AI Is Already Ahead of Its Skis

Shimmy says, Shimmy says, Shimmy, Shimmy, Shimmy says, ask me almost anything! Hey everyone, it's Shimmy. I'm coming at you today from Old Town Dubrovnik in Croatia.

We were in Prague for SUSECON, and we've done a little Europe hopping here. We went to Vienna, Split, Dubrovnik now, and then from here, we're going to Italy. " I read a lot.

I check out a lot of different sites, stories, and news sources, of course, Techstrong's as well, and I definitely see a counter-wave building. It's a counter-wave to AI. And at first, I thought it was just people being malcontents or the cynics, but I think you got to acknowledge that it's maybe more than that.

This week, it became a little clearer to me. It's kind of like, think of AI as a downhill skier. You ever watch Olympic downhill skiing, right?

You got guys that are, and gals for that matter. This is not casual when you're in the downhill. It's a controlled chaos, right?

You're going 80, 90 miles an hour on ice or snow, threading gates, and there's no margin of error. The best in the world, the people who win the medals, they don't play it safe. They live on the edge.

They're living right on the edge between control and chaos. Because if you're not on that edge, you're not winning a medal, and you're not pushing the ball forward. But I think that we miss something in that.

That when a downhill racer gets ahead of their skis, there's a moment, just a moment, maybe they catch a little air or whatever, where they're still upright, they're still moving forward, but they're not in control anymore. At that point, inertia, gravity, competitiveness has taken over. And they're no longer carving, as skiers say.

They're just reacting. And I think that's where we are with AI at the moment. And it's not just the AI technology, it's us as users of this technology.

We look around. On one side, you got companies that are cutting people to take their salary, not to replace them with AI, but to invest that money into AI infrastructure. On the other side, you got salt of the earth, everyday people, farmers, musicians, pastors, activists, all pushing back.

They don't want data centers in their backyards. They don't want data centers, period. They're worried about energy for them.

They're worried about water for them. They're worried about the job situation. It doesn't create the jobs they thought.

They're worried about their IP and copyright that might be taken up into these frontier models. They want to know, why is this being done to us? Not for us, to us.

And that's an important distinction. Until people realize it's being done for us and as humanity, not to us, I don't know if we're going to have to fight that battle. And now even you got Wall Street starting to blink a little bit and push back.

I read an article that the people building this huge data center complex down in Abilene, Texas, had a hard time getting Wall Street to float the money for it because it was Oracle that was the primary tenant of the data center. And they think Oracle is a little bit overextended in terms of how much they're investing in AI. God, they laid off 30,000 people to invest in AI or something like that.

And they balked at it. The data center operator had to get Microsoft to take over that project before Wall Street put the money in. But Wall Street did put the money in.

Let's remember that. " If we look at those points, the people pushing back, Wall Street getting a little antsy about putting more money in, companies cutting salaries, not to replace workers, but to build more AI that will eventually replace the workers. But the fact is, they're not replacing them right now.

When we look at each of those on their own, yeah, you can make an excuse for that one. You can make an excuse for this one. But I think when you start putting them all together, well, that becomes a signal.

And I think that's the signal we're seeing right now. We've seen, maybe not at this scale, but we've seen versions of this before. In the dot com era, right?

All of a sudden, bottom line didn't mean anything. It was all top line, and we really got ahead of revenue. com.

Cloud, in the early days, got ahead of cost savings, right? People rushed headlong into the cloud and then realized they weren't saving any money. Social media grew out of control before we even realized what hit us, and now we're stuck with it.

There's always a stretch with new technologies, with new ways, where the story gets ahead of the math, and I think that's where we are with AI right now. We even have an executive from NVIDIA sayingRight now, it's probably cheaper to have humans do the work that he needs done rather than AI. Think about that, the irony of it.

An executive at Nvidia saying it's cheaper to have humans do the work than AI. And if that's the truth right now, well, then we've got to re-look at where we're going, and maybe it is time to tell the downhill skier he has to slow up a little bit. We start with the economics of this thing, as I mentioned.

For all the talk about efficiency, AI isn't consistently cheaper than people. In a lot of cases, it's more expensive. And as I said, this is coming from an Nvidia executive himself.

If you want references for this, I actually have an article on this up on Techstrong AI you could go to with all of the references. But now, doesn't stop the investment. I'm just saying he said that.

Because at the end of the day, I think what we're really making is a bet on the future. Not tomorrow, the day after, and the day after that. It's not on today's economics.

It's where they think this all is going to land, where we're all going, and the question is when and how fast, right? But in this meantime, we're running models at scale. But it isn't just software, it's the infrastructure.

It's power, as I said, water for cooling. Real estate. It's creating a real estate boom.

The plumbers and HVACs and electricians, the whole supply chain of chips and memory. Costs that don't magically compress because a slide deck says they will. We're living the reality of that right now.

Now, on top of that, you got to layer in, what are these companies actually doing, as I said, right? Meta Platforms trimming headcount, Microsoft doing buyouts. We could call it optimization.

We could call it refocusing. It really doesn't matter what you call it. I tell you what you call it, though.

We're firing, laying off people. The direction here is obvious. Companies think they're going to be able to do more with less.

Less money towards people, more money towards GPUs, data centers, and AI infrastructure. But that only works short term, long term, medium term, if the economics follow this model. Right now, I don't know if that's a great bet, in all honesty.

And you know what? Let me be clear. I'm a huge proponent of AI.

We use AI so much in Techstrong. Agentics have changed the company over the last few months. But I also see patterns, right?

You still need the people. It doesn't do everything we want it to do as well as we want it to do it, as fast as we want it to do it. And that's us.

We're in this tech bubble like you are. But step outside the tech industry, and I think things get even more interesting, radical, scary. This isn't about just us tech people arguing over this anymore.

As I said, you got farmers worried about having water for their crops. You got musicians worried about their work being pirated and ripped off to use to make synthetic music. You even have religious leaders asking what this does to communities and values.

When the local church is against you, it's hard. That's not a coordinated movement. This is kind of random right now, but that's what happens when something shows up in people's real lives.

Because once AI becomes physical, it's not abstract anymore. A data center isn't the cloud when it's in your backyard. It's land, it's energy, it's politics.

And then there's the trust issue, and don't underestimate this, right? The same companies that are building the AI, they're still carrying baggage from social media. People remember how that all played out.

So when they hear, "Trust us," well, you know what they're thinking. It lands just a bit differently, and you can't blame them. Now we're at the part that really shifts this into another gear, financing.

The amount of money is crazy. We're talking almost a trillion dollars. You're starting to see deals so big that they're testing the systems funding them.

Banks like JPMorgan Chase, gold standard, are hitting exposure limits on their AI exposure. Massive infrastructure plays tied to companies like Oracle are getting harder to spread across the market. OpenAI, we were all for it, but how many trillions of dollars are we putting in here?

But yet that's not stopping this headlong rush into AI. But I do think it changes it. I do think we maybe need to just slow down a little bit, get off the downhill.

Let's go to the slalom, not the bunny hill, but get off this downhill. We need to reshuffle stuff. We have to make it tighter.

We have to make the economics work, because if the economics don't work long term, this doesn't work. And we all want that future to get here, but getting there is the key. So let me put this all together for you.

The technology is obviously still maturing. It's getting better every day for sure, but it's still maturing. The economics we have not fully proven out, and I think anyone who tells you differently is lying.

The public is already pushing back. Any one of those individually might be manageable, but when we look at all four at once, that to me, folks, is you're out ahead of your skis. And here's the key, and I mean this.

This doesn't mean AI fails. AI isn't going to fail. It's going to get there.

The capabilities are real. The direction it's taking is clear. But inevitability doesn't mean smooth, and it definitely doesn't mean controlled.

Progress is always lumpy, and those lumps, the upturns, the downturns, people live and die on. And we've got to realize that and try to smooth that out as much as we can, because there's still a gap between what AI can do in a demo and what it can do reliably, economically, at scale, in the real world. That gap, as I said, is where all this tension is coming from, and it's the gap we need to cross, that chasm we need to cross.

From the outside, this all looks just peachy. Companies cutting jobs today to fund something that hasn't fully proven itself yet, though. Communities being asked to absorb the footprint, and the financial system underwriting, it's starting to ask how much is too much.

That's not a stable line down the mountain, if you were a skier. AI hasn't wiped out, not yet, but it's living on the edge, and it's not carving clean turns at this point. And at this speed, you don't get a lot of chances to recover.

You know what? I'm going to be on the road next week. We're on Italy.

I'll try to do a "Shimmy Says" from there. But for now, this is Shimmy. I hope you've enjoyed it.

I'd love to hear what your thoughts are on AI. We'll see you next time. Shimmy says, Shimmy says.

Shimmy, Shimmy, Shimmy says. Ask me almost anything.