Why we’re bullish on the coming application layer.
Every major computing platform has expanded human capability by removing a constraint. Cars expanded mobility, the internet expanded access to information, and smartphones made connection nearly continuous. Each one collapsed a different kind of distance between a person and what they wanted to do. Intelligence represents a different kind of shift, because it can expand our capacity to think and act, not just our capacity to consume or connect. It collapses the delta between our desire and our agency.
At Forerunner, we've spent years investing in products that reshape how people live and connect: OURA, Chime, Faire, Hims&Hers, Town, Fora, Casa, Suno, Superpower, and many more. We think AI is categorically different from what came before. Its greatest contribution may not be what it can do on our behalf, but what it allows us to become. Rather than simply helping us complete tasks faster, it has the potential to expand our capacity to think, create, organize, decide, and act on a level more sophisticated and profound. Intelligence doesn't just make actions more efficient. It makes people more capable.
Much of today's conversation focuses on AI's agency: what decisions it should make, how autonomous it should become, how nuanced its reasoning can become. Those are important questions. But they risk distracting us from one that matters most: how AI expands human agency. That shift, more than greater autonomy of the tech, may prove to be intelligence’s most enduring contribution.
Agency has two dimensions: the ability to follow through on what you already intend, and the ability to attempt things you never could before. Most people are constrained on both. For decades, software has been fundamentally reactive. Whether we were using spreadsheets, search engines, mobile apps, or today's AI chatbots, the interaction almost always began with us. We clicked, searched, typed, or prompted, and waited patiently as software performed the next step.
For the first time, that relationship is changing. Software can now observe patterns, remember context, connect ideas across time, and contribute without being asked. It doesn't simply respond to instructions; it participates alongside us. That stands to be a larger platform shift than conversational interfaces themselves.
Because most software has always waited for us to go first, it never developed the one quality that defines a good collaborator: initiative. The best colleagues don't simply answer questions when asked — they remember commitments you've forgotten, notice opportunities you haven't seen, and quietly move important work forward before anyone requests it. Software lacked this not because it wasn’t intelligent, but because it lacked memory and context. It couldn't notice what you'd forgotten because it couldn't remember what you'd said last week.
For the first time, software can hold enough of your life in context, the kind of compounding context we wrote about in our thesis on Cognitive Effects — and act on it, if you let it. When it can recall, discern, and act, the relationship shifts from operating a tool to gaining leverage with a collaborator. But collaboration is the mechanism, not the destination: the real value is that software can help us become more genuine and high-performing versions of ourselves. A calendar that schedules itself doesn't reduce your agency. It returns the cognitive load of scheduling back to you as usable capacity.
Why does this matter? Because most people don't struggle for lack of ambition, intelligence, or good intentions. They struggle because life is filled with small moments of friction: the conversation they meant to have, the article they wanted to read, the doctor's appointment they intended to schedule, the financial decision they want to tackle, the reconnection with an old friend they can’t quite get around to.
Our aspirations are often larger than our capacity to consistently execute them, and agency is what closes that gap in both directions. As intelligence becomes abundant, the constraint shifts. Information isn't scarce anymore. Advice isn't scarce. Even the ability to create is becoming abundant. What remains scarce is the ability to consistently turn intention into action. And that makes agency the defining constraint, and products that expand it will create extraordinary value.
We're already investing in this thesis. Casa and Town close the execution gap, putting your home on autopilot and learning your professional context deeply enough to act before you ask. Suno and Wispr expand the possibility ceiling: giving anyone the ability to make music, turning spoken thought into polished writing faster than most people can type. Two different expressions of the same idea: AI that makes people more capable, not just more efficient. People don't wake up wanting AI. They wake up wanting to feel healthier, calmer, more organized, more creative, more connected, and more present with the people they care about.
The enduring value of intelligence won't be measured by artificial intelligence. It will be measured by human capability.
