AI Euphoria Or Early Peak

We question whether AI hype has outrun fundamentals while mapping a practical plan to stay invested. We dig into energy limits, data center pushback, circular demand, debt signals, selective stock picking, and tax-smart ways to lock in gains.

• froth signals across AI leaders and suppliers
• circular demand and debt-funded capex risks
• grid and energy constraints shaping data center buildouts
• local approvals, community impact, and siting choices
• selective exposure over single-name concentration
• competition across chips, code, and applications
• practical profit-taking and rebalancing tactics
• donating appreciated shares for tax efficiency
• preparing allocations for the 2025 to 2026 shift

Think about it

Straight Talk for All - Nonsense for None

Please check out our other podcasts:

https://skepticsguidetoinvesting.buzzsprout.com

Disclaimer - These podcasts are not intended as investment advice. Individuals please consult your own investment, tax and legal advisors. They provide these insights for educational purposes only.

Chapters

• 1:14 Framing The AI Investing Debate
• 1:47 Is AI Hype Near A Peak
• 3:38 Locking In Gains After The Run-Up
• 4:34 Circular Demand And Debt Signals
• 6:24 Data Centers, Energy, And Local Pushback
• 9:04 Grid Limits And Community Tradeoffs
• 11:19 Jobs, Productivity, And Adaptation
• 12:44 Who Leads: Chips, Code, Or New Entrants

Transcript

Clem Miller: 0:50

Hello, everybody, and welcome to Skeptic's Guide to Investing. I'm Clem Miller. I'm here with Steve Davenport, and we're going to talk briefly today about AI and where we're going as we end 2025 and go into 2026. Steve, do you think we're peaking on AI, on AI stocks, AI as a technology, or do you think we still have a ways to go on all that?

Steve Davenport: 1:23

No, I feel like we've kind of reached a point where I think we might have already seen the peak. I believe that there's a high likelihood that we're gonna see something pull this AI idea down back to earth. I think it has a great foundation. I think there's a lot of reasons to think it will make a difference. I just don't think it's based on the, you know, the magnitude of that difference is being, in my mind, exaggerated or overestimated by a lot of the market. We are in a place where I honestly believe that it is getting harder to to really see the impact and measure it for a lot of companies that use AI. They say, I know it's impacting things, but is it AI? Is the reorganization around AI, is it businesses that are starting to focus on the right things because AI exists? Those aren't necessarily AI-related improvements. I agree that just as using the personal computer, just as using the internet, all of these things eventually led to businesses becoming more effective. But the impact and the estimates for these companies over the next two, three years ass e perfection. And as we know, Clem, there is no perfection. I just I just want to say I'm thinking of this in relationship to recap of 2025. And what we're gonna do today is record five shorter clips in the 15 to 20 minute range versus one long recap that's an hour and a half, and then people fall asleep in. So this is meant to be focused and staying with how do investors look at AI through 25. And my opinion is we've seen an unbelievable run-up, and we've seen a lot of great, you know, appreciation occur. And as an investor, my first thing is well, let's lock some of it in. Is it too too soon to do that? I don't think so, given the magnitude of what we've seen from NVIDIA and Broadcom and others. And , I look at more and more names who are telling their story better or getting their AI buttoned up. And I think that when we started this year, it was just a complete feverish move towards AI. And we didn't know why, we didn't know what. And then we saw some things during the year with Tr p and tariffs and with other things regarding the movement of AI, where individual companies started the circular rotation of I'm gonna buy more of your chips if you promise to use more of my, you know. And I I think that once you saw that, that's one signal that you're near top. When companies create this movement on the balance sheets that really doesn't, it's like buying your own demand and creating this demand kind of with the resources you have. We also saw in this year, you know, Oracle taking on debt. They didn't have enough cash flow, so they needed to add debt. And, you know, I think that people starting to look at the credit default swaps on the hundred billion of Oracle debt. Is the debt too much? No, absolutely. They could handle it as long as things continue as they are and as they are going to be maintained through a normal growth cycle. I don't think it's it's really that unclear. There's some signs of froth, and that froth is usually an indicator. And do we need all four signs to be lined up in order to have a peak? No, I don't think that's the way froth works. It's an emotion, it's not a you know quantifiable variable. What do you think, Clone?

Clem Miller: 5:41

Yeah, so I agree with you that the market is quite frothy for for some of the of these AI-related companies. Uh and I think that there are some real issues in this whole space. Uh for example, one of them is the this whole issue of energy for server farms. Uh is there enough energy? Um you know, is this has this gone kind of crazy? You got all these companies out there building data centers for AI. Uh and you know, there's a projection of need, but you know, is this you know like a real is this a real estate bubble that's going on with regard to data centers?

Steve Davenport: 6:29

Is I think I don't know, like, does all the do all the data centers necessarily have to have all approvals before they'll start the process? Because I've had a reaction here in my small town in Matthews, North Carolina, where they proposed a data storage for you know, a data center for a small company in Charlotte, and the town rejected it.

unknown: 6:55

Yeah.

Steve Davenport: 6:58

And I was like, should should towns be rejecting an opportunity to have a pretty, you know, what I would call minor inconvenience versus a retail organization or somebody with actual, you know, more negative environmental impact. I I'm not sure we should be rejecting these at the local level. I'm not sure we shouldn't all be saying, hey, build it, let them come. It's it reminds me of you know what the senator there from Arizona did when he he said, let's build the you know nuclear power storage facility in my state. And everybody was like, really? And he's like, Yeah, and let's spend six, ten, twelve billion dollars building it. And then when it gets built, you say, we don't want that here. You know, it it creates these these things where we're all pro-growth until it happens in my backyard, right?

Clem Miller: 7:56

It's all about it's all about employment, right? Building things in your backyard, and then when it's finally ready to go, you you have second doubts about it, right? So right.

Steve Davenport: 8:08

And I just think the so my my question isn't just energy and the sources, and can these grids keep up with the needs? Are they gonna hurt the local people if they're drawing energy and and away from actual usage of of individuals?

Clem Miller: 8:27

Yeah, I mean, can you imagine they're building data centers in Texas and the Texas grid can't even handle a winter out there?

Steve Davenport: 8:34

Correct. There's a lot of there's a lot of reasons to believe that if every component of this comes true, then yes, we could have a very big moment. I I was at a group last night discussing this whole issue of, you know, will AI just pulverize the job market, and all of us will be sitting here looking at each other, going, well, I've been replaced. I don't know what to do. I mean, there is this aspect of AI that nobody really talks about, which is how do we figure out the impact and how does the individual adjust? We can't always say, just get smarter. Because guess what? The rate at which we get smarter and the rate at which AI gets smarter, they might not be the same rate. Um, I know my you know, my brain operates a little bit like a cutting-edge chip, but you know. Um, so I I think that you bring up a good point, Clem, because it it comes down to electrons delivered in an environment. And how do you get the energy for those electrons? How do you get the, you know, is it in the code? We've talked about, you know, there's a deep sea revision coming out this month. And when it comes out, you know, is their code going to be so much better as it was the first time that using chips of lower technology that they still are able to get a better chatbot and to build a better mousetrap. I think that we're we're missing some of the creativity and some of the ability of other countries to come up with a better solution. I'm not saying they're going to or American companies aren't going to succeed. I'm just saying that usually people have an affinity for certain models and certain ways to do things, and they become the leaders. And I'm just not sure we know who the leaders are gonna be. Google this week, you know, released some great data on their Gemini. I I think Google has been underestimated.

Clem Miller: 10:41

Yeah, well, you I I agree, and I'm invested in Google, but have you actually tried to use Gemini?

Steve Davenport: 10:49

No, I don't know. Is it good or bad?

Clem Miller: 10:54

Uh it has a ways to go. All right. Yeah. Uh it has a way to go. I think those of us who have used it will know what I'm talking about. So okay. Uh it's not the greatest. So so yeah, so I think I think it's fair to say, and , you know, if I can conclude this, I think it's fair to say that AI is a still a good investment. AI stocks are still a good investment. I think we need to be selective about which AI stocks to invest in. Everybody thinks NVIDIA, , that's not necessarily going to be the best stock going forward. Uh we just don't know which stocks are going to be the best going forward. So it's it's better to have a group of stocks that represent the the AI space.

Steve Davenport: 11:51

Yeah, I mean one thing I'd say for people is I'm looking like most people at the end of the year as to how do I manage my taxes. And one of the things that there is a small advantage to giving in 2025 versus giving in 2026, depending on your tax and other circ stance. And so all I'd say is AI stocks represent an opportunity for you, but it doesn't always have to be an opportunity where you sell, it could be you donate some of these stocks at the end of the year, and when you do, that that deduction of against your income or against future income could be meaningful to you. So think about it.

Clem Miller: 12:33

Yeah, I mean yeah, to that point, Steve, to that point, and well, a little bit related to that point is you know, I do think, and coming back to what you said earlier, I think now is a good time to take some profits. I think that's been the case over the last few months, and indeed I've been doing that.

Steve Davenport: 12:54

Um, I think we talked about that this year. Yeah, yeah. So my only thing now is try to do some planning, whereas if you have a donor advice fund or you have some places that you want to give. This, I think , the taxes in 25 will be better for you than 26. And I think that in all things, we're trying to help educate. And I think overall, , this is one area that you should look into and understand so that when names go up, you realize there's another choice out there. Yeah.

Clem Miller: 13:29

So thanks everybody for listening. Let's let's conclude this one. Uh and and we're going to be doing a few more of these mini episodes to talk about sort of the turning point between 2025 and 2026. Uh thanks, Steve, and thanks everybody, and stay tuned for our next one.

Steve Davenport: 13:50

All right, thanks.