“The Shakeout”: Markets, the Midterms, and Who Controls the Stack | The Open Current Ep 6
Wall Street, Main Street and AI
Is the AI market correction a healthy reset or the start of a bigger shakeout? That question drives this episode of The Open Current. Co-hosts Alan Shimel and Margaret Dawson take stock of a rough stretch for AI and chip stocks.
Shimel notes that the slide began soon after markets hit record highs in early October. Infrastructure players, application vendors and even security companies have felt the pressure. Meanwhile, the largest cloud providers keep raising capital spending.
Dawson questions whether that spending math adds up. She also points to the heavy concentration of index gains in a handful of AI-linked companies.
Lessons From Earlier Booms
Shimel draws on personal history from the dot-com era and the 2008 housing crash. Both times, he saw warning signs when everyone seemed to be chasing the same trend. He wonders whether the constant AI chatter he heard across Europe this summer is a similar signal.
Dawson pushes back on part of the comparison. AI has already become part of everyday workflows, not just a speculative bet. Still, both agree that circular deals between a few companies deserve scrutiny.
Communities Want a Say
The conversation turns to recent elections, where data center measures appeared on local ballots. Many proposals aimed to cap facility size rather than ban data centers outright.
The co-hosts see this as a bipartisan push for governance and local sovereignty. Residents want clarity on power, water and costs. They also want real jobs and economic benefits in return.
Customer Focus and Open Source
Dawson argues that vendors should focus on what customers need instead of chasing the boom. Shimel adds that execution and focus matter more than being first. Companies that lose sight of customer needs tend to struggle when the market turns.
Both hosts see open source gaining from any AI market correction. Smaller models, edge inference and local hosting all benefit from open innovation. Dawson sums up what AI needs most right now: transparency and trust.
Transcript
" And I'm really happy to have my co-host here, Margaret Dawson. Hi Margaret, how are you? I'm good.
A little tired, just got back from Vegas. I know you. Which always kind of saps you a little bit.
There's something about that place that even when you're well-behaved, you just feel like you've been through the ringer a little bit. No, it takes the energy out of you. Exactly.
But other than that, yeah, I'm good. How about you? You've had a rough couple weeks.
I'm good. I hurt my back, but like a trooper, I sat on a plane for six hours to be in San Jose and six hours home. And I've been to Austin twice.
I've been all over the place and I have to go to Vegas. Yeah, but pros play with pain. What can I tell you?
Margaret, this is, I think, our first episode where you're not in green. I know. I'm looking very Vegas today.
I'm just wearing jeans and a T-shirt. You're looking very Vegas. I'm branded.
Like I'm just so- All right. It's casual Open Current day. And you're not wearing a sports coat.
No. Yes. This is your version of a T-shirt.
Just a button-down shirt. Oh my God. This is a regular sporty shirt.
Well, I've got a confession to make. As soon as we're done with this podcast- Yeah ... I'm running home, getting Bonnie, and we're going to see Rush.
Oh, fun. Yes. So I'm going to see Rush.
How old is our listening demographic? I don't know. But if they don't know who Rush is, I'm sorry, but I'm going to see Rush.
How old are the singers of Rush? They're old because this is their 50th anniversary. Is it animatronic or are they real?
No. The bass player, singer, and guitar player are the original, Geddy Lee and Alex Lifeson. The drummer had passed away, and they've got this woman drummer from Germany, who I understand is amazing.
Neil Peart, the drummer for Rush, was amazing, but she's-- How did we get... Margaret, this is why I love doing podcasts with you. You never know what we're going to talk about.
Because usually there's more fun to talk about than the demise of society, and the stock market, and AI build-out. Sometimes we've just got to let it go. You said it.
You said it. But unfortunately, we are going to talk a little bit about it. Do it.
It was a big week with the election and everything. But you know what? You and I haven't had a chance to catch up and do our show in a couple weeks with all the traveling.
My book came out on October 6th, and it went bonkers. It had number one bestseller, all that stuff. On that day, Margaret, not because my book came out, but on that day, Nasdaq hit a record close.
Right? It was flying high. " And then everything went to s**t.
Yeah. Pretty much. Since then, especially the AI and chip stocks, have lost about five, write this down, $5 trillion.
Yeah. Five trillion. My brain doesn't even wrap around that, honestly.
Like what is that total, exactly? Let me just put it in perspective. But you look at some of those numbers, I don't know how much the Nasdaq or the S&P 500 are down overall in October.
I don't know if you have those numbers, but it's like, yeah, stuff right after that, like everything just went down. They're in correction territory, which is more than 10%. Yeah.
To put it in perspective, in the dot-com meltdown- Yep ... about $2 trillion was lost, and it took, I think, five to seven years to get back. What was the relative percentage of the total value of the market at that time?
I wonder. You might have to ask my AI. It would be interesting to look at that, right?
Percentage to percentage. No, so $2 trillion in their dollars to today's dollars, it's probably about equal, to tell you the truth. Yeah.
It's about two and a half times. The important thing, though, is it took five to seven years for the market to get back- Correct ... to that level.
And I hope it doesn't. Yeah. Do you remember what happened?
What was weird about that, though, and I don't know if this will happen again, if you remember, it took five to seven years, and then in 2008, we had that huge financial crisis again. So it was like- Yes ... 2001 and then 2008.
2008. Yeah. Boom.
And it took us-- And now we've come a long way since then, certainly. But I think, you know what they always say? " Right.
In this case, I think they're more correlated. I think in Main Street, the whole AI thing is making people very jittery as well. I think the American, at least here in the US, the people have spoken, right, in their election.
Well, of course, interesting to your point, I love that you just said that, because there's Wall Street, there's Main Street, and then I would say there's Silicon Valley. I don't know what street that is. Maybe it's in Palo Alto.
Fantasy. What was the Grateful Dead song? " I always think of Sand Hill Road, which is where all the VCs are in- Yep ...
Silicon Valley. But anyway, my point that I was making is that AI is the common thread now, which we rarely see that. Usually, there's differing things.
But you see the Wall Street jitters around AI. You saw OpenAI reduce its raise level. We've got Anthropic getting ready for an IPO.
People are getting jittery over that. You saw Intel have its worst day since 1974, which is just hard to wrap your head around, right? They lost something like almost 30% in value.
You saw Amazon, AWS, having a great quarter, and yet Amazon having negative free cash flow- Cash flow ... which freaks me out. But the whole thing goes back to this huge data center, AI-driven build-out, which as you said, the main streets, and this was very Middle America, which is what I find very interesting.
" And that's just freaking everybody out. So there was this kind of calm from the general election, but then I think the reality of all-- You had it somewhere in your notes. How many referendums were there just about data centers?
There were 41 on the ballot. That's wild. Different bills around data centers.
" So these gigantic gigabyte AI factories, that's what really was freaking people out. And 40 of them were against data centers, and a good chunk of-- I don't have the exact number on all of them yet because they're still tabulating, but a good chunk of them were in fact passed. Look, a lot of the winning, not just Democrats, but mostly Democrats, a lot of the platform was, "We're going to rein in the tech bro billionaires.
We know these people have undue influence," right? " And all of these things. And you're right, a company like Amazon, that's basically a printing press for money, negative cash flow.
Google, too. I don't want to just pick on Amazon. That's what I'm struggling with.
It was one thing, Amazon, in the early days when it was still VC funded- They used to do that all the time ... even though you see that. But this is a mature, highly profitable-- Their margins increased.
So my brain is struggling to keep up with this. " But the level of capital expenditure is just off the charts, right? Yeah.
But you look at some of this, it's not just one sector. The semiconductor industry or semiconductor sector, I guess, fall what, over 10%, maybe it was even 11%, right? So huge hit.
Alphabet, 13%, right? We talked about Intel dropping 30%. But software, and we've talked about this before, that the infrastructure layer is not what's getting hit, but the application layer, ServiceNow, I don't remember the number, it was like 30%.
31%. Gee. And on a weak forecast.
They didn't do terrible, though. Right. Couch here, CrowdStrike, right?
Which CrowdStrike, you think, wait a minute, now we're talking 60. We still need security. It should increase.
I would think security would, right? So it comes down to this. A lot of the AI stocks, there's about nine or 10 companies.
Again, I covered it in my book. There's nine or 10 companies in there that consider it. But then there's the peripheral ones that build stuff for the data centers, the Caterpillars, and these kind of people build generators, power plants.
They're all hearing it. But Margaret, because it was all based on scarcity, right? Right.
Andy Jassy announced that basically their big build-outs were sold out through 2027. Right. They were already taking orders for 2028.
Yep. I don't know what the definition of insanity is, but here's a little bit of what I think is insanity. At the same time, these stocks are getting hit, the most, not profitable, but the companies generating the most revenue in the history of the world are going negative in cash flow.
What's their response? We're doubling down next year, right? Is that insanity, or are they playing 3D chess while we're playing checkers?
Well, it's a good question. I guess my question to you, because you said we've seen this before, do you think this is what we're seeing in October and even into early November is just a correction and it's part of that normal cycle when you've got a huge boom? Or do you think this is going to be a major shakeout in this cash flow situation, all this is a lot more of a demise than we want to believe?
I'm glad you asked. So let me give you a little Alan Shimel personal history. Oh, God.
Okay, Shimmy- Shimmy history ... Shimmy graphics 101 right here. The year is 2000.
Ugh. We just had our first child, and I had bought what was my dream house. Okay.
Right? And we were under contract, and I told my wife, "Don't take any of the old furniture with us. We're going to hire a designer- Mm ...
" Wow, you were going bougie. Okay. Well, I was sitting on a lot of the stock of my company that had just gone public, and things were looking good.
And I was flying home, as like then, like now, I flew everywhere all the time. I was flying home, and I'm sitting next to two-- And I'm not kidding you. Don't be mad at me.
I'm not stereotyping. But two big, heavy guys in those denim overalls with cowboy hats, like farmers. And I hear one talking to the other.
"I don't know. " And the other one starts talking about what stock he's in. " They don't.
And again, it's in my book. It opens up with me sitting at the top of a skyscraper in Houston, right? And it was Enron.
" It's a momentary blip. We're in a new paradigm. Momentary blip.
We're in a new paradigm. Let me fast-forward to 2007, or maybe it's 2008. I'm in New Jersey.
We were living in Florida already, but I landed in New Jersey. I rented a car, and I was on it back to the airport, and I had to get gas because I had to return the rent-a-car with gas. And if you've ever gotten gas in New Jersey, you're not allowed to pump your own gas.
Yeah, Oregon's the same. Yeah. Yeah.
So I pull in, and there's a kid there who's going to pump my gas, and I said, "Fill it up. " And while the gas is going in, he's telling me he's buying a three-family building in the town nearby there because he's going to flip it. " I had a flashback to the two guys in the coveralls.
Right. If the kid pumping gas in the gas station is flipping three-family houses, this thing's about to go. I'll fast-forward now to this summer when I was in Europe.
Because you could say that isn't it great that everyone gets to participate in economic success? You could spin this a different way, right? I could say that- Okay.
But you're not going there ... but I'm Shimmy, and I got the scars to prove it. Okay.
This summer, I was in Europe. Actually, we were in Prague, right, for SUSECON, and then I continued through Europe. People were talking foreign languages.
" See, but that's hard to say because I think there's one difference is that it's become part of everyone's personal workflow. So it doesn't mean they're investing. It doesn't mean there's the financial dependency, but maybe.
You're right. But, and I'm talking now to my American friends, how much of your 401 is based upon you have S&P index funds and 80% of the gains in the S&P index have been directly related to Mag 7? Well, I was going to talk about that.
So you're bringing up concentration, right? And what we're looking at. And so while I'm not an economist, I think that's always something you want to look at.
Even in private enterprise, you always look at the concentration of your customer base, right? It's that 80/20 rule. And if we start seeing that where these huge indexes are so heavily concentrated on these very volatile stocks based on what we're seeing, then yes, then it's a house of cards.
And- This is why I love always having a lot of international shares in my portfolio. Well, I do too. Hello, I do too.
But- I am small cap, international. Me too. I play because I may not win them all, but I'm not going to lose them all is kind of my thinking, especially as I've gotten older.
Yeah, but here's the thing then to think about is we've known about this concentration, but it's worse. So much of the money attributed to this AI boom is money that goes from one spreadsheet to another spreadsheet but never really changes hands. I scratch your back, you scratch my back, we both put it on our books.
Well, and it's a very small number of companies. I think it's slightly different- Again, it's very concentrated ... Yeah.
In the dot-com, it is almost similar to the bank collapse in that way. But in the dot-com era, it was, ah, I feel like hundreds of companies. There were so many companies that made a quick fortune and didn't have a floor to stand on because it wasn't based on anything.
I think of all those- Yeah ... I always think of the Super Bowl commercials in 2000, right? You look at, what was the joke back then?
You look at the Super Bowls in '99 and 2000 and then who's still in business. None of them. com.
Yeah. But we sort of, I don't know if you watched-- Well, you probably did because Seattle was in the Super Bowl this past year. Yeah, they were.
But there was plenty of AI commercials- There were ... in the Super Bowl. There were plenty of tech commercials.
But let me look at it half the glass half full, Margaret. Okay. I think- I love your pause there.
It's like, "Wait, wait. " Why? Is the glass half full?
Let me remember. But no. The fact of the matter is, over the last two or three years, the AI companies have basically written checks to get what they want and their will as part of national policy.
So we have gone forward with these mega factories. We have gone forward with $8 trillion of investments in all of these things. I do think the election changes this.
I think it's primarily Democrat, but all of the people coming into office are being sent with a mandate of put a brake on, run away. It's about governance. Yes.
I don't know if it's put a brake on it, but it's like, can we put a little boundaries over this? Can we add some governance over this? We're going to need- Can we add some oversight- Yep ...
to this before-- Because the problem is, and if we look at financial crises or we look at other things that happened, we always did the governance after the fact, right? Something bad happened- Right. You had to get whacked in the head before someone said- Right ...
" Right. Exactly. And so there's been a lot of talk about can we for the first time, and maybe it's already happening, maybe it's too late, but can we at least start to put some guardrails on it or have some oversight or build those boundaries before we have something worse than Hugging Face, right?
Before we have the nuclear holocaust of AI. But here's the thing. Necessity is the mother of invention.
Right? And I think by having some governance, by saying, "Hey, maybe we're not going to build these huge AI factories. We're going to build smaller, but maybe we're going to use AI, be smarter about how we use AI.
Maybe we're going to run AI at the edge. Maybe we're going to run AI on our personal devices. " We are going to see- More diversity.
Right? And by the way, this plays into the whole sovereignty thing, Margaret. Right?
Because- Oh, it plays into two things. It plays into sovereignty. It plays into sustainability.
I'm sure we could come up with another S, but which Europe brings a lot of this stuff together, like ESG and sovereignty and all these things kind of play on the same field when you're in Europe. But yeah, I think it's huge. How do we think about sovereignty in terms of us controlling what happens to our land, us controlling what happens to our data, us understanding what these models are doing, and just making sure that we all have clarity on the value that is coming?
So it's weird that, I mean, that's not exactly sovereignty, but if these states were- No, but it is. It's freedom over what's yours. It is sovereignty, I believe.
Yes, it's sovereign AI. I think this is where it's funny because we use this all the time at SUSE. We talk about sovereign AI, and this is kind of a different spin on sovereign AI, where it's like I want to have a hand in what is happening around AI in my sovereign location, right?
I get to have a voice on that. Even if it is like you're saying, it's a lower megawatt data center. Whatever it is, it's a local decision.
Because here's the thing, I don't care what side of the political spectrum you're on. If you're going to stand up here and say, "AI is bad, data centers are bad, let's go back to just chiseling letters on rocks," that doesn't work. No one-- Let's not be silly.
AI is an amazing technology. It has all kinds of benefits. We're going to go forward.
What I think people are saying, and I hope this is what they're saying, is that we just want a say in what goes on in our neighborhood. Well, I don't think it's even just that. Yes, and one, this is bipartisan.
You can see it everywhere. It's not just, this is not a political- Yeah ... party, maybe at the party level, but not from the human, the voter level.
It's bipartisan. Two is they want the economic return. I think it's very fair.
They're saying, not only is this hurting our environment, not only is this, we think, sending up our cost of electricity and water and all these things that are scarce already in the world. But is it bringing jobs? Is there an economic benefit?
Well, bill it. And so I think those things together, those are fair questions. If someone had built a huge Walmart in the middle of North Dakota and had zero jobs, the same argument would be happening.
It's not just because it's a data center. I think people are tired of these massive investments where the return is not going to the people who live there, right? No.
There should be some economic- It goes to the billionaires. Right. There should be some economic value for the people, either better schools or better streets.
" It's like, well, how would we know that? You're not sharing the contract with us, right? That's the contract, right?
Well, many times the politicians had to sign an NDA that they can't disclose what was in the deal. See, and that just goes against, that just feels yucky to me. That feels really- It does ...
yucky. Yeah. Let me talk a little more about yucky.
Okay. There's only one of us here who work for a vendor. That would be you.
Wah, wah, wah. Wah, wah, wah. What does this- Am I recusing myself now?
What's happening? No, no, don't recuse yourself because I want your honest opinion. Okay.
What does this mean as a vendor? What does it mean for you? I think about this very simply, and I've been in tech for a long time, as you know.
I always want to understand what this means for my customers. I take that responsibility very literally. And I am always trying to think, like, do we have the technology that is still useful, that is still helpful, that people still need to keep their businesses running, AI or no AI?
And now with AI, are there gaps? Are there new things? Is security a bigger concern?
So I think if I look around the tech industry holistically, what is scary is there is a lot of companies that have been at this a long time, and I feel like they're losing their way. And what they lost their way is what I just said. They're not thinking about, what do my customers need me to do to help them navigate this?
They are focused on, how do I save my company? How do I make more money in this AI boom? They're asking the wrong question because if you help your customers solve the challenges they're having in this era of AI, those other problems will be resolved because you're- Yes ...
helping the market. Because there's so much easy money right now, and the same thing, this is where the parallel to dot-com is kind of freaky because you saw these companies going to market, and no one had that pain. They weren't really doing the foundational product market fit.
They weren't going to customers- No ... " Well, at the end of it, nobody wants to do that. And we know that.
Yeah. Right. So I know this sounds overly simplistic, and it doesn't mean that there aren't going to be companies that end up going belly up.
There always are, right? But why do those companies go belly up? Whether we look at the last digital transformation, we look at the dot-com era, you can look at any industrial or revolution change in technology or in the market, it always goes back to those that understood the need, and stayed a customer first organization and those that lost their way.
At a simple level- And execute ... I realize I'm oversimplifying it. No, no, you're 100% right.
And execute. " I love Brad Feld. " Right?
And that's really it, right? You may not be the best. You may not fire on everything, but you got to suck less.
Well, and you've got to focus. I think the other thing is- Yeah ... you've got to be really careful- Yeah ...
because what happens in these moments of huge change is people start grabbing at too many things. " And it's actually the opposite that you need to do. It's like we've got to double down- Down on what makes sense.
Yeah ... yeah, on what we're good at- And quite frankly- ... and what our customers need.
Sorry ... you see like OpenAI announced that they were buying a company that's going to help them get into like a GitHub kind of business. Anthropic announced something.
They're both trying to climb out of who they are in terms of what they've promulgated. It feels a little bit early to do that, but okay, maybe. That's what they're doing.
They're very, very wealthy companies, so I'm not going to question them at this point. Wealthy companies that aren't profitable, Margaret. Okay, that's fair.
So there's two different things. I always think parallel expansion is smart, right? You look at things that are kind of horizontally related to your customers.
It makes sense if the GitHub thing is because they found developers are their main users, and GitHub as developers is the main user. That makes sense to me, right? You're at least saying we're after the same persona, we're after the same type of organizations.
It's part of the supply chain, so to speak, right? Then it makes sense, but I worry that's not what I'm seeing holistically. Nope.
Last category, and we can wrap. Okay. So I think open source is a huge winner out of this.
Prove me wrong. I always will say open source is a great answer, but I struggle with that because the argument, as I'm seeing it in the market right now, is not about open source versus proprietary. There's a lot of noise about the open weight models.
Yes. And we've had many discussions about the difference between open weight and open source, so- Mm-hmm ... I don't think we need to go through it again here.
But in general, what is the argument saying that there's pain there is that these open weight models, or maybe if we want to go full open models, will put pressure on the cost and take away profitability of these very closed proprietary models, let's be honest what they are, because people will see options that will bring down costs. Now you could say, well, maybe that's a good thing, because when a market gets mature, that's exactly what should happen, right? The proprietary first one out the door that was super expensive because it was the only thing, competitive models come out, whatever it is, whether we're talking about cars or dishwashers or AI.
And so there's some healthy correction that I think happens. Open source will always move the innovation and that pressure faster. I just feel like we're confusing the true power of open source and the innovation it brings with these open weight models, especially out of China, and the economic pressure that people think it brings.
That's all I'm saying. So there are open weight models that are not open source, but they do have a similar sort of liberating effect in the market. But no, what I'm saying is if I can't run my big AI models in some AI gigawatt, multi-gigawatt factory, and I'm going to run smarter AI locally on my device, in my closet, what have you, I need the innovation speed to innovation that open source gives me to invent it as I go, in essence.
And the beauty of it is with AI, I can invent it as I go kind of thing, right? And build- Well, and that is very true. Using true open source, you're able to understand the data, understand the logic, understand how you're training that model so there's not that surprise, and if you're using that, you can host it wherever you want to host it.
And that, so again, open source, it encourages the sovereignty aspect of it. It encourages innovation. It encourages just making a better mousetrap, to tell you the truth, right?
And I think, and this is a good place to end our episode, I think what's gone on in the market, what's gone on with the elections, what's going on with the technology and scarcity and everything else is it's going to force us to build a better mousetrap. Well, the way I would end that to relate it to open source is it comes to two things to me. It's transparency and trust.
And if there's two things we need right now around AI, it's transparency and trust. And we always associate those with open source, but in general- Just AI? We need it all over the place, but that's another story.
Well, yes, but yes. Okay. Fair enough.
Well, that would take another hour for us to solve. Yeah. But you're right.
You're right. Hey, Margaret, by the time people hear this, I'll be out in Salt Lake City at KubeCon. Awesome.
I know SUSE's going to be there. SUSE's going to be there- Love it ... doing a lot of stuff, so check it out if you're in Salt Lake City.
And we'll try to get back the week after when I'm back out home. That sounds great. I'll look forward to it.
Take care. All righty. Thank you.
Hey, I hope you've enjoyed this episode of The Open Current with Margaret Dawson and myself. I don't know where you're listening or watching this from, but you can catch it on all of the usual podcast stations, Apple, Spotify, or our Techstrong TV network. Have a great day, everyone.
Thanks.

