Thoughts on the Market

Our Head of Public Policy Research Ariana Salvatore goes through the main topics on the table during the meeting between Presidents Trump and Xi: Taiwan, tariffs and the Iran conflict.
Read more insights from Morgan Stanley.

----- Transcript -----

Ariana Salvatore: Welcome to Thoughts on the Market. I'm Ariana Salvatore, Head of Public Policy Research for Morgan Stanley.
Today, I'll be talking about expectations heading into the U.S.-China summit this week and what investors should be watching.
It's Wednesday, May 13h at 11am in Copenhagen.
Despite the importance of the upcoming summit, we think expectations for tangible progress should remain relatively modest. Reporting ahead of the meeting indicates that the discussions will focus on trade, Taiwan arms sales, and the U.S.-Iran conflict. Across the board, our base case remains an extension of the current truce with limited areas of relaxation. That's probably enough to support modest upside for risk assets in China, but likely short of the kind of breakthrough needed for a material re-rating in risk premia.
Let's start with trade. We think the discussion here is likely to skew toward phase one style commitments rather than structural policy shifts. That could include additional Chinese purchases in sectors like agriculture and aerospace, or things like high-level trade and investment pledges. Or even limited tariff relief in key areas designed to demonstrate cooperation but without fundamentally changing the competitive dynamic between the two countries.
What we don't expect is a meaningful unilateral tariff reduction from the U.S. side heading into the summit. Remember, China still faces an effective tariff rate of around 30 percent, and it benefited the most of all our trading partners when the Supreme Court struck down the IEEPA tariffs earlier this year. As we noted at the time, that lowered its effective rate by roughly 7 percentage points.
Secondly, we think the administration continues to view higher tariff levels on China versus other trading partners as a strategic imperative. Said differently, the administration appears committed to maintaining some degree of structural separation between China and other trading allies like Europe, Japan, and South Korea. We think that means a large-scale tariff reset is unlikely in the wake of the summit or in the lead up.
On Taiwan, we also see limited room for meaningful policy change. President Trump has publicly referenced Taiwan arms sales in recent comments, but we think a major concession from China would be needed for a meaningful departure from many years of U.S. policy precedent.
The third issue on the agenda is the Iran conflict and the Strait of Hormuz. Reopening the strait is likely the area of greatest uncertainty heading into the summit. The extent to which the U.S. will ask for China's help on this front and whether or not that request will be granted remains a key unknown.
But there's also a technology dimension here worth watching closely. While public reporting indicates that export controls are likely not formally part of the talks, we see a possibility that the discussion could occur, in particular in the context of rare earth relaxations from China's side.
Concessions on rare earth controls likely require some corresponding U.S. flexibility on advanced semiconductor exports, given the chips for rare earths equilibrium that we think underpins the strategic bilateral relationship. We think that's largely what's disincentivized both sides from escalating in recent months.
So, what should markets watch most closely? Aside from tangible trade arrangements or a formal extension of the truce, we think the tone will be crucial. Language around technology cooperation or an agreement to continue negotiating will be critical in assessing how both sides plan on managing the relationship moving forward.
Remember, this event is one of several potential meetings this year, so symbolic commitments toward broader structural concessions in the future could matter. For now, we think the most likely outcome is continued stabilization rather than a transformational reset. That's still constructive for markets at the margin, but probably not enough to eliminate the geopolitical overhang that continues to shape investor positioning globally.
Thanks for listening. As a reminder, if you enjoy Thoughts on the Market, please take a moment to rate and review us wherever you listen and share the podcast with a friend or colleague today.

Our U.S. Healthcare Analyst Erin Wright discusses how health tracking and preventive diagnostics could influence healthcare costs and different industries, from fitness to retail.
Read more insights from Morgan Stanley.

----- Transcript -----

Welcome to Thoughts on the Market. I’m Erin Wright, Morgan Stanley’s U.S. Healthcare Services Analyst.
Today – the emergence of the self-directed patient and its implications.
It’s Tuesday, May 12th at 10am in New York.
A blood test ordered from your phone. A wearable that tracks your sleep or nudges you to move, recover, hydrate, or rethink last night’s dinner. Preventive health is moving out of the clinic and into everyday life. And that shift is becoming an investable theme.
In essence, healthcare is moving from reactive to proactive. Instead of waiting for symptoms, more consumers are using lab tests, wearables, imaging, and digital tools to spot some these risks earlier. And this shift reaches well beyond healthcare.
On our estimates, the U.S. spends about [$]3.4 trillion annually on chronic diseases, including lost economic productivity. About [$]1.4 trillion of 2024 spend was tied to preventable disease. So the big investment question is: can earlier detection and behavior change bend the cost curve?
We think expanded preventive testing, screening, and monitoring can help avoid roughly [$]200 billion to [$]800 billion of U.S. healthcare spend by 2050. That assumes preventive testing reduces preventable disease costs by about 10% to 30% based on our analysis.
Direct-to-consumer lab testing lets people order lab tests directly, often online, without starting with a traditional doctor visit. We see this as a roughly $4 billion U.S. market, which has more than doubled since 2021. And it’s no longer niche. Our AlphaWise survey found that about 34% of respondents completed a voluntary wellness lab test in the past three years. Among users, the average was 3.2 tests, suggesting this is not just a one-time behavior. The most common test was a general health profile, used by about 45 percent of recent testers.
Wearables are the other part of the story. Our survey found that 41 percent of respondents currently use a wearable or fitness device, while another 22 percent are interested in getting one. More importantly, people are acting on the data. 34 percent of wearable users today regularly change behaviors or decisions based on their device, and 52 percent even sometimes do so, based on our survey.
That creates a feedback loop. A wearable might flag poor sleep. A lab test might show elevated glucose. A digital health tool might suggest changes to diet or exercise, or follow-up care. Over time, prevention starts to feel less like an annual event and more like a daily habit.
The sector implications are broad. In healthcare, more testing may initially actually increase utilization as people follow up on results. But over time, earlier detection could obviously support lower-cost of care and better chronic disease management. That also aligns with value-based care, where providers and payers are rewarded for better outcomes and lower total costs, not just simply more services.
In consumer sectors, better health tracking could shape food choices, reduce demand for some indulgent categories, and support products tied to hydration, lower sugar, protein, and functional benefits. Fitness may also benefit as gyms evolve from just workout destinations into broader wellness platforms, with recovery and coaching, and preventive health services layered in. Imaging is another emerging area, as screening shifts from reactive diagnostics toward earlier disease detection.
Of course, there is some risk that these health tracking and consumer-driven diagnostics trends could still prove to be a wellness craze rather than the new normal. Out-of-pocket costs, privacy concerns, inconsistent interpretations, and limited repeat testing are all real issues. But consumers are clearly taking more control of their health and increasingly asking, “What can I learn before I get sick?”
Thanks for listening. If you enjoy the show, please leave us a review wherever you listen and share Thoughts on the Market with a friend or colleague today.

Our Global Head of Fixed Income Research Andrew Sheets explains the economic theory behind the unwavering spending on AI infrastructure.
Read more insights from Morgan Stanley.

----- Transcript -----

Andrew Sheets: Welcome to Thoughts on the Market. I'm Andrew Sheets, Global Head of Fixed Income Research at Morgan Stanley.
Today, a uniquely price insensitive development.
It's Monday, May 11th at 2pm in London.
Elasticity is one of the first concepts that they teach in economics, and for good reason.
It's the idea that our sensitivity to the price of something differs from item to item. If the price of pizza goes up, for example, you may decide to go out for burgers. But if the price for something essential, like electricity, or deeply desired, like tickets to see your favorite artist perform; well, if those go up a lot, you're probably going to complain, but also end up paying anyway.
This latter category is what we would call inelastic. The demand for these items holds up even as the price increases, and maybe if the price increases quite a bit. And that is becoming very relevant as we all debate the AI build-out.
It's not an exaggeration that the investment in AI, chips, power, and datacenters is at the center of many market conversations. It's supporting U.S. growth despite a sharp slowdown in job creation. It's supporting stock market earnings, even as uncertainty over the Iran conflict continues to percolate.
Part of this importance is just the sheer size of this build-out. We estimate about $800 billion of investment by large U.S. technology companies this year, almost double their spending last year and triple their spending in 2024. But it's not just the size, it's the idea that this investment may happen almost whatever the cost.
Specifically, we're looking at a desire by multiple large companies to build out large AI infrastructure all at the same time, and that's increased the price of these components. The copper needed to wire together that data center? Well, it's up about 40 percent in the last year. A gas turbine to power it? Up 50 percent. The memory to run it? It's up 150 to 300 percent over the last year alone. And yet, despite these extremely large price increases, the demand to build in AI has been accelerating.
Our forecasts for 2026 spending have been consistently revised higher. And that $800 billion that we think is spent this year is set to be dwarfed by $1.1 trillion of estimated spending in 2027, based on the view of my Morgan Stanley colleagues.
This idea of inelasticity or price insensitivity extends even to the costs of financing the spending. Debt costs for these companies have increased this year, and yet they continue to issue at a record pace.
A quick aside as to why all this spending may be price insensitive or inelastic. AI is seen by these companies as, without exaggeration, maybe the most important technology in a decade. These companies have financial resources and the patience to wait it out, and they see gains to those who can figure out AI technology, even if the winner is not yet clear.
The inelastic nature of the AI theme is a classic good news, bad news story. To the positive, it suggests real commitment to this technology and that spending won't easily be shaken by outside events. That should help buttress overall growth and should also support earnings this year – a core view of Mike Wilson and our U.S. equity strategy team.
But there are also risks. It remains to be seen what returns can be generated from all of this historic investment. Robust demand for items, even as their price goes up, may cause those prices to increase even further. That's inflation happening at a time when core inflation measures are already well above the Federal Reserve's target. And if companies are less sensitive to the cost of their borrowing to fund AI, well, other companies could find their cost dragged wider in sympathy.
We continue to expect record supply and modest widening in the U.S. corporate bond market.
Thank you, as always, for your time. If you find Thoughts on the Market useful, let us know by leaving a review wherever you listen. And tell a friend or colleague about us today.

As real estate values reset and cap rates widen, net lease is back in focus—but the approach has changed. Ron Kamdem and Hank D’Alessandro explain.
Read more insights from Morgan Stanley.

----- Transcript -----

Ron Kamdem: Welcome to Thoughts on the Market. I'm Ron Kamdem, Head of U.S. REITs and Commercial Real Estate Research.
Hank D'Alessandro: And I'm Hank D’Alessandro, Managing Director on Morgan Stanley's Real Estate Investing Team and Vice Chairman of Private Credit.
Ron Kamdem: Today: a part of real estate that's changing fast and drawing fresh attention from investors. Net lease investing.
It's Friday, May 8th at 10am in New York.
You might not think you invest in net leases. But there's a good chance you do, especially if you have money in a pension fund or another income generating vehicle. Net leases are the kinds of long-term lease assets that can help generate steady, predictable income.
They are no longer a sleepy corner of the real estate market. In fact, they're changing in some really interesting ways.
Ron Kamdem: So, Hank, for listeners who know the term but may not know the structure, what exactly is net lease investing? And why does it tend to come up more often when markets get more uncertain?
Hank D'Alessandro: At a high level, net lease investing is typically associated with long-term leases that can offer durable income streams; typically growing streams, which is why it's often seen as a more defensive part of real estate investing. We see that when investors are thinking more carefully about geopolitical risks, market volatility or say portfolio resilience, this durable cash flow derived from mission critical assets and long lease durations with fixed annual rent bumps can become especially attractive to investors.
Also, with higher inflation likely, net leases are generally insulated from increases in expenses given these are the responsibility of tenants. But what's important today is the net lease is broader than many people realize, both in terms of the property types involved and the range of investors participating in the space.
Ron Kamdem: Let's stay on that idea of a broader market for a moment, because one of the biggest shifts has been the growing role of private capital in the space. What are you seeing there and why does it matter?
Hank D'Alessandro: Well, listen, Ron, there's no question. The role of private capital has grown substantially, including through joint ventures and public real estate vehicles. That matters because it tells you that the sector is attracting a wider range of investors than it has in the past, such as pension funds, insurance companies, sovereign wealth funds. And retail investors are increasingly investing either through traditional locked up funds or through semi-liquid funds. But it can also change the competitive landscape and can influence how capital gets allocated across the opportunity set.
Thus, one's approach going forward from an analysis perspective will need to evolve. More broadly, it's a sign that net lease is being viewed as highly relevant in today's market, not just as a legacy category within real estate.
Ron Kamdem: And that's an important distinction that you make right there, because not all investors are approaching these assets the same way. So, when private capital comes into the space, what separates their underwriting approach from another? And we hear all the time about private credit. How does that play into this?
Hank D'Alessandro: Well, Ron, you know, as we discussed previously, the competitive landscape is changing and therefore underwriting is absolutely critical in this part of the cycle. And so, we believe underwriting both tenant credit, of course, is very important. But we equally analyze the real estate underwriting because we believe that real estate can be a real differentiator over time – both in terms of returns and risk profile.
We think that strong real estate underwriting with strong tenant credit underwriting, both enhances returns over time and reduces risks. So, therefore, that matters a lot. We also believe that by focusing equally on the real estate underwriting, you get a fuller picture of the risk and value, especially as net lease expands into newer property types.
It is an easy nuance to miss, but we believe this distinction is becoming much more important differentiator in how investors assess opportunities in the sector today. And I believe that the most successful managers will do a good job underwriting both tenant credit and real estate.
So, Ron, for a long time, many investors thought of net lease primarily as a retail story. How much has that changed?
Ron Kamdem: Well, that's changed quite a bit. If I take you back 20 to 30 years ago when you thought of net lease, you thought of a convenience store that's, you know, 5,000 to 10,000 square feet. But today, that opportunity has expanded well beyond retail and there's much more attention now on industrial assets. And even increasing discussions around areas like data centers.
I'll give you an example. Realty income made its entry into the data center vertical in November 2023 with a $200 million build to suit JV. That shift matters because it shows net lease evolving alongside where demand and capital are moving.
It also means the sector is becoming more connected to larger structural trends in the economy, rather than being viewed through one traditional lens. At the same time as the mix broadened, investors have to be selective because not every new category will have the same long-term profile that we're used to.
So, as investors look at some of these newer areas, where do you see the best opportunities, Hank? And where would you be more cautious?
Hank D'Alessandro: So first, opportunities. The industrial segment has clearly become a major area of focus. This sector benefits from growing e-commerce penetration fueled by AI, reshoring of manufacturing, and increased defense spending. The ability to acquire mission critical distribution centers in top tier logistics markets or advanced manufacturing assets in innovation clusters is particularly appealing in today's macro backdrop.
Another area that we find very compelling is medical outpatient buildings where the aging demographics can support long-term demand. So, we have great conviction on both of those.
Now, turning to area where we're more cautious. There's been a lot of attention on data centers, you know, as you previously mentioned. But that's an area where investors really need to think carefully about long-term durability. Questions around obsolescence, technological change and whether certain assets fit a true buy and hold strategy are very relevant and need to be considered carefully by investors.
So, maybe to sum up, the opportunity set is definitely broadening, but selectivity in terms of location, asset type and asset specifications remain essential.
So, Ron, the idea of linking property types back to long-term trends feels especially important right now. How do you connect this conversation to the key secular themes Morgan Stanley research is tracking this year. AI and tech diffusion. The future of energy, the multipolar world, and societal impacts. And can you offer a few examples?
Ron Kamdem: There's a couple ways that net lease connects to these broader themes. The first, which is probably the most obvious, is technology diffusion and the future of energy comes through in areas such as datacenters, and that's been a key focus for public investors.
When you think about societal change – that's relevant for sectors tied to demographics like medical outpatient buildings, where you know people go get different services. And multipolar world theme matters because deglobalization and geopolitical fragmentation. Or influencing how investors think about resilience, location, and portfolio construction, which is driving incremental demand for industrial real estate linked to supply chain shifts and defense spending.
So, this is no longer just a sector evolving on its own, it's becoming more closely tied to these macro issues, shaping investment decisions more broadly. And once you widen the lens to that macro backdrop, the conversation naturally becomes more global.
In fact, we saw realty income now generates 19 percent of rents across nine European countries with more than $15 billion invested since 2019. Given this, Hank, how should investors think about net lease and adjacent opportunities outside of the U.S.?
Hank D'Alessandro: The global angle is clearly becoming more relevant. There's growing interest in Europe and the U.K. And one area that comes to mind in this context is retail parks, where rents have reset, yields are wider, and tenant resilience has improved.
Thinking more broadly, international markets can give investors a wider set of ways to think about real estate opportunities tied to the same themes that we've discussed. And add to diversification, as macro drivers continue to diverge and geopolitical risks remain elevated. Even when structures or sector exposures differ from the U.S., which undoubtedly they will, the bigger point is that investors are increasingly valuing opportunities through a global lens.
Ron Kamdem: So, if we pull all this together, what looks like a simple-income oriented category is actually becoming much more nuanced. As we wrap up, Hank, what's the main message you want investors to take away about net lease today?
Hank D'Alessandro: You know, I believe the main takeaway is that net lease remains relevant because of its defensive qualities, and predictable contractual cash flows derived from long-term leases. But the story is becoming more nuanced, requiring a granular focus on the credit, and importantly, the underlying real estate.
With real estate values down 20 to 25 percent from peak levels, replacement cost has elevated, which is keeping supply muted and net lease cap rates wide relative to the last 10 years. This is a very attractive entry point for investors.
Private capital is playing a bigger role, no question. The asset mix is shifting beyond retail, towards areas like industrial. Investors are actively debating the long-term role of newer categories such as advanced manufacturing and data centers. There are selective opportunities to think more globally, which is exciting.
Ron Kamdem: Great. That's very helpful. Hank, thanks for taking the time to talk.
Hank D'Alessandro: Great speaking with you, Ron.
Ron Kamdem: And thanks 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.

Original Release Date: April 28, 2026
Tom Wigg and Stephen Byrd discuss the accelerating pace of AI breakthroughs, the forces driving them and why the next phase of development may look very different from anything we’ve seen so far.
Read more insights from Morgan Stanley.

----- Transcript -----

Tom Wigg: Welcome to Thoughts on the Market. I’m Tom Wigg, Head of Specialty Sales in the Americas at Morgan Stanley, and a sector specialist in Technology, Media and Telecom.
We wake up every day to new AI product releases, so it’s easy to lose sight of the unprecedented non-linear improvement in AI capabilities. But things are about to get weird.
It’s Tuesday, April 28th at 8am in New York.
The market has been thinking about AI in linear terms. But we need to reframe that assumption of only incremental improvement and think about exponential improvement.
That was my takeaway from a conversation with Stephen Byrd, Global Head of Thematic and Sustainability Research at Morgan Stanley. In our conversation, we zeroed in on Stephen’s bull case for broader AI model improvements.
Tom Wigg: First, I want to talk about one obsession that you’ve been writing about for the last several months – is this idea that we’re going to see nonlinear improvements in the frontier models coming out this spring.
Stephen Byrd: Yes.
Tom Wigg: There’s been, you know, some big headlines around new models, benchmarks coming out publicly. Is this, you know, your bull case playing out on these models? And what are the implications?
Stephen Byrd: Yes! Absolutely, Tom. So we have, to your point, we are obsessed. And I know I’m not shy about that – with the nonlinear rate of AI improvement. It is the most important impact to so many stocks that I can think of in the sense that it can impact all industries, all business models. So, what we’ve been saying for some time is, if you look back over the last couple of years at the relationship between the amount of compute used to train these LLMs and the capabilities, we have a very clear scaling law.
And approximately the law is, if you increase the training compute by 10x, the capabilities of the models go up by 2x. Now, as you and I’ve talked about this a lot; just meditate on that for a moment. I think things are about to get weird in the sense that on the positive side, we’re going to see all kinds of underappreciated capabilities across many industries. So this disruption discussion, I think, is going to spread, but it’s also going to require investors to, kind of, be more thoughtful about what they do with that concept. Meaning you can’t sell everything. In the sense that AI will disrupt some businesses.
I actually think this is healthy in some ways because now it forces investors to really look at each business model and assess which is going to get disrupted, which can get supported and enabled by AI, which are immune. Because there are some business models that actually are immune.
But essentially from here, Tom, I’d say we are expecting through the spring and summer to see multiple models that are able to perform a much greater percentage of the economy at better levels of accuracy at incredibly low cost. Which I know you and I have talked a lot about the cost of actually doing this work from the LLMs.
This is massive. This is going to impact so many industries. I think this is all to the good for the AI infrastructure plays because it shows the importance of getting more intelligence out into the world.
Tom Wigg: So, you mentioned the constraints we’re seeing across compute, memory and power. It seems like most of the CEOs of the labs and hyperscalers are talking about this. Investors are bullish in terms of the ownership in, you know, memory, optical, semi-cap, et cetera. But the question I’m getting more recently is around what’s the ROI on all this spending. And does the market action in these hyperscalers, which have been pretty bearish year-to-date, force a cut on CapEx? So, maybe if you can marry that with what you’re picking up on the ground in terms of compute spend and whether the frenzy still continues, you know, versus the ROI? And, like, what could happen?
Stephen Byrd: Yeah. The short answer – I’m going to go through detail – is I think the bullishness is going to get more bullish over the coming months. And let me walk you through a couple of the mathematics and then just what I’m seeing on the ground to your point, Tom.
So the mathematics. We have a token economics model that looks from the perspective of a hyperscaler or an LLM developer in terms of – if they sell their token at a certain price and you fully load the cost of a data center and all associated costs, financing, you name it – in what are the returns? And the bottom line is the returns are excellent.
The other element we spend a lot of work on, and you and I talk a lot about, is the demand for compute. In this world where the LLMs are increasing in capability and the token usage goes way up with agentic AI, video world models, all that stuff, we think that there is a massive shortage of compute. So, if you’re lucky enough to be a hyperscaler with the compute, with the power, we think that they will have a lot of pricing power on the tokens.
Let me explain why we see price power on the tokens. Now I’m going to flip to the perspective of an adopter. Let me give you just rough mathematics. There was a study last year from one of the big labs showing that on average, an enterprise user using an LLM might be able to replace work that would take about one and a half hours from a human. That would save about $55 of cost. A million tokens, depends on whether you’re looking at input or output – but let’s just call it $5 for a million tokens.
The average usage case today for a fairly complex agentic task in an enterprise setting is in the tens of thousands of tokens. Okay? So let’s just do that math again. $55 of savings. A million tokens cost $5, and a typical agentic usage is far less than the million tokens today, though that will accelerate. The economics are a home run for adopters.
So, we’re in a situation where compute is very scarce. I see pricing power all over the place for those who have the compute and have the power.
Tom Wigg: So, when you put it like that, Stephen, it seems so inevitable and obvious. But I wonder why the hyperscalers are trading the way they are? And when do they see the revenue inflection you’re talking about? Is this like a stay tuned kinda 2026 event? Is this something we have to wait for for 2027-2028?
Like, how do you think this flows through to the extent that the market will get more comfortable that all this free cash flow pressure is worth it on the other side?
Stephen Byrd: Yeah. This is, in short, I think this is a 2026 event. But let me dive into that because what you just asked is so important for so many stocks.
So, let’s talk through this. The capabilities of the models are advancing so fast that the average corporate user is not yet keeping up. There is this gap. But that will happen quickly, and we’re seeing signs from these labs of revenue at the lab level that is accelerating. So that’s a good sign.
What we’re seeing, though, among fast adopters is those adopters who really understand the capabilities are quickly realizing just how economically beneficial there is. An example, one of my best friends founded a software company many years ago. Last month was – that was the last month in which his programmers wrote code. They’re done with writing code.
The efficiency benefits for his business are absolutely massive. But he feels like he’s just scratching the surface, and he’s about as technically capable as anyone I know. He has two PhDs in the subject matter. He’s very, very good.
So long way to say that we’re living in almost two worlds where the fast adopters will show what’s possible. The average utilization for enterprises will still take some time. But I do think that the market will react to what they see from the fast adopters in the sense of – the tangible economic benefits are so big.
Now, on the ground, what I’m seeing on the infrastructure side, my friends in power tell me that a couple months ago is when they saw the sense of urgency from the AI community go up a couple of notches for them to get the infrastructure they need. So they saw this explosion in compute coming. In the last two months, the weekly usage of tokens according to OpenRadar is up a couple hundred percent in a couple months.
So, I do think we’re seeing this. So, this is; it’s happening quickly. What I would say is the market will have these signposts in every industry of early adopters showing this benefit. I think that’s enough for us to start to get bullish. We also… I just think when you look at the demand for compute, the compute numbers need to go up. And with that, you know, everything in the AI value chain, infrastructure value chain, the volumes need to go up.
Tom Wigg: One bear case that I wanted to interrogate was – there’s one view that, yes, there’s a token explosion right now. But it’s because the first use case is coding. Which is inherently, you know, very developer-friendly and token-intensive relative to other knowledge work.
Can you talk about, you know, whether you subscribe to that? Or whether the token intensity will be as high or lower as this expands to other areas of knowledge work in the next several years?
Stephen Byrd: Yeah, it’s a great question. The short version is that, yes, it’s true that software usage is more token intensive. However, what we’re going to be seeing – we’re starting to see it – is in almost every knowledge-based job, we’re going to move to agentic AI. And when we do that, you tend to see an explosion in compute.
Let me walk you through the numbers. There are a couple studies that show essentially when you go from a query-based usage of LLMs to an agentic use for any occupation, you see about a 10x increase in token usage per use of those models. And you can see why.
I’ve anecdotes of some of my friends who are newer to this – who set their agents loose overnight to do non-coding work. And in the morning they get some pretty amazing results. But they also used a lot more tokens than they’d expected … (laughs)
Tom Wigg: And a five grand credit card bill?
Stephen Byrd: Exactly. It’s like maybe next time you put a few parameters around that. But long way to say, it’s agentic across every workflow that I can think of that will still result in an explosion in token demand.
Tom Wigg: It’s definitely a good idea to put some parameters around your agentic workflow.
My thanks to Stephen for that conversation. And thank you for listening. Let us know what you think of the show by leaving us a review where you listen. And if you find Thoughts on the Market worthwhile, tell a friend or a colleague about us today.