#77: Opportunities Because of AI That Aren’t AI (Part II)
Industrials, autonomous vehicles, and energy
Last week, software stocks fell, in part due to news that Anthropic’s Claude Code is starting to tackle niche software applications long reserved for incumbent enterprise software companies. Firms like Adobe, Salesforce, and ServiceNow are down roughly 24%, 20%, and 30% since the start of the year, respectively as of February 9.
The growing theory is that as more code becomes AI-generated, traditional software companies will struggle to grow because firms will develop internal tools in-house. Instead of relying on Salesforce for your company’s customer relationship management (CRM) tool, you can use a program like Claude Code to build your own custom solution.
Based on much of the online chatter I come across, I sit with majority who claim that the software stock sell-off is overblown and AI isn’t going to eat the entire software industry overnight.
But then a headline like this one makes me question everything: Goldman Sachs taps Anthropic’s Claude to automate accounting, compliance roles.
Anthropic is beginning to partner closely with enterprises to embed AI into real-world, niche workflows. The takeaway here is less that Goldman Sachs is looking to invest in automation, and more so that Anthropic will learn much about solving the types of problems facing complex, financial firms.
Through Claude Cowork (similar to Claude Code, but for non-technical tasks), users are able to automate daily tasks like market research, synthesizing customer feedback, and reorganizing your file system based on prompts typed into a prompt bar. Their AI is focused on automating as many workflows as possible, it seems.
Now imagine through Anthropic’s work with Goldman Sachs (and likely other enterprises), the level of context they’ll receive regarding which types of problems are worth diving deeper on. When Anthropic rolls out more specific use cases that firms like Goldman Sachs co-sign, it could be the reckoning moment for software companies without large proprietary datasets or omnipresent brand names.
OpenAI is trying to do the same thing too. Not too long ago, I saw a job posting where OpenAI was looking for a relationship manager to partner directly with Deloitte (it’s since been removed). A role dedicated to customizing OpenAI technology for Deloitte and its clients. A recent job post suggests OpenAI is still all-in on building out their enterprise relationships.
My hunch is this is less about making money off one client, and more about learning all the problems their client sees on a daily basis. We’ve talked about it before at Relentlessly Curious: the value is in the application layer. But with what Anthropic is building, the foundational layer and the application layer are starting to blur.
With all this said, the software market is likely to fluctuate over the next few years as the impact that foundational model companies have on this industry has yet to shake out.
So, let’s dive into a few industries that are likely to experience a tailwind because of AI.
Industrials
Big Tech has signaled plans to spend nearly $700B on AI capital expenditures (capex) in 2026. For reference, the country of Argentina has a GDP of $700B. Between Amazon, Apple, Meta, and Google, that single expense line would collectively rank with the 25th largest GDP in the world.
So, where’s all this money going to end up? Building data centers, of course. And the infrastructure to support these massive investments.
Companies directly tied to building data centers, like Caterpillar (construction equipment), Honeywell (heavy industrial inputs), and Martin Marietta Materials (concrete), are likely to go on a tear this year. Caterpillar itself has been directly linked to data center announcements, and the others stand to benefit immensely from the GDP of Argentina being pumped into data centers.
When it comes to industrials, think as basic as it gets: inputs. Steel companies, copper mines, and cooling equipment manufacturers are all poised to win major new contracts this year with AI-related demand.
Caution: although I believe this picks and shovels play is one of the safer AI investments, pay close attention to language surrounding funding “commitments” versus breaking ground. If reports start to flurry in July that Big Tech has barely spent a fraction of their headline commitments, I expect this sector to take a hit.
Prefer an ETF? Check out FIDU from Fidelity.
Autonomous Vehicles (AV)
Uber is far and away the de facto ride sharing platform in the US. In Q3 2025, Uber recorded 180 million monthly active platform consumers, while Lyft notched 25 million. I’d say that’s capturing market share.
The rise of Waymo and autonomous driving is gaining momentum. However, in five years, my guess is that Waymo is not the only AV provider. I believe there will be plenty of new entrants that we haven’t even heard of today that will be major players in the early 2030s.
But you know who doesn’t have to deal with competition in the AV market? Uber. They can serve the same role they do now, as the platform. Buy an AV and let it drive Uber rides while you sit at home and watch TV. Or Uber may partner directly with AV manufacturers to operate its Ride Share network. Either scenario, Uber acts as the platform and doesn’t have to compete on the make and model. It is worth noting that this assumes that the regulatory environment supports such a seismic change in transportation.
Candidly, I’m not sure where Tesla will fall here. Tesla’s revenue continues to fall year over year, and Musk seems more focused on the SpaceX, xAI, and X combination. But I certainly wouldn’t bet against him.
Energy
Those data centers aren’t going to run without energy. With major enhancements to the electrical grid needed to power data centers, electric utilities located near planned data center expansions are poised to benefit from increased power demand.
Plenty of data centers are planned to be built in the Midwest and South. See below for a list of planned data center locations by major tech firms, including but not limited to:
Amazon: Ohio
Google/Microsoft/OpenAI: Texas
Meta: Louisiana
xAI: Tennessee
One approach is to focus on energy companies that serve these regions. American Electric Power covers the Ohio region, while Duke Energy also covers Ohio, as well as the Carolinas, and Tennessee. It’s worth diving deeper to better understand these firms’ energy mix, but their geographic coverage is a competitive advantage.
Prefer an ETF? Check out XLU from State Street.