Software's Headwinds & Tailwinds at Forefront of AI Development

Nicole Pennellides (0:00)

Welcome back to The Watchlist. I'm Nicole Pennellides. Time for our Watchlist panel. We're taking a close look here at software. Joining me for this conversation, Luke Yang, Equity Analyst, Technology at Morningstar, and Stephanie Walter, Practice Leader, AI Stack Hyperframe Research. I'm so glad that you are both with us. And so we think about what slower AI development could mean for the software space. Stephanie, you had some thoughts on this. Let's hear what your thoughts are.

Stephanie Walter (0:30)

Thanks for having me, Nicole. From the AI Stack perspective, slower frontier model development does not mean slower enterprise AI adoption necessarily. Companies still have models and AI tools to deploy.

And if anything, more emphasis on safety makes the software around AI more important. So what do I mean by that? Think about the software that provides security, governance, observability, and the application that actually turn the models into useful work.

Nicole Pennellides (1:06)

As we think about AI production, how quickly it's been happening, Luke, what should we really be focusing on here?

Luke Yang (1:14)

Yeah, sure. That's a really good question.

The main factor we are focusing on right now is how fast these enterprise software companies are coming up with their agentic AI offerings and how they embed those offerings into clients' day-to-day workflows. I think we're still at the very beginning of the organizational wide of AI adoptions into the knowledge workflows. So, yeah, going forward, we expect to see a lot of higher-end AI offerings that can drive the performance of these enterprise software names.

Nicole Pennellides (1:58)

Is that why Intuit is on your radar?

Luke Yang (2:02)

Yes. So, for Intuit specifically, we think this company has been making pretty positive progress with AI over the past few years.

But, you know, with the capability of LLM Keep Refreshing, it's becoming more challenging for Intuit to keep their user experiences at the frontier, you know, to chase after the widest range of customers possible. So, in their latest earnings, I think the company also mentioned they are going to beef up their entry-level experience. We hope these changes can help the company adapt to this pivot to AI faster, and hopefully see some improvement down the road.

Nicole Pennellides (2:51)

And when we want to see the pivot to it happening even faster, I mean, enterprise software has already been moving so fast. Stephanie, what kind of hurdles are we facing?

Stephanie Walter (3:03)

So, I mean, the hurdles that we're facing is, companies have barely begun to use the AI capabilities they already have. They still have to connect models to the right context, fit AI into how customers actually work, and show that the results are reliable.

Just because a company has AI features does not mean those features are actually useful by default. And so enterprises really need to ensure that customers find out value in what they're doing with AI, to keep using them and to keep paying for AI.

Nicole Pennellides (3:43)

Yeah, you know, maybe I should turn the question around because I originally thinking challenges and hurdles, but what would be the tailwinds? Because we are seeing everything we were talking about security, identity, governance, those were some of the areas Stephanie that you were focusing on. What are the tailwinds?

Stephanie Walter (4:02)

So, I mean, I think that, you know, AI work did not stop just because a model release slowed down. Developers still have to choose model, build context from enterprise data, they have to connect tool, test behavior, decide where people must approve an action. So, these requirements grow, and software has to provide this capability across business systems. So, a steady or model release could give teams more time to make applications reliable and it gives a lot more opportunity to the supporting software in the AI stack that makes sure that AI can work correctly and safely.

Nicole Pennellides (4:45)

We're waiting on Intuit for its investor day on Thursday, Luke, and I know in your note, you focused on a few of the things with the do-it-yourselfers.

Credit Karma has done very well for Intuit. I remember that in the last quarter reading that. They also have the Turbo Tax. What are you looking for and what do you expect to hear now from the company on investor day, specifically right here in the New York term?

Luke Yang (5:11)

Yeah.

For the investor day that is going to happen later this week, we are really focusing on the company's new product developments. So far, they already have a lot of AI offerings, but we want to hear more from the management, how to persuade their users to leverage those offerings more, and what are the value propositions Investor Intuit can provide to their users with these AI agents. So specifically, we see Credit Karma is doing very well, and we think AI is a key factor that can drive even more positive user interaction and growth. For the DIY text filing offerings, the performance are a little bit underwhelming. So we also want to see how Intuit management is exactly planning to use AI to renew the entry-level experience.