AI Investor Outlook for 2026 and Beyond

Top 7 AI Bubble are as prevalent as ever, but real-world investors are still bullish. We're digging into the next phase of AI today on Motley Fool Money. Today's Tuesday, January 6th, Welcome to Motley Fool Money. I'm your host, Emily Flippin, and today I'm joined by full analyst, Austin Sharma, and the head of AI here at the Motley Fool, Donato Riccio, to discuss the investor outlook for AI and 2026 report. So I have you both on today because the Fool recently published an interesting report around real-world AI usage. I would you two were obviously integral to its creation. This report, which is called the Motley Fool's 2026 AI Investor Outlook Report, is available for free at full.com/research/AI-investor-outlook for anyone who wants to read it. But don't worry, we do have that link in the show notes for easy access, so you don't have to memorize it. But for anybody who can't read it or just has it yet, I'm really excited to dig into some of the findings here today on the Motley Fool Money podcast. And I want to start with what the report says about real-world investors and what they're doing with AI today. And then we'll move to where Donato, Donato, you think the industry is heading, and then a wrap with Austin's framework for investing in AI, including where the opportunities may be the most ripe. Now the Motley Fool's 2026 AI Investor Outlook Report did survey around 2600 American adults in November of 2025, and the headline is pretty simple, right? Amongst people who already own AI stocks, 36% planned to increase their holdings, 57% planned to keep the same, and only 7% planned to reduce. Moreover, a whopping 62% of respondents said their competent AI heavy companies will deliver strong long-term returns, and that number grows to 93% amongst those who already have exposure. So the gist of this report is, there's still a lot of excitement around AI, even with the hype. Now Austin, I know there's always going to be biases in this type of self-reported data, right? Those who are most excited about AI are probably also the ones who are most likely to respond to a survey about it, for instance. But when you see that people are largely holding or adding to AI in a world that continues to focus on the fact that we're in, quote, AI bubble, what does that tell you? And when I think it reflects a societal learning curve, I'd argue that most people and most investors are much more knowledgeable about the components of AI, machine learning, and generative AI, versus a few years ago. I guess that's obvious. We evaluate businesses in the space. I've noticed that most of us have acquired vocabularity we didn't have in, say, 2022. So we're familiar with terms like GPUs, LLMs, inference tokens, etc. So I think investors have this broad enough understanding to evaluate what type of bubble we're in. So we should spot the average investor some credit here. I think the decision to be invested or to stay invested has more reasoning and rationale behind it than previous bubbles that come to mind. Okay, so along these lines, the mania aspect of this bubble appears comparatively smaller to me against historical bubbles. I'm thinking about, let's say the dot com bubble in 1999, go all the way back to the tulip mania in what the 18th century, 17th century in Holland. That doesn't mean that this bubble isn't going to pop or at least deflate a bit. But investors seem to me like they're in this mode of evaluating the risks, the trade-offs. And they're more willing to demarcate like their personal lines that go between investing and speculating. All right. So here's this paradoxical question, which is sort of hinted at Emily. This gets to the surprising results of our survey. If you understand that we could be in a bubble and you already have exposure to the upside potential of AI and you understand that the market has appreciated for three straight years with a cumulative return of 78% and you know that the S&P 500, which is driven by big tech, currently sits at all time highs. Why would you be planning to add to your AI positions in 2026? And to me, I think it says, okay, number one, you've got an inherent belief that this technology is tied to the creation of value in the global economy, i.e., you believe it's for real. And number two, you think that some companies are going to continue to realize appreciable cash flows from selling either the development or the output of this technology. And you're also researching new opportunities. You're tuned to valuation in the businesses you own and the ones you want to buy. And finally, you intend to be rational in your capital allocation or is that hope of mine? No, I think that that's a fair read, Osset. And one of the things that we don't get from the survey is how much exposure already exists. We talk to people who say they already have exposure. But in terms of a total portfolio, that exposure could be smaller than what somebody may want to allocate. So the intention to add may just be actually building out what would then be a full-size position to exposure to AI. That's defined in 2026. But I also think, and this is maybe the irrational hope of mine, that anybody who answers this survey and says that they're planning on adding or maintaining their AI exposure is doing so with the awareness that I'm going to hold these companies for the extreme long term. So yeah, maybe this is a bubble. Maybe there are risks and we do have a crash, but that's okay because the companies I'm invested in have very real appreciable cash flow. And I believe that a decade from now, even if there is a short pullback and share price of a company, they're going to be bigger, better, more important businesses in the future. Maybe I'm giving too much credit here, but I sure love for investor. Yeah, that's where I hope we're going. Denado, I want to pass the mic to you because obviously you're the head of AI here at the Mollie Fool. And one of the things that I really liked about the report was that the optimism that Osages mentioned, and we talked about it wasn't totally blind to this report. When they asked about risks, the top two risks from respondents were things like data quality, security, as well as a sense of overvaluation in the sector. You're somebody who already spends all of your days living inside the world of AI, obviously. When you see this investor confidence shown in the report, is that matched by what you're seeing in terms of like real world adoption or as Wall Street still early to the party. So the short answer is that I think it matches, and we are currently in a healthier place compared to just six or nine months ago, because at the beginning of 2025, it was many others are starting to worry about, is it a bubble? But the main indicator I monitor is pretty simple. So our people's expectations connected to how the technology actually works, because when the expectations disconnect from the fundamentals, that's when you get a bubble, right? So early last year, I saw this starting to go sideways, because people were getting more and more excited about agents, which are elements that are able to perform more complex actions, and can go beyond just answering questions such as booking your flight or creating an act or matching your calendar. So many people started calling 2025 the year of agents, but I like Andrea Patis framing better, which is that this is the decade of agents, because they're just getting started, and this is an emerging technology. But at the time, the expectations were running way ahead of reality and people were imagining these autonomous entities that could do like everything and run your business alone. So yes, agents work, and they are at the new disruptive technology, that for now proved effective in very narrow scopes and controlled environments. Last year, I've served this gap between expectations to reality, but then two things happened. So first, the sentiment cooled down a bit. The hype around agents got more measured. I think, in fact, we are starting to get a little bit past big hype, because people are getting more realistic now about what agents can do, but who knows what's going to happen tomorrow, right? And the second thing is that the most important part is that agents actually improved dramatically. The technology really caught up with some of these expectations, not all of them, but I'd say enough that this gap narrows. So yeah, I say that we're in a healthier place. I like this direction. The market markets had reached new highs last year, but over the past couple of months, we have been we've been playing flat, and I think that's okay to give people and companies more time to play to experiment. And when we look at actual adoption data and companies, it confers in this direction. So the paid AI adoption across US businesses increased a lot from 2025 to 2023. It was just around 5% to 44% in September 2025. And if we look at revenue growth in AI companies and tropics reported 10x on the revenue two years in a row, cursor the AI coding tool is in a similar situation. Went for four million last year to hitting a billion analyze revenue this year. So I said this is not hype. There is real commercial structure in these tools and we are adoption companies. People are funding real value in these tools. So when you ask Emily if the investor confidence is matched by adoption, I say yes. And the data shows that companies are funding real value. It's almost ironic that we talk about AI as a bubble today when I think the skepticism around AI is probably the highest it's ever been. Unlike bubbles in the past, I think we as investors have a new level of awareness of the things like the hype cycle. And it's when to your point, when things get separate, right? When hype separates from reality, that's when it creates a bubble. But to your point, there is reality backing up a lot of this technology. And I love the fact that the survey shows that investors are still largely leaning in to the AI and adoption. But there's still that awareness, that cautious amount of optimism. Up next, we're going to be getting practical about 2026, including where the next wave of opportunities may show up. Stick with us. Welcome back to Motley Fool Money. Today, we're discussing the AI investor outlook report for 2026 and where AI technology may be headed. Donata as head of AI here at the full, I'd love to dig in a little bit deeper to where you see AI going. Now, you said in this report that the right mental model is somewhere like three to five years in terms of a timeframe for investors in the sector. And that investor shouldn't get too caught up in things like the present day cost of LLM since the intelligence per dollar ratio for models has been doubling roughly every six months. That goes over my head. So when I hear that coming out of your mouth, I mean, can you provide some more context as to what exactly that means? And if AI capabilities keep improving at that same price, where you expect value to accrue in the year ahead? Yeah, that's exactly right. So currently there's a lot of focus on the AGI. This is the big question. Well, will AI become super intelligent? I think this not always the right questions question for investors because it's really impossible to predict that. But I say that the more impactful and important question right now is that when does current level intelligence become cheap enough to be everywhere? And I think this is happening right now. So if we take a look at how costs evolve over the years, just two years ago, GPT-4, which was the flagship model by opening AI available at the time, costs 30 to 60 dollars per million tokens. A million token is around three, four books. And so you need 30 to 60 dollars to process this amount of information. But today you have available GPT-5 mini, which is a way better model, just costs two dollars. So the model's got around 15 to 30 times cheaper for more intelligence in just two years. That's what we call the intelligence per dollar curve and watching that curve as one of the most important indicators. If we take a look at also the how many tokens the companies are processing, Google reported that they're processing quadrillion tokens per month. That's a crazy number here. And so it's a five-axis increase year over year. So people are deploying this at scale. And so how is it possible the cost of falling so fast? It's coming from multiple directions. So first we have algorithmic improvements. There are new reinforcement learning and training techniques like GRTO by Deepseek or RLDR by OpenAI. You can get your better results from less compute. You have more different architectures like Mixer of Experts that can just turn on a portion of your models instead of paying for the whole model. And we have smarter thinking models that can sync actively based on the difficulty of the query. So the thing is we don't even know how to use the intelligence we already have right now. And most companies are really still experimenting to figure out what AI can do and how to deploy it. So I say that the botaniker now is not that AI is not smart enough, but really the cost is that what can transform our industry. And it's really easy for investors to forget about that cost curve, how quickly it can change. We saw it change over the past two years. You can think about how different it will be in 2028. And we can circle back and have this conversation about the cost of models then. And I think the fundamentals and the impact that it has on a lot of the companies that are, you know, say building data centers or using the compute will look fundamentally different. Do not know before and move on though. I do want to also ask how we can apply your technical expertise to an investing framework for our listeners. You've helped lead the change, charge with AI changes here at the full. If an investor is looking to evaluate the investments and performance of other companies, as it relates to their AI ambitions and capital expenditures, what do you think they should be looking for? Yeah, that's a great question. I think right now we're in a phase where companies are just throwing the AI at everything to see what sticks. And honestly, I think that's the pretty healthy because that's how you figure out what works right. You experiment. You don't have all the answers from the beginning. You have to just take risks, see how the products evolve, some fail and some succeed. But I think this phase won't last forever. So eventually the experimentation phase ends and you need three other results in the company. So when a company announces any initiative or a significant AI spending, I'd want to ask a few questions. So first, is this solving a real problem? It sounds obvious, but you'll be surprised about how often the answer is no. So is AI addressing an actual business problem? And I give you a simple test. Can this problem be solved without AI? And sometimes the answer is yes. So simple is better. And if a company is having an AI announcement, just to have an AI announcement, I'd be skeptical. So the second is is this action in production in front of users or is just a demo or pilot because right now companies can still get headlines for a demo. Startups can raise lots of money on a good prototype. But I'd argue that this window is slowly closing because everyone has a demo at this point. But the hardest part is to bring the demo to production in front of real users and scaling the app and making it secure. So that's what I want to see. I love the other one, which is the data advantage. So are they building on proprietary data or just plugging in generic tools? Because the models themselves are becoming more and more interchangeable. You can use GPT, Gemini, Cloud, GROC. They're all great and they all have different strengths. But we all have access to the same models. So what's not a commodity is your customer data, your years of refinement and arbitesting to figure out what your customer wants, its domain expertise, and so on. So I believe the companies that we get real value from AI are the ones using it on data that their competitors cannot access. Because if I can do the same thing, chat GPT or which I pay for their product. So the differentiation lies in the data and in the specific company context. So to recap, the first is AI solving real problem problems. Second, is it a prototype or is it in production? And so does it use proprietary data in assets or just generic tools that other schemes we reproduce? And I believe that the best AI investment sometimes just look boring is the company that may be quite using AI internally to make their people 20% more productive. Those companies compound on the long term and you should have a long term mindset. I think that's where the real values. And I hope everybody listening does have that long term mindset. I love what I love about your response to NATO is it's so incredibly measured. You know, you're the head of AI here at the full and it's easy for people to say, well, we're in a bubble. Anybody who operates in the space of AI is probably over-enthused, over-investing, over-indexing, over-hyping. But the reality is, is that the way that you speak about what you look for and an AI investment totally is actually the same thing that our listeners look for. It's something practical and purposeful. And to your last point there, maybe something that looks a little boring. So don't be afraid of adding boring to your portfolio. Osset not to put you on the spots, but I think you might have some maybe boring. We'll see stock ideas and proof points, I guess, ahead for what businesses may perform well in AI in the year ahead. So up next, we're going to be passing that mic to Osset who do evaluate these investment opportunities as well as some risk management strategies for portfolios. Stick with us. Welcome back to Mellie Full Money. As he wrap up today's show, centered around our AI investing outlook for 2026 reports, I want to pull Osset into the conversation to get a better sense of specific opportunities and some risk management strategies. Osset, you made a really specific point in the report that I want to mention because it highlights something unique, other than like the same big tech names that everybody already knows. You said, and I quote, "For the biggest opportunities, look to smaller semiconductor and data center ecosystem players, such as data and our connect specialists, high bandwidth memory providers, and cutting edge data storage designers." That is also a mouthful. But I think that's a really fairly unique perspective. And I kind of want you to translate that into something specific for me. Are these businesses or stocks that fit that description without those or are those just like AI vaporware? Yeah, such a great question, Emily. And I would argue that for all of these, they're really the concept is simple. In the first case, I'm describing companies that help sling data around faster within data centers when I talk about data center and connect specialists. I'm going to name some names here and these really aren't meant to be, "Hey, these are my high conviction buys. God and load up the truck." But more types of companies, you can start researching, understand they all come with risks. So the first example is Estara Lab Symbol ALAB. This is a company that simply helps different components within a server, talk to one another with lower latency much more quickly. And this is the type of boring thing that Donato talks about maybe on the inference side. So how AI is helping companies. Also for those that play in this ecosystem, they're doing really simple stuff at a high level, at a complex level. So the second thing that you mentioned, which you're referring to our survey, companies that are helping businesses like Nvidia, Symbol NVDA, manage memory within GPUs. That's a persistent bottleneck at that level of computation. And so we talked about high bandwidth memory or HBM providers. An example of this is micron technology, Symbol MU. This is a business that for a long time, played in a very boring space of the memory market, but low and behold, it has a very good technology to help sling data around a GPU faster than existing methods. So they're seeing some love in the marketplace. And thirdly, these cutting edge storage designers, these are businesses that are building specialized memory storage that are used within AI data centers. Your computer, my computer need memories to operate. Actually, my brain needs memory to operate. That's why I try to sleep at least seven hours a day. It's not so much different within a data center. So these information workloads that move around, you need storage drives for those. And that's a commodity business, but there are a handful of companies that are sort of at the bleeding edge. At the end of the day, what they're doing is making storage that's faster to access. It's very configurable. And it provides a lower total cost of ownership over the life of that component to the operator of the data center. So pure storage, Symbol PS, TG, and that's a company you and I have both studied. It's a great example of a business of this type. So there's a common thread running through all of these, essentially, if investors understand the inherent value of an Nvidia or advanced micro devices, chief competitor to Nvidia, to the AI story, I think in some ways, much of the value is sort of priced in. These businesses have had a great run. And investors are naturally looking now to suppliers within the spectrum of the value chain that exists between your keyboard where you input a query and your screen, where you get the response back from chat GPT. So what happens in between? It's not all about the GPU makers. Yes, like valuations are elevated. They feel sort of dangerous to me right now. So let's make sure we're clear about risk here. You know, since we commented on these high bandwidth memory providers in the survey, those have been under just like acute supply change shortage. So since we mentioned, they've really run up even more. So I hesitate even to talk about a micron, but there you have it. Be careful out there. Anyway, overall, I think there's going to be many investable opportunities outside of the GPU builders or the cloud hyperscalers that make up, you know, the rest of big tech, think Amazon, AMZN, Microsoft, MSFT, or alphabet, GOOG over the next few years. And that's why we want to think in holistic terms about a whole industry that's being built up. I think that's a wonderfully measured approach. And I love the fact that you mentioned the risk associated with a lot of these names, interesting companies across the different value proposition of AI. And the last question I want to pose to you, Asa, before we sign off here, is around that risk. Like when you're building your own AI investing framework, do you have any rules, things like position sizing, time horizon, milestones, anything like that? Sure. I have some rules. My first personal rule is to stay invested in the AI leaders. I own many of the companies I just mentioned, but especially those bigger names like Nvidia AMD, avoid concentrating in any single idea. Asa, you've done that before earlier in your investing career in tech companies. It didn't work out. Now, speaking of concentrations and other personal rule, when I'm assessing ecosystem players, I try not to shy away from customer concentrations for very specialized suppliers. It sounds counterintuitive. Like why would you buy a company that only has a few businesses, even though they're gigantic businesses as customers? Well, there's an example in a business like Arista Networks, Symbol, and ET, which for a long time was highly concentrated in those cloud hyperscalers, like Amazon and Microsoft, but it grew well over the years, and it's a little more diversified now. You're going to see this time, again, in this infrastructure because supply chains are limited and specialists abound. There are fewer players that have enough skill and technology to serve everyone, so their supply is getting snapped up by just a few players, but I position size accordingly, because there are so many concentrations. If I enter a new position of a company that I've been interested in, it comes in somewhere half percent or percent of my total portfolio, even sometimes a little bit less. And then finally, personal rule for this year, drill down into sectors and industries that are outside of my own core expertise. Look at last year, Emily, construction companies with expertise in building these complex mechanical electrical and plumbing systems for data centers, they just had a stellar year. It was outside of my wheelhouse. I really didn't pay attention until it was a bit too late, but I learned the lesson, the breadth of the AI trade and the opportunity, both are very wide, but you have to be willing to turn over some new stones to benefit, I think, in 2026 and beyond. I love that. It's a bit of curiosity, but also the all-important patience for investors. I know after our conversation today, it's clear to me that investors still have an appetite for AI. I hope that's cleared everybody, but they're also naming a lot of really key risks that are worth considering when managing investments and portfolios, both for 2026 as well as obviously the many years ahead of us of which I hope everybody is staying invested for. As a reminder, anybody who wants to read more can always access the Motley Fool's 2026 AI Investor Outlook reports at full.com/research/ai-investor-outlook. Again, don't have to memorize that, that link will be in the show notes. Donata and Asset, thank you both so much for joining today. As always, people in the program may have interest in the stocks they talk about, and the Motley Fool may have formal recommendations or/or against. So don't buy yourself stocks, be solely on what you hear. All personal finance content follows the Motley Fool editorial standards, and it's not approved by advertisers. Advertisements are sponsored content and provide for informational purposes only. To see our full advertising disclosure, please check out our show notes. For Asset Sharma, Donata Richio, and the entire Motley Fool's 20 team, I'm Emily Flippen. We'll see you tomorrow.