the

inflation the way it did. Uh we're not

in that situation right now. Monetary

policy is not accommodating uh

inflation. We're at 4.3% M2 growth on a

year-over-year basis. So nothing like

the high 20s, low30s in in CO. AR CEO

Kathy Wood there alongside Bloomberg's

Anna Edwards. And look, let's talk more

about the war in Iran because it is

fueling volatility across global

markets. But cryptocurrencies are

actually emerging as an unexpected

bright spot of late. Bitcoin and its

peers have rallied during these times of

geopolitical stress. For more, Bloomberg

Crosset reporter Isabel Lee joins us.

Look, on the day a little bit of

pullback over the last few weeks, we've

actually finally seen some what more

buying or less bearishness?

>> I think it's both. Bitcoin has been

resilient. It's now on a six-w week high

and unlike its other asset classes,

gold, equities or even other asset

classes, it's se it's seen a relative

calm. The volatility of Bitcoin has been

kind of flat and really you can point to

it to many things. Maybe um it's

institutional buying or maybe it's just

a narrative although what a lot of my

analysts are saying at least when they

talk to me it's really more institutions

especially corporate treasuries for

every fall they absorb it and they buy

more of the supply. There is a um a

technical or I guess a better phrase for

it transactional part of this story

which you write about that the the rally

is driven in part by traders unwinding

their options bets. Could you explain

that quickly Isabelle?

>> This is the thing with Bitcoin. It's

interesting because more than the

sentiment it's really all about this

mechanic. So we have traders unwinding

their options bets those that bet that

Bitcoin will continue to fall. And so

when they unwind that, as traders close

out their negative positions, Bitcoin

rallies and we have about $1.5 billion

of Bitcoin puts clustered around the

60,000 level and now we're around 75,000

level, which is again a six-month high,

so which is really impressive. And there

are 1.3 billions of calls at 75,000. So

that's why you see Bitcoin rally, but

again, it's more about uh narrative.

We're moving beyond that. Now it's about

mechanics. And when you look at ETF

flows, $1.5 billion have flown monthto

date. So that's really a sign of

confidence. Then

>> Bloomberg's Isabelle Lee. Thank you.

Cara, what are you looking at?

>> Well, I just want to look at what's

happening in the world of stable coins

because a bit of an acquisition has

happened with Mastercard and BVNK. Look,

this is a company that they're buying

for $1.8 billion to really ride into the

infrastructure play of stable coins.

Remember, Coinbase were looking at that

asset and they decided to walk away when

it cost them 2 billion. That were the

reports at least.

[music]

Welcome back to Bloomberg Tech. One of

the other stories out this morning is

Qualcomm buying back another $20 billion

of shares, also boosting its dividend

from 89 cents to 92 cents. The story is

pretty simple as Qualcomm tells it. They

want to bolster shareholder returns

while also continuing to try and

diversify this business from smartphone

to things like automotive and

increasingly and more recently data

center. That's what the CEO Cristiano

Aman is talking about in the statement.

Stock up 2%. And then another check on

Nvidia. We're now modestly lower, 4/10en

of 1%. The peak of Monday's session

during Jensen Wong's keynote was a gain

of 5% which very quickly faded as the

market interpreted the 2025 to 2027

outlook or forecast of a trillion

dollars of demand for its AI compute.

We're now down half a percentage point.

Car.

>> Yeah. and two and a half hours of

speeching it would feel. Jensen Hang

didn't actually only just forecast that

trillion dollars. A core said he also

talked about what the company will need

to get there including more copper and

optics capacity. Look those comments

really rattles shares of companies that

make data center optical components.

Let's talk about it with most common

Ricky. Look, we're off by one and a

half% on Corning and and some other

suppliers at the moment. But what's been

interesting is Lmentum and Coherent,

they've been running up into this

announcement on the anticipation that

we're all in on optics and it looks as

though there's a doublebarreled strategy

here. Yeah, totally. And it's so

interesting because we see those run-ups

and then we just have one little comment

from Jensen Wong and it just, you know,

unraveled right away. So yeah, I mean

what he said was basically that, you

know, copper is still going to be or is

going to remain important in these data

center buildouts, which just I guess

made some of the investors a little bit

concerned about all of these, you know,

optics components. The thing that's

interesting, a lot of analysts did say

this morning in some notes following GTC

that, you know, both are definitely

going to remain important in these

buildouts going forward. So there's

probably not huge cause for concern

here. And we did see some some of those

stocks recovered today. I think momentum

did kind of go back up and so is in

positive territory now.

>> Common, what's the experience of a GTC

like for the equities team? Every single

time on stage there was a name check of

any given company, you'd notice a little

tick higher uh in some of those stocks,

which were the ones of substance and

which were just literally name checks.

>> Yeah. I mean it ends up being such a

like a flurry of excitement as we're all

rushing to, you know, send headlines and

watch these shares move. I mean we did

see some things, you know, move quite

materially, you know, uh outside of

Lmentum Coherent, we saw uh shares of

Uber and Lyft jump on partnerships that

were announced. Um IBM shares also

jumped and then there were some other

little smaller ones where things quickly

faded. did I mean one that was actually

kind of interesting is Nvidia itself had

a very big spike you know around that 1

trillion figure and then actually paired

most of those gains sort of as the

market digested it but yeah overall we

really are seeing again that you know

Jensen's comments have the ability to

move markets to move these stocks and it

actually is a little bit of a flip from

last year we didn't see a ton of

movement around GTC investors were

really sort of concerned about the macro

and you know had a lot of AI skepticism

so seeing a little bit of a trend and

kind of maybe back towards normal this

year where you know they have the power

to move stocks of other companies.

>> Bloomber's Colin Rhiniki thank you very

much. Let's stick with what's going on

in broader markets today, particularly

for tech, and bring in Carol Schlife,

chief market strategist at Beimo Wealth

Management. And on a very serious note,

like with everything going on, um you

know, the war in Iran, uh considerations

around trade, a number of headlines

relating to trade this morning, even a

decline of 4/10en of a percent uh for

Nvidia, it's the second biggest points

drag at the index level for the NASDAQ

100. Carol um you know how closely were

you watching GTC over the last 24 hours

for some macrolevel impact?

>> I think not necessarily a macrolevel

impact and and it's really important to

zoom out from what's going on in these

day-to-day basises and try to think

intermediate longer term. Clearly

markets are trying to get beyond that.

And I think it's one of the reasons why

you've seen the aggregate indexes hold

in so tightly because realistically

given all the news we've had since

basically the first weekend before

January 4th, we've had several different

wars, conflicts started in here, lots of

stuff that have hit the markets and yet

the aggregate averages are still hugging

that close to all-time highs even though

there's been a lot of turbulence

underneath. But I think that's

indicative of the fact that people are

really leaning into you've still got

growth stories going on. You had GTC

reaffirm that growth story. I mean a

trillion dollars if you will out over

the next couple of years. And so you've

got a lot of momentum underneath and

investors don't want to be out of the

market when when some of that macro

[snorts] clears.

>> I find that interesting. You know, as

Jensen Wong tells it, the global economy

is in the early phase of a transition

spanning aentic AI through to robotics

and the physical AI space. You know, if

you've stud classic econ economics, um

how does one prepare for that as an

investor to understand where in that

transition the global economy is?

>> Well, I think one of the things you do

is understand we are in the middle of a

transition. and they tend to be really

muddy when they're you're in the very

short term or the new new phases of a

technology. I was rolling it back

thinking to when we first got Excel and

Word Perfect and and some of the other

early software and everyone was trying

to figure out how do I use it? Do I have

to put everything into a spreadsheet or

just some things into a spreadsheet? And

when you have that sort of recon or

reconfiguration if you will, it takes

some period of time to figure out what

the impacts are. But rolling it back and

looking at

company after company, country after

country, reinvesting or investing for

the first time in a long time in some of

those really important infrastructure

and capital investments. And that has

long long live and long tail to it, if

you will. But it's also important to

remember from economics 101 that a lot

of the economic stats we have were meant

to measure a very different society than

we have now or are emerging to. And so

that'll be part of the challenge. So as

investors I think part of it is stay

diversified, stay long, lean into it and

don't get too

uh don't hyperventilate too much about

[clears throat] day-to-day uh activity.

Should you be leaning more into the

compute buildout, the AI infrastructure

buildout? That's what's worked for the

last few years. And I'm looking at Nebas

today for example, it's selling well

convertible debt to be able to continue

to build out its own NeoCloud offering.

We know it's got to deal with meta.

Should you be long that space? Well, I I

think a piece of it is is to understand

not only the buildout, but to to look

far and wide for those companies that

are leaning into using the new

technology as well, because there's lots

of different applications, different

places that that it has the chance to to

um to supplement, if you will, not

necessarily replace everybody, but

supplement. And so where are those

leaders that are encouraging and their

employees, if you will, and prepping

their employees to be able to lean in

and use those new technology?

>> I mean, look, Carol, they've got to lean

in cuz many of them are going to be

forced out. Look, there's another

headline in your space of of banking,

Norda, which is a Nordic bank, saying

they're going to be laying off some 5%

of staff because of AI productivity.

Now, you can call it AI washing, but

when someone like Jack Dorsey is laying

off 40% of staff and he's thinking he

can do that because of AI, are we just

going to see more and more of that

impacting the labor force? I I think the

it's interesting because I do think

there's a t a a hint if you will of AI

washing to some of it, but there's also

the issue that these leaders are looking

for people in their organizations who

and they're giving them seats at the

table, the ones that are rolling up

their sleeves, getting messy, trying

this stuff out and figuring it out. But

also, there was a report out recently

that talked about the bulk of the

spending being done in AI is done on the

technology, not on teaching people how

to use it. So there's a piece of it

where we have to create an environment

where it's okay for people to experiment

with it. We're not going to replace

overnight some of these macro systems

and especially in highly regulated

industries like banking. It's hard to

believe that you're necessarily going to

totally displace an old software and

allow aentic AI to take over all of it.

But it is going to supplement what each

of us are doing and and how we're doing

it. And you go back and look just in my

business in analysis where we started

with handplotted charts and hand

calculated moving averages to deploying

new technology all the way along makes

me a lot more productive dayto-day and I

think that's what companies are looking

for but pe each of us individually is

going to have to figure out how to lean

into that too.

>> Well said Kawish Life chief market

strategist at Beimo Wealth Management.

We always appreciate your time. Let's

talk about that disruption a little bit

more now for labor because China, it

faces a key test as the nation's AI boom

really reshapes industries while putting

millions of jobs at risk. Now, analysts

warned that up to 142 million urban jobs

could vanish by 2049 due to rapid

AIdriven automation. China correspondent

Minau reports from Beijing.

That almost looks like me, but it's not.

It's an AI generated video based on a

single screenshot. Tools from Alibaba,

Tencent, Quisho are making AI video

generation accessible to millions,

sometimes for free. Our company, I can't

imagine without AI, how can we survive

at the beginning

>> without AI, I don't think our game can

actually be done in one year.

It's brought huge productivity gains for

this game developer, but also raised

alarms in the entertainment industry.

Disney and Paramount have accused Bite

Duns of IP infringement after its video

generator produced near cinematic scenes

from just a text [music] prompt. The

fear is that AI could replace not just

mundane tasks, but jobs across creative

industries.

>> [music]

>> the painting if we use like labor works

it's like 2,000 to 4,000 for one piece

[music] but with AI we only use like

probably two R&B for one piece [music]

>> for policy makers the challenge is

balancing growth we will nurture

emerging industries and industries of

the future

>> with disruption

>> the rapid development of AI is having a

profound impact on employment

>> the China econ Economic Journal projects

that more than 30% of urban jobs in

China could be lost to AI [music] by

2049. That's 142 million jobs. That's a

scenario that could threaten social

stability. Beijing is aiming to create

another 12 million jobs this year with

just as many graduates set to enter an

already slack labor market

>> in finance and IT. If there weren't for

this kind of regulatory barriers, about

30 to 40% of the job could have been

lost right away. If you're thinking

about the automation

AI replacement of Junior Rose, it's

already happening massively in China.

>> That leaves Ciinping's government with a

dilemma. China can't afford to fall

behind in its tech ways with the United

States. But the implications of the tech

revolution are far [music] from clear.

>> If we let AI replace the jobs without

taxing the AI appropriately, then then

it can really get to the core of the

consumer economy. And we got to think

about sort of tax policy that is

targeted specifically at what is driving

those job losses.

>> It will be in the regulatory framework

at some point. I just don't see it as a

choice yet because we're at a very early

stage of promoting the application of

AI. So the gray area must be kept uh

actually by a very wide margin. Last

December, state media published a

landmark arbitration case in Beijing

that set some early guardrails.

Dismissing an employee because of AI is

illegal because it's a business decision

for profit, not an uncontrollable event.

That means companies must prioritize

retraining and reassignment before

dismissing an employee. But for now, all

signs at the two sessions suggest the

government is going allin on tech at the

risk of leaving some people behind.

That was Bloomberg's Mim Laauo. Coming

up, Gecko Robotics makes a deal with the

US Navy to monitor and maintain its

warships. We speak with CEO Jake

Lucerarian. That's next. This is

Bloomberg Tech.

>> [music]

>> The US government, well, it's stepping

up its use of AI to monitor aging

infrastructure and modernized military

systems. In a new push, Gecko Robotics

has announced a $71 million partnership

with the US Navy, deploying its AI

powered robots to assess the condition,

the readiness of American warships.

Joining us now, Jake Lucazarian, his

Gecko Robotics CEO and co-founder. Jake,

can you measure what your robots are

able to achieve in terms of real term

military readiness?

>> We're all about measuring the real um

the real information and details on the

ground. And so we swarm our robots all

over these ships as you see in the

videos. And what they're doing is

they're gathering ground truth and

information that would typically take 3

or 4 months to be done and gather 1

millionth of the amount of information

in data set that is not filtered to any

source of truth and software that could

be you know gathered and then and then

in perpetuity evaluated to help with

planning into the future. What we're

providing to the to the Navy and kudos

to the Navy for adopting this technology

you know in a way that's uh giving them

an advantage over others uh that deal

with the same problem every single day.

Um they're taking very seriously this

demand of getting to 80% readiness of

their fleet and the robots collect all

this information and data that can help

perpetuate goodness both now saving 3 to

four months of cycle time um in

maintenance cycles but in but also into

the future to plan smarter so that we

have less and less days of downtime for

the vessels.

>> I mean you're seeing them flying we're

seeing them climbing we're seeing the

hardware but you're also about the

interpretation of the data as well. How

are you developing your own models to

ensure that the right information is

getting to the end user?

>> Yeah. Well, it's it's uh it's this

example and what we say at the at the

company a lot is if it's not ready, it

doesn't count. And that's very important

as it relates to the readiness of our

fleets as it relates to deterring

conflict, what's going on right now in

the Middle East and then also in the

Pacific side. And you know, we've been

doing this for 13 years. I started the

company of a college dorm with this idea

of what if you could diagnose the health

of the built world. you know, what could

that enable as it relates to predicting

what the future structures, how they

should be built, and then what what

they're going to what's going to occur

um and how to prevent that from

occurring. And so, you we've been um

improving these models and believing

that if you can gather information and

data using robots, turn atoms into bits,

you can have such an incredible

advantage as it relates to how to make

infrastructure smarter, make it make it

um more efficiently, and make even new

structures. And so these models are all

being fed into a central source of

truth, a digital thread for critical

infrastructure called canver.

>> Jake, you uh you presented basically at

the winning the AI race summit in DC

last year which was very much focused on

this administration's policy platforms

for AI. I'm curious in the time that

that since past, you've done a deal with

the Navy. you got it done. And I

wondered what that's like, you know, to

see a project through to fruition with

some of the military apparatus of this

nation.

>> It is so exciting and so incredible to

see the Navy being a leader in the world

as it relates to giving our war fighters

an advantage, an advantage that no other

um no other Navy around the world,

including China, has. And that's the

ability to have this incredible

advantage with robotics and AI to speed

up how quickly we're gathering

information and then deploying it. We're

not just doing that, you know, for the

US government. We're we're doing that

for the largest companies in the world.

Companies like Adno. Adno who's who's

giving this AI and robotics native

approach um in such an incredibly uh

advanced way. We're bringing that same

technology and beyond to the Navy and

and ensuring uh as well that it's not

just about talk about 5 years or 10

years where you're going to see

autonomous systems and and robots and AI

affecting the the Navy and our and

helping our war fighters. No, this is

actually happening today. And that is

something that the administration cares

deeply about. That's something that

Secretary Failen of the of the Navy

cares deeply about. If you want to have

the best Navy in the world, the best

assets and programs in the world, most

robust. You need impact today. And as

you're seeing today matters and it

matters more than ever,

>> right, Jake? Uh, what is your core

competence at Gecko Robotics? Are you

good at hardware or are you good at

software? We're good at building robotic

systems that go out and gather

information and data sets that no one um

one knew ever how to ever capture

before, but then two interpreting it to

allow for decisions to be made to give

you advantages that can speed up times

to to bring critical infrastructure

that's you know constantly delayed,

constantly over budget because of how

hard it is to find out where are things

broken, how to fix it. uh the robots are

able to gather all the information, not

have to destroy the infrastructure and

assets to gather it and then provide the

ability to optimize where to make the

repairs, the replacements, and then into

the future how to plan for that to

potentially never even have a shutdown.

You know, we've been able to prevent

shutdowns at oil and gas facilities that

were going to occur um that would cause

a big explosions and for the Navy to be

able to accelerate and give them

advantages in terms of how to get our

ships out of dry dock and defending

defending our values. And so we're just

so so incredibly proud to be serving u

the Navy in this way and and just uh

kudos to the Navy for taking a big bold

step as relates to giving us this

advantage.

>> Jake Lucerarian of Gecko Robotics, great

to have you back on the show. Thank you

very much.