The Strategic Battle for AI Supremacy Among Tech Titans
Apr 14, 2024
Evan Kirstel
Explore the cutting-edge with Michelle Brophy of AlphaSense as we dissect the evolving landscape of AI and cybersecurity investments. This episode takes you on a journey through the technological advancements propelling the industry forward, where generative AI isn't just a buzzword but a transformative force reshaping how businesses operate. Michelle, with her deep understanding of tech trends, sheds light on crucial infrastructure upgrades and the race among tech giants like NVIDIA, AMD, and Intel to dominate the accelerator market. We also dive into the cloud's expansive role in AI, discussing how its growth is not just inevitable but essential in supporting complex AI applications.
Venture into the strategic realm of big tech's influence on AI innovation, where companies like Amazon and Apple aren't just participants but pioneers charting the course of the future. In this episode, we unravel Amazon's multifaceted AI strategy, from their Trinium chip endeavors to bold investments in AI startups, offering insights into their vision of consumer choices and technological breakthroughs. Meanwhile, Apple's flirtation with Google's LLM Gemini signals a potential shift in collaboration and competition. Beyond the sparkle of innovation, we confront the gritty realities that enterprises face in proving the value of AI amidst data privacy concerns and cybersecurity challenges. Join us for a compelling narrative that dives into the intricacies of AI's impact on business and the global tech ecosystem, guided by Michelle's expert analysis.
Explore the cutting-edge with Michelle Brophy of AlphaSense as we dissect the evolving landscape of AI and cybersecurity investments. This episode takes you on a journey through the technological advancements propelling the industry forward, where generative AI isn't just a buzzword but a transformative force reshaping how businesses operate. Michelle, with her deep understanding of tech trends, sheds light on crucial infrastructure upgrades and the race among tech giants like NVIDIA, AMD, and Intel to dominate the accelerator market. We also dive into the cloud's expansive role in AI, discussing how its growth is not just inevitable but essential in supporting complex AI applications.
Venture into the strategic realm of big tech's influence on AI innovation, where companies like Amazon and Apple aren't just participants but pioneers charting the course of the future. In this episode, we unravel Amazon's multifaceted AI strategy, from their Trinium chip endeavors to bold investments in AI startups, offering insights into their vision of consumer choices and technological breakthroughs. Meanwhile, Apple's flirtation with Google's LLM Gemini signals a potential shift in collaboration and competition. Beyond the sparkle of innovation, we confront the gritty realities that enterprises face in proving the value of AI amidst data privacy concerns and cybersecurity challenges. Join us for a compelling narrative that dives into the intricacies of AI's impact on business and the global tech ecosystem, guided by Michelle's expert analysis.
More at https://linktr.ee/EvanKirstel
Speaker 1:
Hey everybody. Super intriguing discussion today on where AI spending goes next with a true insider and expert in the field, michelle from AlphaSense. How are you?
Speaker 2:
I'm doing great. Thank you for having me today, Evan.
Speaker 1:
Well, thanks for being here, Really intrigued by the expertise and insights that you and the firm provide. Before we dive in with lots of questions and your thoughts, maybe introduce yourself and your role at AlphaSense and who is AlphaSense for those who may not be familiar.
Speaker 2:
Yeah, of course. Let me start by letting you know my name is Michelle Brophy. I am the Director of Research for TMT with AlphaSense. I've spent the better part of my career as an investor and an analyst on the buy side of Wall Street about 10 years with a long short equity hedge fund in New York City as a senior analyst covering technology and following that with a family office as a technology portfolio manager and senior analyst, and particularly as a former buy-sider. I'm really excited to be part of AlphaSense, as I've been really pretty inspired by the power of the AI platform and, as you mentioned, Evan, for those of you who may be unfamiliar with AlphaSense, AlphaSense is a AI-powered market intelligence and search platform that helps enterprise customers.
Speaker 2:
We have about 4,000 enterprise customers at this point, Evan, so including the majority of the S&P 500. And our goal is really to help them find the answers they need to make better decisions quickly and confidently. The company started in 2011, actually, and we're up to 1,300 employees worldwide. We cover ranging industries beyond technology energy, industrials, consumer goods, just to name a few and just an exciting plug for everyone, we just crossed the $200 million revenue mark this week, and so that's really exciting for our company's history. So, AI-powered market research, it's a really exciting place to be.
Speaker 1:
Wow, you're in the middle of it all now with all the excitement around AI, investment and spending something you spend a lot of time thinking about. Maybe tell us about what's top of mind for you these days around AI and the issues that are top of mind to your clients?
Speaker 2:
Yeah, as Director of Research for Technology, I head one of the areas of our company called Expert Insights, and those insights are used to uncover some proprietary data points, as you mentioned, for more of the strategic business decisions, and being in the thick of it in AI and hearing from CISOs and CIOs and CTOs in terms of what they're thinking, in terms of what they're planning, is really the place to be and how we really glean some really interesting information. So, on this topic, the first thing I want to say to you about generative AI, evan, is, from our experts' opinions across the board, generative AI is really here to stay. We can argue about the hype and the reality and what stage we're in. We're in early stages, in my view. The competitive landscape is increasing, but so is the pace of innovation and development, and the rapid pace at which constraints are being put on existing infrastructure and security offerings is fairly evident to us right now.
Speaker 1:
Yes, indeed Exciting time, so let's dive right in. I mean, where should company business leaders focus their spending in areas like AI, gen, ai, to enhance their capabilities, let's say efficiencies, revenue, on and on.
Speaker 2:
I think the first place that we need to talk about is really infrastructure. People talk about the first tranche of infrastructure really being in place at this point. The big name is, of course, nvidia, but there's a lot more that's being considered. Name is, of course, nvidia, but there's a lot more that's being considered, and we've seen that in the last earnings reports, especially with names like SMCI and Dell coming up big in the hardware category. Our experts are actually saying that hardware spend is going to be 50 to 60% of spend in the next few months or a year. So that's kind of an interesting you know, when we first started talking about AI and the hype and it came out like in December of 2022, it was all about how it was going to change the workplace automatically and everything was going to just automatically change. The truth of the matter is that the existing infrastructure out there needs to be improved and upgraded, and so that's why this picks and shovels trade seems to be working at the moment is because those companies are really making money and they're really setting the tone for what the future will look like, and that's why sort of reiterate, it's not going to be at such a hype cycle as some may think it's going to be. Our experts really believe that the setup is there for a big tailwind in the future.
Speaker 2:
I'll start off by saying NVIDIA is clearly the leader. We're talking about a $90 billion calendar 2024 accelerator market. Estimates have put that at $200 billion for 2027 calendar year. Our experts see the competitive field evening out over several years, with hyperscalers also trying to build out their in-house designs. You hear about a number of companies out there big tech, amazon, trinium and Infringia chips. You've heard just this week Google's improved their own TPUs, which is a similar offering as an NVIDIA GPU. For those who are unaware, meta's announced their MTIA and Microsoft is out with their Maya chip. You've also got the traditional companies like AMD and Intel introducing competitive offerings and it remains to be seen whether those companies are going to catch up with NVIDIA. But the market cap is there and we're talking about TAMs of $200 billion in just a couple of years for the chips alone. So I'll stop there if you want to follow up on any of the chip discussion, and beyond that there's. There's obviously more to talk about on the hardware side.
Speaker 1:
Yeah Well, it's so much to unpack, so fascinating to get the nail. Sketch Enterprises, as you know, have been on a tear the last decade, plus on cloud investment, public cloud in particular. Do you see an increase on the cloud spend for AI related applications as well?
Speaker 2:
applications as well. Yeah, 100%. I mean, I think that it's been pretty clear from the big tech earnings report that the 2023 story of optimization has broadly passed us. What we're now talking about a reacceleration of that cloud spend overall, but once again, I think a lot of it comes down to cloud infrastructure and what we're playing on.
Speaker 2:
Microsoft and OpenAI just came out with this plan to spend $100 billion on data centers, calling it a Stargate supercomputer. I don't know how I feel about that term and whether or not that's really the way we're going, but Sam Altman's also talking about starting his own AI semiconductor startup and there's a number of other announcements out there that really point to, you know, a real intangible tailwind in this group. But once again, I point back to things like, you know, high band memory and servers and semiconductors outside of NVIDIA really seeing a tailwind as sort of this picks and shovels play outside of NVIDIA really seeing a tailwind as sort of this picks and shovels play, where you know we really need to see the infrastructure improve before the productivity and efficiency gains happen in the cloud, but that spending is certainly coming.
Speaker 1:
Interesting to see. So I've been spending a lot of time thinking about cybersecurity. With RSAC coming up in about a month, the big industry get together and AI is changing everything on the defensive side, and the bad actors, of course, have this technology in hand as well. But what do you think about cybersecurity spending related to AI, maybe protecting AI infrastructures? There's probably going to be a big investment there as well.
Speaker 2:
Yeah, I agree with you, it's going to be a big investment, no matter what, whether it be in the cloud or the hardware. It seems like everyone's sort of coming to the realization that not only do regular infrastructure services need to be upgraded, but the security along with that needs to be upgraded. There's a couple of things I will say. Cybersecurity seems to be presenting a tale of two cities for AI. On one hand, we have AI applications increasing the shot fund goal by bad actors, as you mentioned, and breach headlines, by the way, also out there in an increasing clip, but it's also paving the way for a flourishing period.
Speaker 2:
Of three key areas of potential focus for cyber with respect to AI applications, the number one I would name to you, evan, is data leakage. It seems that data accessed by the models, data that may be leaked enterprises really need to be concerned with a way to create classification and cataloging standards to ensure that that data is appropriately protected really, and accessed by those Gen AI apps. That's a real area of concern and focus for the companies that we've been talking to. I would also say number two is data lineage. Number two is data lineage. The conversations that are happening with our CISOs and CIOs seem to point to partners and customers talking a lot about how to delineate inputted data versus outputted data with respect to the future of Gen AI applications.
Speaker 2:
Lastly, I'll just say, on AI security, overall, there's concerns about, you know, just protecting that surface area of Gen AI. It's on the rise. You know ongoing conversations about how to protect those attackers that are coming through. So those are the three areas that we've identified so far in terms of being the key areas of focus, but the bottom line is the threats are really changing and the ability to mass customize is going up, and so the increase in frequency and capability to mount attacks is also going up. On the flip side, again, security companies can utilize AI and do more with fewer people. So it remains to be seen how that shakes out in the longer term, but there's really interesting things happening in this area.
Speaker 1:
Yeah, I know it's fascinating to watch and we're seeing. You know some industries adopt AI technology more aggressively already than others, and you know we've seen in the past certain industries kind of fall behind in adopting cutting edge tech. I think of healthcare, personally, where, despite all the amazing science and medical advances, you know we still use fax machines and clipboards at our local doctor's offices. But so what industries do you think are at risk of disruption if they just ignore this whole AI investment?
Speaker 2:
Honestly, I think they all are. I mean, I think that, truthfully, you know the crux of AI can really improve, you know, any vertical, from healthcare on down the line. I will say that healthcare applications seem to be the most exciting, especially being able to being able to, you know, discover new genes, for instance. Are it's like 20% productivity increases in terms of being able to run the data faster and find the anomalies within the data. That's what our experts are saying. Anyway, as preliminary as it is, I would say that any evidence that's being able to be harvested in mind from the healthcare sector is going to be something that's going to see a vast improvement from AI applications, for sure. I think some of the areas you may see some slower adoption is areas like financial services and insurance companies, where you have some more risk involved. We haven't quite worked out some of the privacy issues involved with generative AI and I think there may be some slower adoption in terms of the consumer or maybe those companies that touch consumer with more intense privacy requirements.
Speaker 1:
Yeah, for sure, great, great point. It seems every big tech company has gone all in on AI and investment and innovation. It's had Google Cloud Next just wrapped up this week. Amazing news all week. Yeah it's been amazing to watch, both as a user, customer of Google services, but also just as an observer of the industry. But who do you track as the key players leading AI innovation? Is it really the big tech companies?
Speaker 2:
I think we're seeing big tech really leading the way because they have the cash right. I mean, at this point, we're not talking about, we're not talking about, you know, really nascent players unless they're getting the funding. You know, I think something like 90% of the private investments are coming from the corporate level. So, you know, the startups are really getting funded by the Googles and the Amazons of the world. Big tech looks like they're trying to find their lanes, but they're also still looking for optionality.
Speaker 2:
I'll talk about Amazon for a second. As we're talking about that, right, they're home growing their own chips in Trinium and they're also looking at Bedrock. But, as they're looking at their customer needs, they're also investing in Thropic. So they're still doing $4 billion investment with an outside LLM. Rather, you know, the reason being is Andy Jaffe said it yesterday on in an interview he wants to keep optionality for his customers and let the customers decide what they prefer at this moment. So it seems like we're at a key juncture of that innovation really taking place. But big tech's able to lead the way because they're able to invest in-house and out-house in a way, because they can afford to. So it's really a tide lifting all boats at this point in terms of who's innovating, but the cash is definitely coming from big tech.
Speaker 1:
For sure, and do you see a lot of collaboration there continuing with big, with big tech and startups, and is is that a key requirement for this space to really get off off the ground?
Speaker 2:
I think it is a key requirement. I think there could be an issue we run into from a regulatory perspective as things mature. You know, I don't think that any of the big tech firms can go out and acquire these people under. You know these companies. Rather, under current circumstances, I think there would be some anti-competitive issues to deal with in terms of in terms of the current regulatory environment in the United States. But it really remains to be seen whether or not, whether or not they'll they'll stay friendly in terms of partnering. I mean, you'd have to imagine that at some point, when there's more competitive pressure as the industries mature, perhaps they won't be so friendly. I mean, I'm just thinking back to the Apple days before. They were really the main competitor in the mobile space. Right, they were friendly with everybody, but that wall tends to close as things get more competitive.
Speaker 2:
I've been thinking about this with respect to a report that came out a couple of weeks ago that potentially Apple would go with Google's LLM Gemini for their next iteration of iPhone, and it's perfectly to your question. I mean, will they partner with Google? Would they partner with you know someone who is their competitor on something so important as an LLM? And I guess the other prong to that question is well, it depends on how far behind they are the curve. We don't really know. We're never going to really know what's happening at Apple until something happens.
Speaker 2:
So our experts are opining on it. I've been asking them myself. I'm super fascinated by what's going to happen next. For those who are listening, ai at the edge really means AI over your device. We're currently talking about AI applications happening in a cloud environment and things really change if we're starting to talk about a mobile climate, and Google seems really well set up with Gemini and their LLM efforts there. They've been working with Samsung and Android models and it seems to be moving forward, and we don't really know how far behind Apple is. Albeit they have their own chips in-house that seem to be able to handle the traffic in-house that seem to be able to handle the traffic. It stands to reason that they may be behind, since they've been focused more on the auto market.
Speaker 1:
Yeah, it would be fascinating to watch particularly as an Apple customer and fan to see how that evolves. You talk to so many different enterprises out there. What are the main challenges and roadblocks you see in the enterprise with AI adoption and deployment at the moment?
Speaker 2:
I think the number one main challenge for enterprise. Well, there's two key challenges for enterprise. Right, there's there's ROI is number one. We've been in an 18 month cycle of what LLMs can offer the AI world, but it's expensive and it hasn't gotten any less expensive, particularly due to hardware constraints. So enterprise, I think it's really struggling with how they can show value somewhere along the line.
Speaker 2:
We've seen it in the last couple of earnings periods where, conceptually, it's been interesting and a lot of companies are out there talking about roadmaps but they're not monetizing quite yet and I think the general consensus is they will mostly, but I think the investor base may run out of patience in the meantime for some of the companies somewhere along the way. So there's a bit of a dichotomy going on between what investors are seeing on the bottom line versus what is being invested into this world. But I do think it's super exciting and those who have genuine roadmaps that they can articulate clearly in the next 12 to 14, 12 to 24 months are really going to stand out from those who can't. That's number one.
Speaker 2:
Number two is privacy. I think enterprise is just really concerned with uptake and what it means for their internal security. Can they have their employees exposed to data leakage? Will they have all of this in the cloud? Will it be a hybrid model? Are they going to go on-prem with some of the data? These are things that I think that enterprise is really dealing with right now. In terms of where the data resides, I think that we're leaning more towards some of it will be in the cloud and then some of the more intricate company sensitive information will be perhaps on smaller LLMs within their own enterprise setting. So I think we are moving kind of an interesting way. So it'll be. It's kind of happening right now, in my view, of a hybrid, more hybrid environment where enterprise can feel safe that their data is safeguarded, but they are getting some of the elements of AI.
Speaker 1:
Yeah, I would agree. As a small business owner myself, I've used 12, 15 different tools already, so I'm already spending in the thousands on AI technology. So I can only imagine that being multiplied by millions of businesses. So very exciting. And speaking of AI adoption, you at AlphaSense talk about incorporating AI into your practice, and so how does that work? What does it mean for you and your colleagues and your customers?
Speaker 2:
Yeah, we deploy AI applications on a range of items.
Speaker 2:
We've been moving aggressively to deploy AI for our customers in terms of our search capabilities.
Speaker 2:
So if you think about our expert insights, you know we do a lot of ETL work, expert transcript, library work, where we're building our library, amassing a huge amount of data, and our customers are able to search and have that functionality and look for the insights they need.
Speaker 2:
Outside of the expert insights arena, alphasites in general caters to a broad array of customers, and so we also offer filings and private company reports and press and broker research, and our clients are able to search all of it for whatever their market intelligence needs are. So it's a really powerful platform, whether you're an investor or you're on the corporate side, and you're able to really tap into whatever those sources of information you're looking for are whether it be a filing on Edgar. Tap into whatever those sources of information you're looking for are whether it be a filing on Edgar, whether it be a broker report from Southside or whether it be an expert transcript that you're looking for and therefore, and beyond that, you're able to tap into the expert library and then reach out to that expert yourself and conduct your own interview. So it's really a powerful way of market intelligence. Layered on all of that is AI capability. So things like smart summaries of transcripts are available, things like smart search are available A lot of.
Speaker 2:
AI capabilities that we offer are client-based.
Speaker 1:
That's phenomenal. I have to give it a look. Well, it's been very interesting and informative. I learned a lot just in this short 20 minutes. What are you looking forward to over the next few weeks, months, any events or meetups or other travel?
Speaker 2:
I'm looking forward to seeing what happens in June with Apple. I want to see if they're going to launch an iPhone with an LLM in it or some kind of Siri capabilities that embed AI functionality.
Speaker 2:
I think that's going to be really interesting to see. Beyond that, let's see what happens with the latest earnings reports. We're coming up on April earnings in May and, as I mentioned in our commentary, some of these companies are going to mature and do phenomenally and some of them may lag behind a little bit. So I've been really tracking it with a lot of enthusiasm.
Speaker 1:
Fantastic. Well, really a delight catching up and look forward to another chat in six months and see where we are in this industry. Thanks so much, michelle, and thanks everyone for watching.
Speaker 2:
Thank you, evan, appreciate it. One last plug for AlphaSense. I'll just encourage our viewers to check us out on Twitter and check out our website if you're interested in finding out more. So thanks so much for having me, evan.
Speaker 1:
I really appreciate it. I would agree. Yeah, you guys put out some very informative and insightful content, so very educational. Thanks, michelle.