AI startups are operating at speeds faster than ever before. Teams are always experimenting, learning, and building in places where there’s rarely a clear playbook. If you want to win, you need to stay nimble, move fast, and, above all, keep everyone on the same page. Sam Altman recently said at the JPMorgan Investor Conference: "We’re going to see 10-person billion dollar companies soon." That’s not just a prediction. It’s already becoming reality. Sam Altman recently said at the JPMorgan Investor Conference "We’re going to see 10-person billion dollar companies soon." Some of the fastest-growing companies in AI today are proof: Cursor: $0 to $100M ARR in 21 months with just 20 people Bolt: $0 to $20M ARR in 2 months with 15 people Lovable: $0 to $10M ARR in 2 months with 15 people Midjourney: $0 to $200M ARR in 2 years with 10 people Mercor: $0 to $50M ARR in 2 years with 30 people Cursor: $0 to $100M ARR in 21 months with just 20 people Bolt: $0 to $20M ARR in 2 months with 15 people Lovable: $0 to $10M ARR in 2 months with 15 people Midjourney: $0 to $200M ARR in 2 years with 10 people Mercor: $0 to $50M ARR in 2 years with 30 people The reason this has become possible today is because AI is collapsing the boundaries between roles. From my own experience running product and design at a high-growth AI startup, Rilla, I’ve learned that the best outcomes come when these two roles converge. Not just working closely. But thinking as one: solving problems, crafting experiences, and shaping product direction in a unified way. In these environments, you don’t need to “hand off” a spec or “loop in” design late in the game, because the thinking happens together, from the start. Building Faster by Staying Aligned One of the biggest unlocks I’ve seen is just getting product and design closer to the customer. Not through decks or market research, but by actually talking to people. Sitting in on support calls, visiting customers IRL, and testing new ideas alongside customers. When you hear what’s confusing or frustrating in someone’s own words, it lands differently. At Rilla, every engineer spends one week doing only support. During this time, they don’t ship code or attend engineering meetings. They talk to customers, field issues, and experience pain points firsthand. Some of our best product ideas have come out of these support weeks. It’s often the week engineers come away saying, “Why didn’t we build this sooner?” “Why didn’t we build this sooner?” We’ve also seen how alignment between product and design directly impacts our speed. There have been multiple instances where we’ve designed and shipped a new feature in a single day, from idea to implementation, because the people involved shared full context from the start. There’s no need for lengthy specs or back and forths. We spot a problem, discuss it together, sketch it out, and ship. The more traditional companies we’ve worked at treated product and design as distinct lanes: product defined what to build, and design figured out how it should look and feel. That separation often created friction. Shipping took longer, context was lost, and the end result often missed the mark. Despite these issues, that setup might be fine when you’re working on well-scoped features. But with AI products, where behavior changes weekly and user expectations are still being defined, that gap becomes a real liability. What has worked better is treating product and design as one role with shared context and ownership. Over time, this builds instinct. You start to see patterns earlier, make faster calls, and avoid overthinking things that don’t matter. It’s not perfect, but it’s closer to how real progress happens when you’re building in a space that’s constantly shifting. Teams like Linear have fully embraced this model, deliberately structuring their organization so that engineers and designers own the core product work. As CEO and co-founder Karri Saarinen puts it, “No product managers, just a head of product. PM duties are distributed across engineering and design.” It’s a clear bet on tight, multidisciplinary teams that build with context, not layers. Karri Saarinen puts it “No product managers, just a head of product. PM duties are distributed across engineering and design.” Taste as a Core Product Advantage In today’s AI-powered landscape, baseline functionality is expected. Most products “just work.” What separates the forgettable from the beloved is taste. Design can’t be a final layer or a polish pass; it has to be embedded from the very beginning. Jony Ive said it best: “When somebody unwrapped that box and saw somebody gave a sht about me, I think that’s a spiritual thing… it came from a place of love and care.”* “When somebody unwrapped that box and saw somebody gave a sht about me, I think that’s a spiritual thing… it came from a place of love and care.”* That kind of care is what makes a product feel intentional. It shows up in small, thoughtful moments. A helpful message timed just right, a smooth interaction that anticipates confusion, or the decision not to add a feature that clutters the experience. These aren’t surface-level choices. They come from shared ownership and instinct, when product and design think as one, not in handoffs. In my experience, teams that prioritize taste make better calls, faster. They focus less on checking boxes and more on how something will actually feel in the hands of a user. And in a world where AI handles more of the building, that kind of human judgment becomes the real differentiator. The Evolution of the Product Builder Role AI startups are starting to hire very few PMs, if any at all. Instead, product responsibilities are shared across the team. Designers are sitting in on customer interviews, engineers are helping shape feature priorities, and everyone is involved in defining what success looks like. This shift isn’t just about leaner teams. It’s about moving faster by reducing handoffs and increasing ownership. When the people building the product also understand the “why” behind each decision, they don’t just complete tasks—they care about the outcome. They make smarter calls, notice edge cases sooner, and adjust quickly when things change. AI tools like Cursor and Figma Make make this even more accessible. Even if a designer isn’t fluent in code, they can ask AI to tweak a hover state or fix spacing—and the code updates itself. The feedback loop is much tighter and so the product gets better, faster. This sentiment has been echoed by Brian Chesky, CEO of Airbnb, when he publicly announced that he was restructuring roles within his company. He said, "We got rid of the classic product management function... We elevated design to be alongside product, so it's engineering, design, and product." This approach emphasizes the importance of multidisciplinary collaboration and shared ownership in product development. "We got rid of the classic product management function... We elevated design to be alongside product, so it's engineering, design, and product." The Future of Product Teams Peeking ahead, you’ll see the lines between product, engineering, and design blur even more, thanks largely to AI tools that speed everything up. Designers might pull components straight from a shared library and drop them into production. Engineers could dive into design systems and lay out entire user flows without waiting around for a mockup. This shift will let teams collaborate seamlessly and own the product from end to end. Designers, engineers, and product folks will all be writing specs and documentation together, each contributing directly to how the product takes shape. As AI keeps automating parts of design, writing, research, and development, everyone can focus on what they do best. The “product builder” of tomorrow will be a hybrid: part engineer, part designer, part product thinker. One moment they’re tweaking a model, the next they’re sketching a new feature or digging through user metrics. It’s messy, often overwhelming, and not always clear where the work should start. But when it clicks -- when the right person is close to the problem and empowered to solve it -- the results are fast, focused, and far more impactful.
