DiscussionResearch

AI’s Next Winners: Signals From Private Markets

Thoughts on the Market10m 8s

Morgan Stanley researchers discuss findings from their Spark private company conference, revealing that AI startups are scaling at unprecedented speeds with tangible customer returns, while identifying key distinctions between coding/inference tools and application software in terms of growth trajectories and competitive defensibility.

Summary

The podcast features three Morgan Stanley strategists—Sabrina Howe, Adam Wood, and Brian Nowak—sharing insights from their Spark private company conference in California, where over 100 startups across software, robotics, and AI infrastructure were represented. The discussion centers on three major themes: growth velocity, the evolution from AI assistance to autonomous agents, and competitive moats in an AI-driven world.

On growth, the speakers highlight extraordinary scaling rates among private AI companies, with some approaching $4 billion in annual recurring revenue in two years and others reaching $1 billion run rates in under a year. However, Adam Wood cautions against simply contrasting private versus public company growth, instead distinguishing between two software categories: coding and inference tools (building blocks that can be constructed and adopted quickly while companies are still building bespoke solutions) versus application software (which requires complete business solutions, data integration, workflow understanding, and controls—taking longer to build and implement). He notes that fast-growing AI divisions within large public software companies may be obscured by slower-growing incumbent businesses.

Regarding agents, Brian Nowak discusses the shift from AI assistance to agents capable of executing work, sizing addressable markets at $30-40 trillion on both consumer and enterprise sides. On the consumer side, early leaders like Muse are emerging, with broader adoption expected in 2027. Enterprise agents are developing both as vertical-specific solutions (legal, healthcare) and as broad-based offerings from larger private companies.

On competitive moats, Adam Wood introduces a "moat and journey" framework distinguishing between the durability of existing businesses and the speed of AI transitions. Fast-growing coding and inference companies face questions about defensibility despite rapid monetization, while embedded incumbents with complex workflows are harder to displace but face greater demands in introducing reliable AI. He emphasizes that performance, model customization, and integration—not just model access—create differentiation, exemplified by legal AI companies hiring forward-deployed lawyers to understand workflows deeply.

The speakers note that models are increasingly becoming table stakes rather than competitive advantages, with value shifting to orchestration layers, data, workflows, and implementation expertise. They identify opportunities for software companies to help customers route tasks to appropriate models cost-effectively. Additionally, Wood suggests a potential gross margin tailwind if inference costs fall significantly while software companies maintain pricing, contrary to prevailing assumptions about margin pressure.

Finally, Nowak argues that competitive advantages may increasingly reside at the intersection of digital and physical worlds—in logistics, manufacturing, robotics, agriculture, and defense—where AI's physical manifestations through drones, warehouse robots, and agricultural tools will drive substantial business transformation over 2-5 years.

About this episode

<p>With AI startups scaling at record speed, our analysts Sabrina Hao, Adam Wood and Brian Nowak share takeaways from Morgan Stanley's Spark Conference on AI returns, competitive moats and implications for public markets.&nbsp;</p><p>Read more <a href="https://www.morganstanley.com/insights?cid=mg-SM_CORP-insights-17607" rel="noopener noreferrer" target="_blank">insights</a> from Morgan Stanley.</p><p><br /></p><p>----- Transcript -----&nbsp;</p><p>&nbsp;</p><p><strong>Sabrina Hao:</strong> Welcome to Thoughts on the Market. I'm Sabrina Hao, Morgan Stanley's AI and Tech Thematic Equity Strategist.&nbsp;</p><p><strong>Adam Wood:</strong> And I'm Adam Wood, Head of U.S. Software Research at Morgan Stanley.&nbsp;</p><p><strong>Brian Nowak:</strong> And I'm Brian Nowak, Head of U.S. Internet Research.&nbsp;</p><p><strong>Sabrina Hao:</strong> Today, what private companies are telling us about the AI opportunity and what it means for public markets.&nbsp;</p><p>It's Tuesday, October 20th, at 10am in New York.&nbsp;</p><p>Last week, we were at our Spark Private Company conference in California, bringing together more than a hundred startups across software, robotics, AI infrastructure, and beyond. There was even a company helping to manufacture pharmaceuticals in space.&nbsp;</p><p>It was a really interesting week to be at the conference given everything that was happening in the market. But really, we came away feeling even more positive about the AI cycle.&nbsp;</p><p>While public market investors are waiting for returns to justify trillions of dollars in AI spending, we're already seeing that pay off in private markets. AI demand is real, companies are delivering tangible returns for customers, and startups are scaling faster than ever.&nbsp;</p><p>Adam, let's start with growth. We heard from startups scaling at extraordinary speeds. One was approaching [$] 4 billion in annual recurring revenue in two years, another reaching a billion in under a year.&nbsp;</p><p>What really stood out to you about the pace of AI adoption, and why aren't we necessarily seeing that same growth at established software companies?&nbsp;</p><p><strong>Adam Wood:</strong> So yeah, absolutely, the pace was remarkable. And clearly one of the standouts from the conference when we used to discuss how quickly a company could get to a $1 billion so-called unicorn valuation. And now we're discussing how quickly they can get to a unicorn revenue run rate of a billion dollars, which for me is quite incredible.&nbsp;</p><p>But I think I wouldn't make the debate simply about private versus public companies. I'd make the distinction between two different types of software.&nbsp;</p><p>I think we have to make a distinction between largely coding and inference tools. There were a number of coding and inference software companies at the conference, and application software. And I think there's a big difference in terms of time and complexity to bring these products to market.&nbsp;</p><p>I think the coding and inference, which tends to be the building blocks, as I would call them, underneath the application layer, they can be constructed much more quickly. And I think because we can adopt when we're still building bespoke, which I think is a lot of the phase we're in today, we're still building bespoke software rather than buying off the shelf. They can be adopted earlier as well.&nbsp;</p><p>So, we've got faster time to build and faster time to adopt by companies. So, I think it's natural you can see very fast growth in those types of businesses very quickly.&nbsp;</p><p>I think in the application landscape, it's a little bit different because those applications have to deliver complete business solutions for customers. In finance and customer service, that means integrating data, understanding workflows, and putting controls in place.&nbsp;</p><p>There's much more to build and much more to implement, and generally, the adoption of those tools takes place a little bit later.&nbsp;After we've gone from the bespoke phase where companies are building themselves, we go to a phase where companies see that there are off-the-shelf packages that are available, that companies have built out.&nbsp;</p><p>So, I think it's both a complexity of build debate and also, a time in terms of where we are in the software adoption cycle.&nbsp;</p><p>The second thing I'd say is, you know, in public software companies that have a big incumbent existing business, sometimes there can be that big incumbent business that's not growing very quickly. And a very fast-growing AI business in there as well. And that very fast AI growth business can be hidden a little bit by that large public software company that exists.&nbsp;</p><p>So, I don't think it's as straightforward as just saying, you know, private's growing very quickly, public not. I think there's a couple of layers of complexity in there in terms of what type of software you're dealing with. And then how big the existing kind of business is in a software company and what is kind of existing software growing a little bit less quickly. And then the AI bit that could potentially be growing very fast.&nbsp;</p><p><strong>Sabrina Hao:</strong> Absolutely. It makes a ton of sense. Maybe on this theme of full stack solutions, you know, Brian, we heard about the shift from AI assistants to agents that can actually execute work. You've done a lot of work on this, both across consumer and enterprise.&nbsp;</p><p>What did you learn about how quickly agents are advancing? And how does it shape the outlook for the internet names that you cover?&nbsp;</p><p><strong>Brian Nowak:</strong> This is a very exciting time for agentic adoption on both the consumer side and the enterprise side. You know, we've sized potential addressable markets of $30 to $40 trillion of spend on both the consumer side and the enterprise side that we think can be addressed by agentic tools. On the consumer side, people are shopping, they are planning their travel, they're planning their holiday shopping. On the enterprise side, we're seeing both vertical specific agents start to emerge. So, agents within the legal area or agents within the healthcare area. And then we're also seeing broad-based agents from some of the larger private companies also start to bud through.&nbsp;</p><p>So, there is a smattering of agents that are popping through. On the consumer side, Muse is your early leader, and we do expect more to come in 2027. And then I think on the enterprise side, they're both vertical specific as well as broader based agents expected to grow.&nbsp;</p><p><strong>Sabrina Hao:</strong> Agents really raise an important question around disruption. Maybe for Adam, you've written about the moat and the journey in software. It's not just about if a company has a competitive advantage, but how it navigates that transition to an AI-driven world.&nbsp;</p><p>Did Spark change your thinking about which software incumbents are best positioned to defend their businesses or capture new growth?&nbsp;</p><p><strong>Adam Wood:</strong> So, I think Spark really confirmed a lot of our views on that moat and journey framework that we've put forward. The moat protecting the existing business that you have and the journey turning AI into new growth opportunities. And in some ways, the deeper the moat for businesses, the longer that journey can be.&nbsp;</p><p>An incumbent embedded in complex mission-critical workflows is hard to replace because of the data integrations and customer trust it has by definition. But those same requirements can make introducing AI into its platform more demanding. You can't compromise the reliability of that work.&nbsp;</p><p>So that's why we've separated the durability of the business from the speed of its AI transition.&nbsp;</p><p>For some of the coding inference businesses we heard from – again, these incredibly fast run rate, fast growth businesses, the monetization has obviously moved very, very quickly. And I think there the investor debate's going to be a little bit the opposite way around. How durable and how big a competitive advantage have these companies built on that very fast-growing revenue base?&nbsp;</p><p>And that's not to say these businesses can't have moats. They can. You know, performance, model customization, and integration can create real differentiation. But I think there is a big difference to be made between rapid growth and defensibility.&nbsp;</p><p>And, you know, to give a couple of examples here in terms of the moat that some of these AI native vertical and business software companies are leveraging. You know, there's a legal AI company that's hiring former lawyers to become forward deployed lawyers.&nbsp;</p><p>We had forward deployed engineers, now we've got forward deployed lawyers. But the reason there is they bring that understanding of legal workflows to the product and its implementation.&nbsp;</p><p>So, the value isn't just access to the model. It's making the model and the technology work in the customer's business and understanding in depth those business workflows.&nbsp;</p><p><strong>Sabrina Hao:</strong> And we've talked about this theme of AI adopters, and really what we need to get this trade to work are those quantifiable benefits from AI. So open weights are a key part of that.&nbsp;</p><p>Maybe to pull on another thread, and you, kind of, alluded to this, the models themselves are increasingly becoming table stakes rather than moats. And the value is shifting to everything around the model – whether that's your data, your workflows, the tools that make AI useful.&nbsp;</p><p>So where did you see this most clearly at Spark? And does it change your view on where durable value will accrue in software?&nbsp;</p><p><strong>Adam Wood:</strong> So again, I mean, I think I'd come back to the growing importance of open weight models. I mean, it was a recurring theme at the conference. Almost every company spoke to us and spoke to, you know, the implementation. And again, it reinforces that view of a multi-model world.&nbsp;</p><p>You’ve got the inference companies talking about bringing down the costs of using the models. Again, that diffusion concept we talk about. At the application layer, a legal software company said it benchmarks 18 models across 200 criteria to decide which model should run the task. So, you know, you're getting incredible gains in efficiency.&nbsp;</p><p>I think there are two implications for public software vendors. I think, you know, first of all, there's an opportunity here for software companies to help their customers with orchestration. So, there's going to be a layer around how do we get the right task to the right model at the right price. And I think a lot of the public software companies, you know, at the hyperscaler level, they can help at a broader enterprise adoption level.&nbsp;</p><p>And then down at the application layer, whether we're talking about CRM, ERP, HR, there's an opportunity for software companies to help with orchestration at that layer as well.&nbsp;</p><p>I think the second thing, which is probably a little bit less talked about, is maybe there is a gross margin opportunity here. I think in software, all we've discussed so far is the pressure that's going to come on gross margins from having to include the cost of running tokens in a software business model and offer that to customers as part of the package.&nbsp;</p><p>If we can bring down inference costs dramatically, and you can hold on to some of the pricing as a software company – maybe there could be a little talked about bull case in the medium term. That maybe software margins are not going to be as pressured as people think. And we can maybe see gross margins hold in a lot better than people think.&nbsp;</p><p><strong>Sabrina Hao:</strong> Brian, I want to end with something that may seem counterintuitive. In a world where software's becoming easier to build, some of the strongest competitive advantages may actually be in the physical world.&nbsp;</p><p>The conference had many companies, whether that was logistics, manufacturing, real-world operations. You know, why is that intersection of the digital and physical world so defensible? And how does it influence your view of who ultimately wins in an agentic world?&nbsp;</p><p><strong>Brian Nowak:</strong> I think it's important for people to remember that AI is not just software. AI is going to get physical through new types of robots that impact the entire globe. We're seeing it in multiple industries already. We're seeing it on the defense side with drones and other types of mechanisms that are going to change the way that we build out our overall defense tools. We're seeing it on the industrial side, in warehouses, on delivery. We're seeing it on the agriculture side, the way that we're going to the future of farming.&nbsp;</p><p>And so over the long term, while I think we will have humanoids and we will have robots in our homes, I think it's important for people to keep in mind there are going to be a lot of physical and robotic implementations of AI over the next two, three, five years that are going to really change the way we do a lot of business.&nbsp;</p><p><strong>Sabrina Hao:</strong> Absolutely. Lots of exciting things happening. Adam, Brian, thanks for joining me. And thank you for listening. If you enjoy Thoughts on the Market, please leave us a review wherever you listen and share the podcast with a friend or colleague today.</p>

Key Insights

  • Coding and inference software companies are scaling to $1 billion revenue run rates in under a year because these building-block tools can be adopted during the bespoke software phase when companies are still constructing custom solutions, enabling both faster construction and faster adoption compared to complete application software solutions.
  • Fast-growing AI businesses within large public software companies with substantial incumbent operations may have their growth obscured in financial reports, making the comparison between private and public company AI growth less straightforward than it appears on the surface.
  • The debate around fast-growing AI companies is shifting from 'how quickly can they grow' to 'how durable and defensible are their competitive advantages,' requiring investors to distinguish between rapid growth and actual defensibility through model customization, performance, and deep workflow integration.
  • Models themselves are becoming commoditized table stakes rather than competitive moats, with durable value accruing instead to orchestration layers that route tasks to the most cost-effective model, data assets, workflow understanding, and implementation expertise like forward-deployed lawyers in legal software.
  • Physical implementations of AI through robotics, drones, and autonomous systems across defense, industrial, warehouse, agriculture, and delivery sectors represent some of the most defensible competitive advantages over the next 2-5 years, as these require both digital expertise and real-world operational constraints that are difficult to replicate.

Topics

AI startup scaling speeds and growth trajectoriesSoftware categorization: coding/inference tools vs. application softwareAI agents and autonomous work executionCompetitive moats and defensibility in AI-driven marketsModel commoditization and value shift to implementationIntersection of AI, robotics, and physical operations

Transcript

Welcome to Thoughts on the Market. I'm Sabrina Howe, Morgan Stanley's AI and tech thematic equity strategist. And I'm Adam Wood, head of US software research at Morgan Stanley. And I'm Brian Nowak, head of US internet research. Today, what private companies are telling us about the AI opportunity and what it means for public markets? It's Tuesday, October 20th at 10am in New York. Last week, we were at our Spark private company conference in California, bringing together more than 100 startups across software, robotics, AI infrastructure, and beyond. There was even a company helping to manufacture pharmaceuticals in space. It was a really interesting week to be at the conference given everything that was happening in the market,…

Full transcript available for MurmurCast members

Sign Up to Access

More from Thoughts on the Market

Get AI summaries like this delivered to your inbox daily

Get AI summaries delivered to your inbox

MurmurCast summarizes your YouTube channels, podcasts, and newsletters into one daily email digest.