But here’s the thing—I’m not chasing the next agent. I’m hunting for the next generation of software.

That might sound unsexy compared to the sweeping autonomy of agentic AI, but if history is any guide, it’s not agents that reshape industries—it’s applications. Tools like CRMs, email, spreadsheets, and calendars are the real engines of work. And every time computing undergoes a paradigm shift, it’s these tools that get rebuilt—and in doing so, they redefine how we work.

When the web overtook client-server, Siebel patched its software while Salesforce reimagined it. When mobile became dominant, the winners weren’t the ones who augmented with an app, but those who rethought entire products—WhatsApp, not Skype; Instagram, not Flickr.

Now with AI, we’re seeing the same dynamic: incumbents layering generative features on top of legacy systems—“AI-powered” summaries, data-entry assistants, and chat bots—but all within rigid paradigms from a pre-AI world. Even as tech giants like Salesforce and Google pour billions into AI and acquire early-stage disruptors, they still can’t rewrite the core assumptions baked into their legacy products.

It’s lipstick on a pig.

Yes, they have distribution—but in every platform shift, distribution advantages erode when the foundation of the product changes. The true innovation comes not from bolted-on features, but from rethinking what the product is.

The opportunity isn’t adding AI to CRMs. It’s entirely reconceptualizing how all software works—starting with the tools we rely on every day. Imagine a CRM that doesn’t just log data but generates it—hunts for your leads, drafts your outreach, bids with context, preps you for meetings, and coaches you in real time. A spreadsheet that interprets intent and self-builds models to expose insights to you. A calendar that understands not just when you're free or busy, but that a regular 1:1 can be moved to make room for a board meeting—or that a visiting exec with limited time should take priority over a standing internal check-in. That’s what AI-native looks like—not just automation, but a fundamentally new interaction model that reshapes the flow of work and decision-making.

AI agents will play a role in this future, but they’re not the destination—they’re ingredients. The magic happens when they’re orchestrated into cohesive product experiences built for how we’ll actually work in an AI-native world.

So yes, agents are impressive. But I’m looking beyond the next agent. I’m looking for the next generation of applications.

‍