have spent years betting that artificial intelligence would transform the economy.
Well this month they got real proof and then they started selling, as we saw with the market reaction, AI disruption fears and Nvidia's blowout earnings.
The bar is set high regarding guidance and the known unknowns.
While we saw Nvidia fall over 5% last Thursday and another 4%.
On Friday, as investors worried about the sustainability of hyper scalar AI spending while free cash flow faces pressure from massive capbacks, there's also skepticism over a heavy reliance on a few big clients.
Right now, AI data centers are seeing explosive growth even though power constraints are a major hurdle.
The North American vacancy rate has been 1% for two straight years meaning there is pretty much no data center space unoccupied.
Well joining me as we kick off a new trading week is Jeffman, managing director for Getman asset management.
Jeff, good morning.
Thank you so much for joining us.
Thanks for having me.
Well, we all know that the data center expansion has hit a critical inflection point here so how sustainable.
The problem that we have is we have this world where we think that everything is in abundance and unfortunately some of the resources that we need are definitely in scarcity mode right now.
When we look at data center build out, we have scarcity of power right now.
We have scarcity of water, we have scarcity of special components that are needed like rare earth.
And minerals in the process, but then we also have government regulation and labor shortages.
So we're coming up to this brick wall of these five kinds of things that are preventing the build out of data centers that we need.
There are some predictions that we'll be about 15 gigawatts short of needed data center energy supply by 2030.
Even if all the current massive data center sites are actually completed and built, and that's a question mark whether that can actually happen right now.
Yes, so those 5 points that you mentioned, they're all points that we're continuing to monitor.
So given that there is this potential power shortage, what is actually being done to address it?
You would think there would be a lot being done to address it, but there's not.
On the critical minerals side, there's a lot being done by this current administration to reshore critical minerals and to lock down contracts and supply because we're dealing with the competitors of China who I don't know whether it's foresight or not, but years ago I saw this coming and unfolding and locked down a ton of these rare earth mineral contracts and supply chains that we didn't, and we let a lot of our supply chains go.
It takes 30 years to re up a mine and get a mine into production and facility and have the infrastructure to actually take the stuff from the mine and make it.
Ready for the grid.
We're not doing enough on the grid side.
You know it's unfortunate.
I would say across the board we're not doing enough.
We're not doing enough job retraining.
We have this impact that we're facing with AI where AI is causing massive layoffs.
We saw the block layoffs and other layoffs this week which were pretty large and mass for a company to lay off 50% of its labor.
Force and say that it was confident that AI could replace all of that work that was being done we're gonna see more and more of that especially in the soft skills and the white collar skills and the tech skills we should be working to transform education towards electrical infrastructure grid infrastructure and the types of things that we'll need to try to keep up with this race to.
Nowhere I like to say, but maybe it's a racist somewhere.
I don't know.
Jeff, as you mentioned this build out here, it takes years and it is very resource intensive.
So when we're talking about a process that is so intensive and laborious, time consuming, what about the data center build out?
Where is this happening across the US and what more needs to happen?
So it's interesting.
Virginia was the number one data center location in the country that is actually being surpassed this year by Texas so Texas is coming on quick and Texas is facing this interesting battle because Texas is projected to run out of water, like substantial water availability by 2030 potentially.
At the same time they're building massive data centers.
I mean it looks right now that energy from AI by 2030, the AI that needs the AI usage of energy by 2030 could be 20% of all energy available in this country, and then add in the water problems which Texas is really facing huge water problems, so.
We are running as fast as we can without thinking very clearly about the 2nd and 3rd order effects of where we're running to and what that impact is having on our environment, on our workforce, on our grid.
We've had instances already with rolling brownouts near big data centers like Microsoft's data center.
And Microsoft was like, we didn't know it was going to cause any problems, but you can't take all that power away from the grid and expect on especially a high winter day in Texas, which we've seen again this year, that you're not going to have issues with the grid infrastructure and grid availability. and Jeff finally before I let you go, you mentioned block those layoffs as well as those concerns about AI disruption, how it's going to affect all industries of all companies that are listed on an exchange such as the New York Stock Exchange, but there's a lot of uncertainty here and there are new terms.
Apocalypse right being named.
So what do you make of this?
How much of this is hype and how do you think all of this is going to shake out?
Look, I think IT skills are going to be replaced quickly, more quickly than we think by AI.
I think soft skills that humans have like understanding how to have a conversation, understanding nuance in voice, as much as AI can sound like you, it can't pick up all of these things.
It can't be empathetic in the way that humans can.
There was a big piece this weekend about replacing financial advisors.
You're not going to call your When your spouse dies and you need to talk to somebody about what you need to do, so I think those industries will be protected for a long time.
I think in the tech sense you need to be rethinking what your future job prospects are.
A lot to keep our eyes on as we head into a new trading month.
So I appreciate your time and thank you so much for sharing.
