Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:02):
Bloomberg Audio Studios, Podcasts, radio News. This is Bloomberg Business Week,
insight from the reporters and editors that bring you America's
most trusted business magazine, plus global business, finance and tech
news as it happens. Bloomberg Business Week with Carol Masser
(00:23):
and Tim Stenevek on Bloomberg Radio.
Speaker 2 (00:26):
It is a Bloomberg BusinessWeek. That is Carol Masser. I'm
Tim Stenevek. We're going to continue our conversation about in
Vidia's plunge today. In Vidio erased five hundred and eighty
nine billion dollars in value. It was the most in
market history.
Speaker 3 (00:38):
Carol.
Speaker 2 (00:39):
It was the worst day going back to March of
twenty twenty for Invidia.
Speaker 3 (00:44):
Yeah, it's pretty rough.
Speaker 2 (00:46):
That's a lot of money gone.
Speaker 3 (00:47):
Yeah, if you were if you were in all in,
if you were leveraged to the upside, that hurt today.
Speaker 1 (00:52):
Yeah.
Speaker 2 (00:53):
Jensen's net worth at one point today was down about
thirteen billion dollars.
Speaker 3 (00:57):
Yeah, that's like, hey, honey, how was your day to day?
Not so good?
Speaker 2 (01:00):
I think he's okay. He still got that in Vidia
tattoo in the leather jackets.
Speaker 4 (01:03):
That's cool.
Speaker 3 (01:04):
I'm curious if investors were pulling money out of things,
and that's what our next guest gets to.
Speaker 2 (01:08):
Yeah, Evan Faguns is watching this close to these portfolio
manager and senior equity analyst at TCW, where he co
manages the TCW Global Artificial Intelligence Strategy in Nvidia is
the largest holding in the fund. Evan joins us from
Los Angeles. Evan, good to have you with us. Just
give us an idea of how you watch a day
like this as the co portfolio manager of this AI strategy.
(01:32):
Do you sell, do you buy? Do you sit on
your hands? What do you do?
Speaker 5 (01:35):
Yeah? Absolutely, Well, thanks for having me first off, and
you know, obviously today was a tough day, and I
think it really hit on, you know, one of the
bare theses of the AI trade that's really played out
over the past two years, that all of this investment
by the hyper scalers and all the capital expenditures, it
wasn't necessary and there'd be a breakthrough in the way
that the models were architected, and you know, it would
(02:00):
be as expensive to continue to see the technology to progress.
But I think, you know, I think that last part
is very key. The technology needs to continue to progress.
The models aren't good enough yet for so many use cases.
I'm sure most people have used you know, gener of
AI and seen some hallucinations, and they need to continue
to get better. And you know, I think Deep Seeks model,
(02:23):
even though it brought down prices, it didn't actually push
out the capabilities frontier. So I think I think that's
important to note, and I think the models need to
continue to get better.
Speaker 3 (02:32):
But price is a big one, right Evan, Like, I'm curious.
I mean, this is your you know, this is your world,
and Nvidio is as to mention, the largest holding in
the fund, Meta is in there, Amazon, Alphabet, Microsoft, some
of these were going to start hearing right in terms
of earnings. But I mean, is there any kind of
major rethink, you know, when it comes to the news
(02:53):
that we got over the last few days, and really
it started coming out of Davos last week, we're some
kind of a who's who who in the tech world.
The publicly held companies and investors are being like, yeah,
we need to kind of notice what these guys are doing.
Speaker 5 (03:08):
Yeah, I mean, I think I think it's very important,
but it's also you know, the price for a given
level of performance for an AI model, it does come down.
If you think about GPT four to oh many, you know,
the price par token is down about ninety nine percent
from their original GPT four for pretty similar performance. So
once you get to a certain level, Yeah, we always
expected that the cost will come down pretty quickly, but
(03:29):
we think you know, those savings will you know, they'll
keep on going up and keep on pursuing performance rather
than costs. We think that's still more important at the
stage of the AI buildout.
Speaker 2 (03:40):
So today, with Nvidia down more than seventeen percent, and
is a buying opportunity, do you add to your position?
That was a question, by the way, Is it a
buying opportunity? Yeah, do you add to your position?
Speaker 3 (03:49):
Yeah?
Speaker 5 (03:49):
I know, I certainly. I certainly think it's it's a
buying opportunity. And I also think, you know, it's not
just the hardware players, but it's also the software players
because you know that the price coming down does mean
that it's it's cheaper for people to run these models.
So I think you did see some of that, but
I think you know that will continue and then the
ones that got punished today, there there's also a buying
(04:12):
opportunity there because it's not just software scaling and the
way that you know, these these models are architected, but
you still need the hardware side and continue to scale
with with computing.
Speaker 2 (04:21):
Does your strategy for the TCW artificial intelligence strategy change
at all?
Speaker 5 (04:29):
No, I don't think so. I think this is still
a very long term trend. And I think we always expected,
you know, costs for a given level of performance to
to come down, and I think, you know, that's moving
it out. And I actually think if you think about
the AI ecosystem as a whole, I think, you know,
after you know, the past week or two, we actually
would expect more AI to be used, you know, say,
(04:50):
five years from now than than we did, you know,
a month or so ago. So it's you know, I
think it's a breakthrough that's potentially you know, bullish for
the overall ecosystem.
Speaker 3 (04:58):
Is it safe to say, though, maybe the AI darlings
of today might not be the AI darlings you know,
four or five years from now. Is it possible?
Speaker 5 (05:08):
I think, you know, in such a rapidly moving area
like AI, you know, it's always going to be product
leaders and technology leaders that they and those those absolutely
do change from time to time. You know, I do
think in video will be one of them, you know,
to get to your your question on you know, whether
we're adding because you know, it's just hard to replace
(05:28):
that for for model training and then really cutting edge
and pushing that frontier.
Speaker 3 (05:33):
Remember Intel was once the one to be many moons ago.
Speaker 5 (05:38):
It is true, that is true. We're not seeing anything
that would make us think that if you know, in
video is going the way of the until I think
there's still so clearly uh the leader in accelerated computing.
And again, we just think that it really pays to
be you know, pushing the boundary and AI right now
(05:59):
because there's still so much on top potential. There's still
so many use cases that have yet to be at
locked by more powerful models. So I don't think that
opportunity has disappeared just because you know, the one level
of compute has gotten more cheap or gotten cheaper.
Speaker 2 (06:14):
Pat Gelsinger of Intel, you've mentioned Intel, Carol, He's out
with a tweet and look, I got it. You gotta
be careful with this stuff these days because like anybody
can have a blue check mark. Yeah, it comes to
this stuff. But this is a Twitter account that looks
to be Pat Gelsinger's. He says, it's a long tweet,
but I'll read some of it. Open wins deep Seek
(06:35):
will help reset the increasingly closed world of foundational AI
model work. Thank you deep Seek team. He says that
wisdom is learning the lessons we thought we already knew.
Deep Seek reminds us of three important learnings from computer
history at Computing History, and that third one is the
one I read. What do you think of him thanking
the deep Seek team here?
Speaker 5 (06:57):
Yeah, I certainly think what they've done is very interesting.
They tried a few different things and how they've they've
trained them all. They've really taken a DIFFERENTIATI approach. Obviously,
the open source is a big one. You know, Lama
was the only major model that was also taking that approach.
They did some things like like changing the way the
model was trained. They skipped some of the supervised fine
tuning aspects of it, so they definitely brought some innovation
(07:21):
on the software side to it. So I think there's
a lot to admire about what they've accomplished with you know,
pretty bare bones budget, But I think it really just
bodes well for again just bringing down the price and
increasing adoption of AI overall, I think it's very healthy
for the for the ecosystem, for you know, this kind
of innovation. It's really exciting.
Speaker 2 (07:41):
And by the way, if I wasn't clear, Pat Keelsinger,
former CEO of Intel, so right, but.
Speaker 3 (07:46):
Knows a thing or two about the semi world to
say at least well, Evan, I think this is what
we're trying to be really smart here at Bloomberg in
terms of the folks we were talking to and all
reporting is kind of put it in context. This is
one day. We'll see what happens tomorrow in terms of
the trade, but it is certainly something that all of
us a sudden Now the AI conversation is going to
include this company, and you know, all the CEOs on
(08:08):
the earnings call are going to be asked about it.
Again in terms of the AI investment rethink, is there
a major change in the narrative, Is there a major
change in the investment play or you're waiting for what
if there is like a change?
Speaker 5 (08:23):
Yeah, I think that you brought up a good point
and I do think you know, the next week or
two with some of these hyper scale earning supports are
going to be crucial for where it goes. You know,
I think you guys mentioned just before I came on
about how Meta just announced their increased cap backs guidance.
You know, it's hard to imagine that they would back
off of that. So soon after we'll also hear from
(08:44):
Amazon and Microsoft and Google, and I think they'll continue
to say that their full steam ahead of investing and
pushing the envelope and unlocking new use cases. So I
think it's actually pretty fortunate the timing of when this
came out for the stock market, because you know, hear
from a lot of the titans in the industry very soon,
and hopefully they'll they'll double down on these investment.
Speaker 3 (09:05):
All right, we're going to double down. Ask you one
more time, so are you going to buy in Vidia today?
Have you been buying in Vidia today? Just got about
twenty seconds.
Speaker 5 (09:12):
We're going to buy more of these AI enablers, which
you know includes everything that goes into the AI eco system,
includes in video and then some of the other lists.
Speaker 3 (09:21):
Yeah, because they're on sale today. I'm just going to say, hey,
and if you believe the fundamental long term story, right,
they're cheaper than they were on Friday. Or there's absolutely
all right, We're going to leave it there. Hey, Evan,
nice to get some time with you. I really appreciate it.
Evan Figgins. He is portfolio manager and senior equi aanalyst
over at TCW. He co manages the TCW Global Artificial
(09:42):
Intelligence Strategy and yeah, I'm looking at it. TCW A
R T I F Well.
Speaker 2 (09:48):
Southern California will clock a few more hours of showers
continue to soak the burnscarred landscape around Los Angeles, depending
on where you are. The rains certainly though, helping fight
those wildfires that have killed at least twenty eight and
destroyed more than sixteen thousand structures. They also, though, raise
the risk of deadly landslide and debris flows that can
(10:09):
overwhelm a house with mud in just seconds. A short
section of the Pacific Coast Highway west of Tapanka Canyon
Boulevard was shut due to mudslides and four vehicles had
to be pulled out near Woodland Hills on Saturday. Carol,
this is how it plays out in southern California. In California,
I mean, you know, the vegetation gets destroyed by a fire,
then the rains come and the mudslides are destructive as well.
Speaker 3 (10:29):
Right, there's nothing to hold right the back the land back.
Speaker 2 (10:32):
Doctor Carolyn Kuski is Associate vice president for economics and
Policy at the Environmental Defense Fund. It's outside of that,
she's also vice chair of the California Climate Insurance Working Group.
She joins us from Philadelphia. Carolyn, good to have you
with us. The working group charged by the governor to
quote explore innovative strategies that reduce the risk of climate
related threats. We certainly saw this play out this month
(10:53):
in California. Carol and I have spent quite a bit
of time over the last few weeks talking about the
economic effect not just on the insurance industry, but also
of course, on the way of life for many people
who've been affected by these fires in southern California. Have
you been able in your analysis to put a number
on this at this point.
Speaker 4 (11:13):
I've seen several estimates of damage and they keep getting higher.
And of course a lot of the estimates we're seeing
are the direct property damage, but we know events like
this cause a lot of non property loss as well, right,
lost income and higher rents that people have to pay
and so on, and it can really cascade through the economy,
so these become incredibly costly.
Speaker 3 (11:33):
You know, what I'm wondering is what's the future of
the insurance industry in the state of California. And I
think increasingly, you know, areas that are being impacted over
and over and over again by climate change or new
areas that never seem to see unpredictable weather that are now.
So let's start with California. What's the future there when
it comes to making sure that those that live in
(11:54):
the state, have homes in the state, can be insured
properly going forward.
Speaker 4 (11:59):
Yeah, very active question right now. We saw going back
to the twenty seventeen and twenty eighteen wildfires, which until
the recent blazes, were the two most destructive in the
state's history, and that really started a period of destabilization
in the insurance market in California, where we saw lots
of carriers pulling back, either not issuing new policies or
(12:20):
not renewing policies that they already had, particularly in the
highest fire risk areas. And as the private sector pulled back,
we saw lots of residents of the state having to
get their insurance through the state's fair plan. Through that
state program to provide insurance for those who can't find
it in the private market. In response to all these shifts,
the California Department of Insurance has changed some regulations and
(12:44):
made some updates that they hoped would keep insurers in
the state. But now we'll have to see how they
assess fire risks coming out of these blazes.
Speaker 2 (12:53):
How do you see that affecting the way that people
live and work in California. This is something that Carol
and I've been asking quite a bit. I'm a California
and my parents certainly are still in the central on
the Central Coast, and think about fire risk all the time.
Does cause Does it cause pricing to go down because
of the difficulty in getting insurance? Do people move away
(13:15):
because they can't get sufficient coverage? Do we see a
fundamental shift in the world's fifth largest economy.
Speaker 4 (13:23):
Yeah, we have some initial research evidence that as insurance
prices go up or in terms availability goes down, that
is capitalized into housing prices. So you see declining home
values and you see people taking that into account when
deciding where to live and where they want to move.
The cost of insurance can be a large part of
the total cost of occupancy in the house, and I
(13:45):
think we need to focus on the fact that insurance
is more expensive and harder to get when the risk
is really high. And so really the only long term
strategy out of that is how do we lower our risk?
And if we can lower our risk, we can not
only stabilize insurance markets, but we can keep people safer
and lower the economic costs of these as well. And
that starts a whole bunch of conversations that are really
(14:06):
outside insurance, right, questions about how we build and where
we build and what our communities look like. But those
are the conversations that the state's now starting to have.
Speaker 3 (14:15):
All right, so Q sarcasm on my part, But Carolyn,
since we're doing so well on raining in temperatures and
the rising temperature of Earth, it sounds like we're going
to have to lean on things like different building materials,
maybe building different in different spots, Like I mean, I mean,
tell me how you you're in the thick of it.
(14:36):
And we have been talking about climate change now for
a long time, and yet here we are, and the
expectations are that things just get even hotter. And so
this is our new world order. So what will be
the changes. Is it really how we build? I wonder
if that's really first and foremost how we move forward,
And what are your expectations around that? Mostly cures about
(14:58):
the cost equation around that, and that once again, does
that become problematic for homeowners?
Speaker 4 (15:04):
Yeah, those are such good points and really important questions.
And absolutely we are now locked into an environment of
increasing risk. And I think it's that fact that risk
is continually growing, it's not plateauing at some new level.
That's really hard for us to kind of wrap our
minds around and to design our institutions for as well.
So twenty twenty four was the hottest year on record,
(15:26):
but it may well be one of the coolest that
your children experience. And I think that frame can kind
of help us wrap our heads around what we're seeing.
And that's going to turn into more frequent and more
severe disasters, and so how we can protect ourselves from
them really does become paramount. And when it comes to wildfire,
there's actually a range of things, so we need to
think about the structure itself and making the home safe.
(15:48):
And then also I'm sure lots of Californians have heard
lots of talk about this. The defensible space, you know,
right around the home. And then it also because fire spreads,
it's not just what you do to your home, but
it's what your neighbors do and your community as well.
And then if we're talking about fires that are in
the forest, which is different from what we just saw
right in LA, we also have to think about forest
(16:09):
management as well. So we really have to think about
risk reduction at all these different scales and align them all.
And some of these things can be done really cost effectively.
They can pay for themselves in lower future damages. Some
are a little bit more costly, and some are just
behavior changes.
Speaker 2 (16:24):
And yeah, we only have thirty seconds left, but I
want to jump in. Do you think do you think
we see fully rebuilt in Altadena, Pacific Palace States and
other areas that have been hurt so bad?
Speaker 4 (16:37):
Yeah, I mean, communities want to come back, and I
think it's a difficult time to have the conversation and
the tragedy of post disaster, but an important time to
think about what they want the future of their community
to be and how to improve their safety in light
of these growing risks.
Speaker 3 (16:52):
Yeah, a lot to consider here the way forward and
how to prevent certainly the impact if the expectations are
to continue to see a warming Earth and more fires
and the extreme impact of climate change on those fires.
Got to figure it out and how we survive better
going forward, doctor Caroline Kuski. She's Associate VP for Economics
(17:14):
and Policy at the Environmental Defense Fund. Right here on
BusinessWeek