Latest / Bloomberg Intelligence / Global Tech Selloff, DeepSeek Buzz

Transcript

Global Tech Selloff, DeepSeek Buzz · Bloomberg Intelligence

• 0:02
  Speaker 1: Bloomberg Audio Studios, podcasts, radio news. You're listening to the
• 0:10
  Speaker 1: Bloomberg Intelligence Podcast. Catch us live weekdays at ten am
• 0:14
  Speaker 1: Eastern on Applecarplay and Android Auto with the Bloomberg Business App,
• 0:18
  Speaker 1: Listen on demand wherever you get your podcasts, or watch
• 0:21
  Speaker 1: us live on YouTube.
• 0:23
  Speaker 2: Let's take it one step further. Gene Munster, managing partner,
• 0:26
  Speaker 2: co founder for Loop Ventures, really one of the leading
• 0:29
  Speaker 2: voices on all things technology really for the last twenty
• 0:33
  Speaker 2: five thirty years. We appreciate getting a few minutes of
• 0:37
  Speaker 2: his time. Gene, I'm just gonna speak for myself, but
• 0:39
  Speaker 2: I probably speak for most of my radio and YouTube audience.
• 0:43
  Speaker 2: I did not know what deep Sick was, deep Sik
• 0:45
  Speaker 2: was before this morning, but now I'm learning. What should
• 0:50
  Speaker 2: I be learning? What should I takeaway be today?
• 0:54
  Speaker 3: I think that the highest level this whole chapter is about.
• 0:59
  Speaker 3: The first takeaway is that investors are nervous that this
• 1:03
  Speaker 3: eighty five percent moving the NASDAC over the past couple
• 1:05
  Speaker 3: of years is creating an environment where any sort of
• 1:09
  Speaker 3: ripple in what that AI trade looks like is going
• 1:11
  Speaker 3: to cause a significant impact on some of these valuations.
• 1:14
  Speaker 3: So I think that's I think this is a temperature
• 1:16
  Speaker 3: in terms of where the market is at as far
• 1:18
  Speaker 3: as kind of the company itself is concerned deep seek.
• 1:21
  Speaker 3: What's most important here is that they are advancing or
• 1:27
  Speaker 3: presumably advancing the cost of training these models you've talked
• 1:30
  Speaker 3: a lot about on the show today. And why that's
• 1:32
  Speaker 3: important is there's a question about how much hardware is
• 1:35
  Speaker 3: needed for that, and that obviously impacts the big hardware companies,
• 1:38
  Speaker 3: and then separately, how does that accelerate the adoption? So
• 1:42
  Speaker 3: there is the truth is always several layers below the surface,
• 1:46
  Speaker 3: and I think when it comes to Dee Sep there
• 1:48
  Speaker 3: are still some pretty significant unanswered questions, and I'll start
• 1:52
  Speaker 3: with one of probably the biggest one is that this
• 1:54
  Speaker 3: five to six million dollar training number that has the
• 1:57
  Speaker 3: market upside down today, that was not their most recent model.
• 2:02
  Speaker 3: They have an R model. It's unclear, you know, maybe
• 2:05
  Speaker 3: that was twenty five fifty one hundred million dollars to
• 2:08
  Speaker 3: train it. It probably was below what the most recent
• 2:12
  Speaker 3: version of Opening Eyes model, which is anticipated or estimated
• 2:16
  Speaker 3: to be about a five hundred million dollars in terms
• 2:19
  Speaker 3: of training. So I think big picture here is that
• 2:23
  Speaker 3: how we are delivering AI is evolving and there may
• 2:26
  Speaker 3: be some pretty significant improvements in terms of the ability
• 2:30
  Speaker 3: to do that at a lower cost.
• 2:33
  Speaker 4: So to that pointing, how much if we look at
• 2:36
  Speaker 4: Envidia in particular, it's down about two hundred and eighty
• 2:38
  Speaker 4: three points. What part of that is justified?
• 2:44
  Speaker 3: So I think the big picture here, I think that
• 2:46
  Speaker 3: this is an overreaction. I think when you look at
• 2:49
  Speaker 3: a company like Nvidia, it's a company we do own,
• 2:52
  Speaker 3: and I think that this is an overreaction, and just
• 2:55
  Speaker 3: specifically around that, if we can kind of zero in
• 2:58
  Speaker 3: on this concept that these models are becoming more efficient,
• 3:02
  Speaker 3: let's just take deep seek at face value that they've
• 3:05
  Speaker 3: had some sort of improvement. Is that that improvement if
• 3:09
  Speaker 3: you believe that the US tech companies are competent, that
• 3:13
  Speaker 3: improvement is around some architectures that have been talked about
• 3:16
  Speaker 3: for the last couple months. So those companies, the big
• 3:21
  Speaker 3: tech companies, the companies that are behind all the announcements
• 3:25
  Speaker 3: last week Meta increasing their CAPEX spend, and I think
• 3:31
  Speaker 3: that that is all that is a belief that they
• 3:35
  Speaker 3: still need this hardware. So when it comes back to
• 3:37
  Speaker 3: AI and comes back to the Nvidia trade, is that
• 3:41
  Speaker 3: I believe if they are reducing the costs to train
• 3:45
  Speaker 3: these models that I actually can build. I think a
• 3:49
  Speaker 3: credible case that that could increase the demand for some
• 3:52
  Speaker 3: of this hardware if we are getting closer to general intelligence.
• 3:57
  Speaker 3: If that potential is even closer, it's just going to
• 3:59
  Speaker 3: be an insatiable arms race. It's going to continue. And
• 4:02
  Speaker 3: so I understand why and video is down today. I
• 4:05
  Speaker 3: think it's actually a healthy thing. But ultimately I think
• 4:07
  Speaker 3: that their business is going to be just fine. We
• 4:09
  Speaker 3: have to wait twenty one trading days before we hear
• 4:11
  Speaker 3: from at least in VIDI on that, but I think
• 4:13
  Speaker 3: it's going to be fine.
• 4:15
  Speaker 2: So, Gene, was it a coincidence that on Friday, Meta
• 4:20
  Speaker 2: upsets capex forecast dramatically, which by a magnitude I've never seen,
• 4:24
  Speaker 2: from fifty billion to sixty five billion, and then today
• 4:28
  Speaker 2: we get this news on deep sick? Was Deep seek?
• 4:31
  Speaker 2: Was that kind of I don't know, coincidence?
• 4:35
  Speaker 3: I think, what's I think that's just purely coincidence. But
• 4:39
  Speaker 3: I do kind of go back to what we're just
• 4:41
  Speaker 3: talking about a minute ago. Is I think it's important
• 4:43
  Speaker 3: to note is that that announcement from Meta, if you
• 4:47
  Speaker 3: assume a Meta is Meta is competent, and I believe
• 4:50
  Speaker 3: that they are. That announcement came with the full knowledge
• 4:53
  Speaker 3: of what deep Seek was doing, so it's new to
• 4:56
  Speaker 3: most of us today. But for these companies who have
• 4:59
  Speaker 3: been making these huge investments, the concept of what deep
• 5:02
  Speaker 3: Seak had been trying to build has been They've had
• 5:05
  Speaker 3: a model that's been out. This has been something that's
• 5:07
  Speaker 3: been aware for the last couple of months, and so
• 5:10
  Speaker 3: I think that is important. Is that I think again
• 5:13
  Speaker 3: that this commentary from on Friday from Zuckerberg about the
• 5:16
  Speaker 3: increased capex factors in some of these potential breakthroughs that
• 5:23
  Speaker 3: we've seen with deep Seek. I want to stop short
• 5:25
  Speaker 3: of saying it's definitively a breakthrough because a lot of
• 5:28
  Speaker 3: unknowns around how good the models are and what the
• 5:30
  Speaker 3: true costs are. But I think that the AI investment
• 5:34
  Speaker 3: phase is still alive and well, this is still very early.
• 5:38
  Speaker 4: So if this could be an overreaction, we don't know
• 5:42
  Speaker 4: fair enough. In Nvidia, obviously an overreaction that stocked. The
• 5:45
  Speaker 4: way you look at it, does it make you think
• 5:48
  Speaker 4: though a rating of forty one times estimated PE.
• 5:55
  Speaker 3: So from where my head's at and thinking about calendar
• 5:58
  Speaker 3: twenty six at this point, and based on the calendar
• 6:01
  Speaker 3: twenty six numbers, I think it's closer to like a
• 6:03
  Speaker 3: twenty five, and I think that they can grow earnings
• 6:06
  Speaker 3: more than that. This is where the basically the rubber
• 6:08
  Speaker 3: hits the road, is that ultimately what happens in calendar
• 6:12
  Speaker 3: twenty six, all this stuff we're talking about with deep
• 6:14
  Speaker 3: seek today really impacts Ynvidia in twenty six. And if
• 6:17
  Speaker 3: in fact this dramatically changes how the kind of hardware
• 6:20
  Speaker 3: that people need less hardware, then I'm going to be wrong.
• 6:23
  Speaker 3: But if this does create this accelerated arms race, then
• 6:26
  Speaker 3: I think that they're going to grow faster. And so
• 6:28
  Speaker 3: I consider a one peg on that twenty six earnings
• 6:31
  Speaker 3: is actually attractive valuation relative to the rest of the group.
• 6:36
  Speaker 3: So I'm comfortable with the evaluation.
• 6:38
  Speaker 2: Just a red headline crossing the Bloomberg criminals we speak,
• 6:41
  Speaker 2: deep Seek says it is subject to large scale malicious attack.
• 6:46
  Speaker 2: So that's a headline. We'll have more reporting on that
• 6:48
  Speaker 2: going forward. Gene, just real quickly, we're going to hear
• 6:51
  Speaker 2: from some of the other big tech companies, Microsoft, Meta, Tesla,
• 6:54
  Speaker 2: how do you think they're going to address this issue?
• 6:57
  Speaker 3: It's obviously this is going to be front and center
• 6:59
  Speaker 3: the topic, and I think they're probably going to address
• 7:02
  Speaker 3: it by saying they still have plans to invest meaningfully
• 7:06
  Speaker 3: more in counter twenty five over twenty four when it
• 7:08
  Speaker 3: comes to Capex, and I think that that will play
• 7:11
  Speaker 3: part of a for reassuring we're long ways away from that.
• 7:14
  Speaker 3: We got two and a half trading days awave before
• 7:15
  Speaker 3: we start to get some of that commentary before the
• 7:18
  Speaker 3: mic gets turned over to the other side of the
• 7:20
  Speaker 3: equation here.
• 7:21
  Speaker 2: All right, Gene, thank you so much for joining us.
• 7:23
  Speaker 2: I know you're super busy today. Appreciate getting a few
• 7:25
  Speaker 2: minutes of your time. Gen Munster, managing partner co founder
• 7:28
  Speaker 2: Lupet Ventures, taking a little bit of a I guess
• 7:31
  Speaker 2: a longer term view, in a more broader view of
• 7:33
  Speaker 2: what this can mean for the indusuay.
• 7:36
  Speaker 1: You're listening to the Bloomberg Intelligence Podcast. Catch the program
• 7:40
  Speaker 1: live weekdays at ten am Eastern on Applecarplay and Android
• 7:44
  Speaker 1: Auto with the Bloomberg Business app. You can also listen
• 7:47
  Speaker 1: live on Amazon Alexa from our flagship New York station,
• 7:51
  Speaker 1: Just say Alexa play Bloomberg eleven thirty.
• 7:54
  Speaker 4: All right, let's get more on this text sell off
• 7:56
  Speaker 4: here Kim Forrest, founder and CIO of Boca Capital Partners.
• 7:59
  Speaker 4: She's also a former Chips person, like she did the stuff,
• 8:02
  Speaker 4: so she knows the things.
• 8:04
  Speaker 5: Clearly.
• 8:04
  Speaker 4: The deep Seek news is upending most of the tech industry, right.
• 8:08
  Speaker 4: You get the socks and DEXes down hard, you get
• 8:10
  Speaker 4: all the hyperscalers are down hard, all the chip makers,
• 8:13
  Speaker 4: all the power makers all down hard. Is this the
• 8:17
  Speaker 4: reckoning or is this a by the dip moment?
• 8:21
  Speaker 5: Well, that is the question, isn't it. I would say
• 8:26
  Speaker 5: it's a little bit of both. I think this is
• 8:28
  Speaker 5: a good level set, re leveling, resetting our expectations. So
• 8:34
  Speaker 5: the first thing is, we have to really see if
• 8:36
  Speaker 5: this is true. Now, apparently the company is in deep
• 8:41
  Speaker 5: Seek is an open source company. They have released how
• 8:46
  Speaker 5: they've done this, but you know, replication is key, so
• 8:50
  Speaker 5: we're gonna have to see if this actually works. That's
• 8:53
  Speaker 5: thing number one, Like, don't get too upset about major
• 8:57
  Speaker 5: developments until you know that they're real major developments, right,
• 9:01
  Speaker 5: So that's thing one. Thing two is and I think
• 9:05
  Speaker 5: this has been troubling me for a very long time,
• 9:09
  Speaker 5: especially with like last week's Stargate announcement. It seems as
• 9:15
  Speaker 5: if people have been expecting AI, which we don't know
• 9:21
  Speaker 5: what the payoff is yet we have a pretty good
• 9:23
  Speaker 5: idea it might help us out. But we're going to
• 9:26
  Speaker 5: cover the earth essentially in data centers and use far
• 9:29
  Speaker 5: more power than we create right now. So those two
• 9:32
  Speaker 5: things are kind of troubling that people have been making
• 9:35
  Speaker 5: investments in companies going to produce far more than they
• 9:39
  Speaker 5: can produce now. And do you see what I'm saying here?
• 9:42
  Speaker 5: We had these really big expectations, and I didn't think
• 9:47
  Speaker 5: that they were going to come to happen in short
• 9:49
  Speaker 5: or long term because of the well physical limitations of it.
• 9:53
  Speaker 2: So Kim, at this very early stage, I think I'm
• 9:55
  Speaker 2: probably representative of most of our listeners and viewers had
• 9:58
  Speaker 2: not heard of deep Seek before this morning. Sure, and
• 10:00
  Speaker 2: so that's calling into question kind of what we think
• 10:03
  Speaker 2: we understood about AI and its implication. It's not just
• 10:06
  Speaker 2: for technology preference stock market in general. My only question
• 10:09
  Speaker 2: that I think I have now that has any relevance
• 10:11
  Speaker 2: is do I have to rethink my spending associated with AI.
• 10:18
  Speaker 5: It depends on how long your timeframe is. If it's
• 10:21
  Speaker 5: the very short term, probably not you you know you'll
• 10:24
  Speaker 5: get rewarded because well orders are in and all that
• 10:28
  Speaker 5: kind of good stuff. And by short term I'm talking
• 10:30
  Speaker 5: a year, but let's say three to five years maybe,
• 10:33
  Speaker 5: And why is that? Well, if we can really train
• 10:36
  Speaker 5: models more rapidly with less input, which is a good thing, right,
• 10:42
  Speaker 5: where were we going to get all this electricity? That
• 10:45
  Speaker 5: was like my biggest question. But anyhow, back to your question,
• 10:50
  Speaker 5: I think you know this is part of being an
• 10:56
  Speaker 5: investors understanding how long was I play? I'm keeping this,
• 11:01
  Speaker 5: So that's part of the problem. We have two more
• 11:05
  Speaker 5: problems that are not being clarified by deep Seek. One is,
• 11:09
  Speaker 5: and remember I used to do this, so I'm way
• 11:12
  Speaker 5: down the road from most investors in thinking about things.
• 11:16
  Speaker 5: But as I use it, I'm not getting the results
• 11:20
  Speaker 5: that I need, And by that I mean it's not
• 11:23
  Speaker 5: right enough. The error rate that I get back from
• 11:26
  Speaker 5: the questions that I'm asking are anywhere between five and
• 11:30
  Speaker 5: maybe thirty percent, which that's huge. Like I can't just
• 11:34
  Speaker 5: use AI. I have to babysit AI. So these are
• 11:38
  Speaker 5: problems that aren't solved by anybody at this point. And
• 11:41
  Speaker 5: then there is how are we actually going to use it?
• 11:44
  Speaker 5: I think there was I think a Bloomberg report that
• 11:48
  Speaker 5: was saying that probably two hundred thousand people could be
• 11:51
  Speaker 5: laid off those lower level entry level people into finance
• 11:56
  Speaker 5: and then that mid tier that are now you know,
• 12:00
  Speaker 5: kind of like the AI for investment banking and equity research.
• 12:06
  Speaker 5: But if we were that wrong, we wouldn't last right, Like,
• 12:10
  Speaker 5: if you're five percent wrong in investment banking, your career
• 12:13
  Speaker 5: is really short. So you know, these are problems that
• 12:17
  Speaker 5: aren't being solved. So that's another issue for investors. How
• 12:22
  Speaker 5: right How long is it going to be till these
• 12:24
  Speaker 5: get right enough to replace a person?
• 12:26
  Speaker 4: That's such a great point because I was also reading
• 12:28
  Speaker 4: an article that talked about the limitations so far that
• 12:30
  Speaker 4: have been seen about Gee, what is it deep Seek?
• 12:35
  Speaker 4: I want to say, geek see deep Seek was its
• 12:40
  Speaker 4: lack of inability to discuss jijenping or to give real
• 12:44
  Speaker 4: handswers on Tanam and square, just those topics that are
• 12:47
  Speaker 4: near and dear to China's heart. And that's in part
• 12:51
  Speaker 4: some of the issue that you're talking about. So who
• 12:53
  Speaker 4: wins in this right now?
• 12:56
  Speaker 5: Who wins? Well, I think investors are losing in the
• 13:00
  Speaker 5: short term. But if you like me think that AI
• 13:05
  Speaker 5: is a path to productivity, and productivity always wins. I
• 13:09
  Speaker 5: think you could be a winner. I think maybe companies
• 13:14
  Speaker 5: like open Ai and Microsoft, you know, by dint of
• 13:19
  Speaker 5: you know, investing in them, and then everybody else who's
• 13:23
  Speaker 5: developing AI, they might be winners. Maybe the what is
• 13:28
  Speaker 5: it Stargate? I don't know why. I can never remember.
• 13:30
  Speaker 5: That's my break up here. I can ever remember that
• 13:33
  Speaker 5: because it has nothing to do with stars or gates,
• 13:35
  Speaker 5: that Stargate might be the loser in those people because
• 13:39
  Speaker 5: we're not necessarily going to need to cover the earth
• 13:42
  Speaker 5: in data centers.
• 13:43
  Speaker 4: Right, yep, all.
• 13:45
  Speaker 2: Right, Kim, thanks so much for joining us. Always appreciate
• 13:47
  Speaker 2: getting your perspective of your technology background. Kim Farst, founder
• 13:51
  Speaker 2: and chief investment officer of Book Capital Partners, joining us
• 13:56
  Speaker 2: from Pittsburgh via the zoom Thing.
• 14:00
  Speaker 1: Listening to the Bloomberg Intelligence Podcast. Catch us live weekdays
• 14:04
  Speaker 1: at ten am Eastern on Applecarplay and Android Auto with
• 14:07
  Speaker 1: the Bloomberg Business App. Listen on demand wherever you get
• 14:10
  Speaker 1: your podcasts, or watch us live on YouTube.
• 14:14
  Speaker 4: Happy mindy, everybody, Alex Deal here alongside Paul Sweeny. This
• 14:16
  Speaker 4: is Bloomberg Intelligence Radio. We are broadcasting to live for
• 14:19
  Speaker 4: Interactive Brookers Studio right here in Midtown Manhattan. Also check
• 14:23
  Speaker 4: us out on YouTube dot com. Or clearly have the
• 14:25
  Speaker 4: tech sel off underway really concentrated in certain names the
• 14:29
  Speaker 4: mag seven obviously, in Vidia, the chip stocks, although Apple's up,
• 14:33
  Speaker 4: I should say, but in Vidia obviously getting pounded, the
• 14:35
  Speaker 4: chip stocks getting pounded as well. So we wanted to
• 14:37
  Speaker 4: continue the conversation about AI and you really need to
• 14:41
  Speaker 4: rethink this investment thesis. Joining us now is Josh Pantoni.
• 14:45
  Speaker 4: He's CEO boosted dot Ai Now. The company has helped
• 14:48
  Speaker 4: dozens of investment managers who's AUM totals over one trillion
• 14:53
  Speaker 4: dollars and how to implement machine learning in their portfolios.
• 14:56
  Speaker 4: First off, Josh, have you seen this new geek seek? Geeks?
• 15:01
  Speaker 4: Why do I keep saying geek seek? It's not geeks
• 15:04
  Speaker 4: seek deep see see if you try deep seak? Have
• 15:10
  Speaker 4: you talked to your clients about it? What do you think?
• 15:13
  Speaker 6: Yeah? So, I mean it came out I think last Monday.
• 15:16
  Speaker 6: We've already ran a whole suite of benchmarks against it.
• 15:19
  Speaker 6: Try that out over a bunch of things, and I'd
• 15:20
  Speaker 6: say we're pretty familiar with it. I think it's impressive
• 15:24
  Speaker 6: in a number of ways. It's impressive how how much
• 15:27
  Speaker 6: power you get for the cost. I think the flip
• 15:29
  Speaker 6: side is it's actually not quite something that we could
• 15:33
  Speaker 6: even consider implementing it to our process. And I think
• 15:36
  Speaker 6: there's a lot of clients that will reach the same
• 15:37
  Speaker 6: conclusion after they start working with it. Why is that.
• 15:42
  Speaker 5: So?
• 15:42
  Speaker 6: A lot of how we help our clients is with
• 15:44
  Speaker 6: things like trying to identify risk and trying to identify
• 15:49
  Speaker 6: to build out these different workflows for different parts of
• 15:51
  Speaker 6: the financial process. And a really big part of that
• 15:54
  Speaker 6: is related to trust. You need to be able to
• 15:56
  Speaker 6: give answers that you're very confident in. You need to
• 15:58
  Speaker 6: give answers that you can sort of deeply into it
• 16:01
  Speaker 6: where it's coming from. And I think one of the
• 16:03
  Speaker 6: challenge of this model is that it has some very
• 16:05
  Speaker 6: clear political biases right off the bat. So if I'm
• 16:08
  Speaker 6: trying to use it to say understand what my portfolio
• 16:10
  Speaker 6: exposure is to China's going to be a challenge to
• 16:14
  Speaker 6: use it for things like that. Also, a lot of
• 16:17
  Speaker 6: the things we're most excited about are things like being
• 16:20
  Speaker 6: able to do computer use, being able to sort of
• 16:21
  Speaker 6: teach the machine how to like interact with different apps
• 16:24
  Speaker 6: and things on the computer, image recognition, trying to extract
• 16:27
  Speaker 6: out charts and things like that, and a lot of
• 16:30
  Speaker 6: those kind of capabilities that actually doesn't really have it
• 16:32
  Speaker 6: and seem to be as sophisticated in so it can
• 16:34
  Speaker 6: do very well in certain benchmarks, but when it actually
• 16:37
  Speaker 6: comes to trying to apply it for at least most
• 16:39
  Speaker 6: of the use cases that we're doing, it's not quite
• 16:41
  Speaker 6: there yet.
• 16:42
  Speaker 4: What is there for you right now? Like what does work?
• 16:46
  Speaker 6: I think the most interesting thing for me is actually
• 16:48
  Speaker 6: on the reasoning side. So of course, like one of
• 16:50
  Speaker 6: the really big pushes Opening Eyes had right now is
• 16:52
  Speaker 6: like with one and three, you've got these models coming
• 16:55
  Speaker 6: in where in theory they can take a task, break
• 16:58
  Speaker 6: it into subcomponents, and then start executing those components. And
• 17:01
  Speaker 6: I think in that area it's actually quite good. I
• 17:05
  Speaker 6: also think, you know, just from a cost basis, it's
• 17:08
  Speaker 6: extremely impressive they've been able to do. There's been some
• 17:11
  Speaker 6: controversy about exactly how much it costs to actually train
• 17:14
  Speaker 6: the model. What I can confirm though, is the actual
• 17:16
  Speaker 6: inference cost. The cost of running the model is extremely
• 17:19
  Speaker 6: cheap compared to what you're getting with like oh one
• 17:21
  Speaker 6: and oh three, and so just the fact is even
• 17:23
  Speaker 6: possible to do that as a major technology breakthrough.
• 17:26
  Speaker 2: If nothing else does Deep Seek just highlight the cost
• 17:31
  Speaker 2: issue and potentially the I don't know, the movement down
• 17:37
  Speaker 2: in costs of implementing AI.
• 17:40
  Speaker 6: Yeah, the way I tend to think about it is
• 17:43
  Speaker 6: the cheaper it is to train AI systems, the more
• 17:47
  Speaker 6: advanced capabilities you can train.
• 17:49
  Speaker 4: In the stress you lost your audio Josh, you there, Yes,
• 17:52
  Speaker 4: I am.
• 17:52
  Speaker 5: Here, still hear me. We're good?
• 17:54
  Speaker 6: Okay, perfect? Yeah. What the way I like to think
• 17:57
  Speaker 6: about it is the cheaper it is to train these
• 18:00
  Speaker 6: AI systems, Then the more advanced capabilities you can train,
• 18:04
  Speaker 6: the more advanced capabilities you can train, the more use
• 18:07
  Speaker 6: cases you unlock. The more use cases unlock, the more
• 18:10
  Speaker 6: you actually see it accelerate. So I actually think this
• 18:12
  Speaker 6: model will probably cause an acceleration in the capabilities of
• 18:16
  Speaker 6: other models that are getting built as more folks start
• 18:18
  Speaker 6: to adopt it. So for me, I actually see it
• 18:20
  Speaker 6: as completely positive.
• 18:22
  Speaker 4: What outside of deep Seek and other models, like, have
• 18:26
  Speaker 4: you worked with that you did?
• 18:27
  Speaker 6: Like yeah, so, I mean behind the scenes, we work
• 18:31
  Speaker 6: with in thropic models we work with open AI models.
• 18:34
  Speaker 6: We work with a bunch of fine tune models that
• 18:36
  Speaker 6: we build in house ourselves. You know. I like to
• 18:40
  Speaker 6: sort of describe it as an orchestra of different types
• 18:42
  Speaker 6: of models. And one of the things we've sort of
• 18:44
  Speaker 6: noticed is there's a lot of differences in capabilities. So
• 18:48
  Speaker 6: something like the inthroduct models and thropic models tend to
• 18:50
  Speaker 6: be better at very long context window where you try
• 18:53
  Speaker 6: and handle like huge amounts of text, whereas something like
• 18:56
  Speaker 6: the GPD form models tend to be a little bit
• 18:58
  Speaker 6: better at like foreign language and just sort of the
• 19:03
  Speaker 6: general verbiage it uses as it's giving outputs. So we
• 19:06
  Speaker 6: use a whole bunch.
• 19:08
  Speaker 2: Where how are your clients and the asset management business,
• 19:12
  Speaker 2: Josh using AI these days, in these early days.
• 19:15
  Speaker 6: Yeah, So the way I like to think about it
• 19:17
  Speaker 6: is we give them the ability to create this sort
• 19:18
  Speaker 6: of team of little AI workers where you teach them
• 19:22
  Speaker 6: how to do some kind of task, and then they're
• 19:24
  Speaker 6: going to do that task on a continuous basis. So
• 19:26
  Speaker 6: let's say you to do something like write an investment
• 19:29
  Speaker 6: thesis on a company, or let's say you had to
• 19:31
  Speaker 6: do something like write an ESG report or Let's say
• 19:34
  Speaker 6: you wanted to continuously monitor the world for any kinds
• 19:38
  Speaker 6: of updates to something that might happen in the air space.
• 19:42
  Speaker 6: These are all examples of sort of workflows you can
• 19:45
  Speaker 6: teach the system and have the system start to automate.
• 19:48
  Speaker 4: Well, we really appreciate your time. It was so good
• 19:50
  Speaker 4: to get that perspective. It's good to kind of get
• 19:52
  Speaker 4: the user mindset in for this. Josh Josh Pantoni, CEO
• 19:55
  Speaker 4: a boosted dot AI on deep Seek and sort of
• 19:59
  Speaker 4: the pros and the hans there.
• 20:01
  Speaker 1: This is the Bloomberg Intelligence podcast, available on Apple, Spotify,
• 20:06
  Speaker 1: and anywhere else you get your podcasts. Listen live each
• 20:09
  Speaker 1: weekday ten am to noon Eastern on Bloomberg dot com,
• 20:13
  Speaker 1: the iHeartRadio app, tune In, and the Bloomberg Business app.
• 20:17
  Speaker 1: You can also watch us live every weekday on YouTube
• 20:21
  Speaker 1: and always on the Bloomberg terminal.