Who is actually going to win AI's next profitability test?

As AI demands more compute power as the infrastructure build-out extends further, it brings the focus back onto whether these investments will ultimately be profitable.

Today's Market Hang panel consists of Business Insider Today executive editor Dan DeFrancesco, Yahoo Finance's Pras Subramanian, SHAKTI partner Liz Harrow, Wall Street veteran Turney Duff, and Zacks Investment Management chief market strategist Brian Mulberry.

I think there was a headline that said like AI versus everything else. and I'm like, what are you talking about? Three years from now, it's one trade. Like AI is going to touch every single part of your life. And these non-tech companies are going to be dramatically different. The entire world is going to be dramatically different. and I don't even think we're scratching the surface. I don't think we're in the first inning. Like I I I'm long everything and I don't care if it's down 15, 20, 30%. Like this is so transformational like

when we do our homework, it's all about who's going to monetize this and do so profitaly, right? That's what we feel like where the market is going to go and the energy is certainly there. There's a reason that we're trading at such a high premium in the S&P 500. You know, it it's all of the capex, it's all of the AI anticipation, all of the good vibes that are baked in there. But you know, the cautionary tale is like Uber. Uber blew through their entire annual budget of IT in Q1, and what did they get from it, right? Did they were they four times more profitable? No, they had an okay quarter and they're looking to have okay earnings growth. So again, what's the return on investment? Other companies are having a different experience, but the demand overall is very strong. Every CEO that reports earnings publicly says, hey, if we had more access to more AI tools, we probably would have been more profitable. Now it's just simply that supply and demand picture where there's too few tokens and they're getting to be more expensive if you're wanting to use the frontier models. On the other side, you have more or less super search, right? The large language side is really getting cheaper and cheaper to interact with. So, how are you going to put all of this together and increase your own productivity? And that's kind of the the yin and the yang of the AI trade right now in the moment is we're we're finding that CAPEX is still the driver of valuation in the market that makes you want to favor the infrastructure trade today, but a year from now, it starts to move over to those companies once more compute is available, using the tools that Tony's talking about and becoming more productive, more profitable with less capital investment on their own. So, we're waiting for that pivot point. At the moment, it still favors kind of the infrastructure trade and and and the chips and and memory and power and all that kind of stuff. But AI will absolutely be additive to the economy. It's just going to take more compute online to get there.

Right. But the question is, are we overbuilding the infrastructure right now? No. Based on the demand structure that I just described, I really don't think so. And so that again, everything is okay as long as the CAPEX gets spent and we start to see, you know, really positive results downstream from that like we have been all year this year.

But I guess that's the the catch, right? Is you may bring up the Uber example, like spending a ton and like productivity's kind of shrugged shoulders. I that that's where I get a little nervous because you you hear stories about like, yeah, the ROI, we're not quite finding it.