The Next Era of Product: Continuous Everything

A wild vision for how AI + humans unlock the golden age of software delight and how you can ride that wave now to accelerate your product and career.

What’s the Problem? is a new publication exploring the future of product building, written by a longtime product leader whose experience spans both the largest (Amazon, Airbnb, CBS) and the smallest (Haven, SET, Jacent) companies around. Like any good product discussion, we start with the question: What problem are we trying to solve?

What’s the Problem? How can PMs use this moment to accelerate their products and their careers?

AI quietly killed the product development process as we know it. Most teams just haven’t noticed. The bottleneck now is the speed of our thinking—and the quality of our taste and judgment. Here’s a vision for how product teams will work in the age of AI.

I’ve lived through more reinventions of the software development process than I care to admit. I started in the era of Waterfall, 100-page BRDs that no one read. Then came Agile, Lean Startup, MVPs, and Design Thinking, and countless others. But this time really is different. As I write this, my AI assistant is quietly churning away, building me a mobile game in the background.

We’re still in the “look at this cool toy” phase. Yes, some teams are using it seriously, but most haven’t yet evolved their processes around it. I have no illusions about predicting the future, so hello, reader of 2030. I hope you find all of this hilariously quaint.

In future posts, I’d like to dig further into what it means in detail for each phase of the process, what it means for startups vs. larger organizations; but for now let’s stay high-level.

Each role in the three-legged stool (PM, Design, and Eng) will change. My optimistic view: this leads to more and better products, not fewer people, though the distribution will be uneven at first. Every technological shift presents a moment for early adopters to leap ahead. This is that moment. To help with that adaptation here is what I imagine the life of a product manager becoming.

Product Discovery

Finding a customer problem worth solving is where we all start, and while most of us have no shortage of inputs and backlog; AI helps scale how we run discovery, and how we vet the inputs. New tools will mine huge troves of behavior data, analytics, customer feedback, emails, support tickets, and community chatter. They won’t just find net-new problems, they’ll also validate, de-duplicate, cluster, and size every hypothesis already on your list.

Historically “more problems” was actually… a problem, because humans had to sift through them one by one. But Discovery Agents change the math. They will continuously surface new issues, attach evidence, and rank them. Humans stay in the loop, but the loop gets smaller, faster, and more evidence-based.

Prioritization

Even with AI everywhere, someone still has to decide what matters most. But a lot of the mechanics—scoring, sizing, estimating, comparing get dramatically easier.

Imagine a product control center that combines:

• your discovery backlog

• your data warehouse

• your past experiment results

• your historical burn-down charts

• your team’s goals

Your Prioritization Agent could generate a RICE-scored roadmap in minutes. (Yes, assuming clean data. No, clean data does not exist. But your AI agent will help there too.)

In this world, discovery becomes automated and continuous, truly continuous discovery. Your Discovery Agent flags new opportunities; your Prioritization Agent checks if they outrank current work; the PM gets pinged only when the roadmap should actually change. You’ve reached the pinnacle of agility.

Design & Build

So we know what problem we’re solving, and in what order, but now we have to design the solution. In today’s world, this is often a prolonged game of telephone between product, design, and engineering:

• The PM writes requirements and hands them to design.

• Design comes back a week later with something in the ballpark.

• The PM gives feedback, and another week goes by for revisions.

• Eventually something “good enough” emerges and gets handed to engineering.

• Engineering spends a sprint or two building it.

• The PM looks at what was built and tests it.

By the time the PM finally gets to use the real thing, it’s weeks—sometimes months—after the original idea. And only then do they realize, “Ah, this isn’t quite the right solution,” no one’s fault, it just requires getting your hands on it to tell. But the deadline is here, so the PM decides better to deliver some customer value than none.

In our new AI product world, that loop collapses. The PM still writes something like requirements, but the primary consumer is their Coding Agent. It pulls from the team’s design system, component library, and APIs, and produces a working proof of concept in minutes.

The PM tries it, asks for adjustments, and gets a new version a minute later. Then another. Instead of two or three design cycles over a month, you might run through twenty in an afternoon. What used to take months of back-and-forth thinking now happens in a single session.

And this speed extends to customer testing. Because the marginal cost of another variant is basically zero, many features will ship as agent-coded prototypes first, letting teams A/B test multiple directions instead of sweating over a single “final” design.

Impact

So what does all of this add up to? From an industry perspective it’s great news: we’re entering the golden age of product. When you remove the friction of long development cycles, something magical happens: you can afford to obsess over details. Think Apple-level polish, but for far more products and far more niches.

For you, my product friends, I have seen the future, and with the removal of so many constraints it is the most fun time to be a product manager I’ve ever experienced. But you must evolve to be more of a jack/jill-of-all-trades than ever before. If you aren’t especially technical, time to level that up, at least in the early days that will still be important. The operator as a product manager will diminish in value, as great product sense will become more important than ever.

We’re entering a world where product managers are dramatically more empowered, not because AI does the job for us, but because it removes every drag coefficient that limited our ability to get the vision from our head to our product with minimal compromise and deliver a valuable and delightful product for our customers.

Agree, disagree, have a different take? I’d love to hear it, please leave a comment (or have your bot do it) and subscribe.