In May, Polish tech had its champagne moment. Viktor.com (founded by ex-Meta engineers Fryderyk Wiatrowski and Peter Albert) raised $75 million from Accel, the largest Series A in Polish tech history. Their product is a digital AI employee that lives in Slack and Teams, remembers everything, and hands you finished work. Ten weeks after launch it was earning $15 million a year.

Five weeks after the round, Anthropic shipped Claude Tag: an always-on Claude that lives in Slack, remembers everything, and hands you finished work. The same product, as a feature, available overnight to every Anthropic enterprise customer.

Anthropic wasn’t copying anyone. It had been using Claude Tag internally before the launch, and 65% of its product team’s code now comes from it. Viktor’s founders were building, without knowing it, a feature that was already on the lab’s roadmap.

That collision is why your AI product idea probably doesn’t matter.

Intelligence is becoming table stakes

The old playbook for AI founders was simple: pick a use case, wrap a frontier model, move fast. It worked while the frontier was scarce.

It isn’t anymore. Epoch AI tracks the gap: the best free, open models are about four months behind the frontier, and models that run on a regular gaming PC are six to twelve months behind. The intelligence you pay OpenAI for today will run free on a $2,500 gaming rig about a year from now.

Meanwhile the top itself has converged. Last week Grok 4.5, GPT-5.6 and Meta’s Muse Spark 1.1 all shipped within 48 hours, each claiming its own sliver of state of the art.

When everyone has superb intelligence, intelligence stops being a business. Even OpenAI’s GPT-5.6 pitch wasn’t “smarter” but “more useful work out of every token.” Usefulness means product, and the labs are coming for it.

The kill zone

Andrew Ng has said it for years: in the AI stack (chips, cloud, models, applications) the application layer has to make the most money, because it pays for everything underneath. The labs read the same map. So they moved up.

Anthropic shipped Claude Code, then Cowork, then Tag. OpenAI shipped ChatGPT Work. Google bundles Gemini into everything with a login. And SpaceXAI skipped the building part: on June 16 it bought Anysphere, the company behind Cursor, for $60 billion in stock, the largest startup acquisition ever.

Why would Cursor sell? It earned about $4 billion a year, more than any AI app in history. But every Claude token it served, it bought from Anthropic at full API price, while Anthropic sold Claude Code, a direct competitor, powered by the same model at cost. Within a year Cursor’s share of AI coding fell from 41% to 26%; Anthropic’s grew toward half the market. Cursor did everything right and still needed a $60 billion lifeboat, because its supplier had become its competitor.

That’s the squeeze from above. There’s another from below. Wispr Flow sells dictation for $15 a month; free tools running open models on your own laptop now do the same job at 95-97% accuracy. Nobody bought or copied Wispr Flow. The free floor simply rose to meet it.

With labs above you and free models below you, the thin layer in between is the kill zone.

The apocalypse that wasn’t

In February the market briefly panicked about all software. When investors saw what Claude Cowork’s agents could do, software stocks lost roughly $300 billion in a single day, the “SaaSpocalypse.” By April the prices had mostly recovered.

The panic taught a useful lesson. The software that died was the simple kind, $29-a-month tools whose one trick a general model now does for free. The software that survived was the deep kind: systems like Epic (hospital records) or IQVIA (clinical trials) that hold years of data, integrations and regulatory approvals. Nobody canceled Jira and asked an agent to rebuild it, because ripping out twenty years of process is harder than paying for it.

The labs won’t build a hospital records system. Neither will your customers. That’s the gap where durable businesses live.

What to do instead

Go deep in one industry, and know what you’re really building. Harvey, the legal AI company, is worth $11 billion and serves most of America’s biggest law firms, but its moat isn’t data. The legal data belongs to LexisNexis and Thomson Reuters, and Harvey isn’t even allowed to train on it. What Harvey owns is everything around the data: it sits inside Word, passes the strictest compliance reviews, and knows how law firms actually work. In most industries someone already owns the data; the workflow, the certifications and the relationships are still up for grabs.

Or sell a service instead of a product. Software is a $1 trillion market; services are $16 trillion. General Catalyst has spent $1.5 billion buying accounting firms and call centers and rebuilding them around AI. One of those bets reached $100 million of yearly profit in under two years. The same play works alone: one consultant with AI can serve what used to take a team of ten. And a service has a defense software never will: a human who answers for the result.

Or sell to the labs instead of competing with them. They rent their data centers (Anthropic runs Claude on CoreWeave’s machines) and they buy the tools they work with. Neptune.ai, a Warsaw platform researchers use to monitor model training, counted OpenAI as a customer for a year; in December OpenAI bought the whole company for up to $400 million. The labs also can’t build neutral tools. Most big companies run several models side by side, and the scoreboard comparing them can’t be made by one of the players. Just avoid anything a free plugin can replace: Pinecone, once a billion-dollar database startup, was undercut by a free Postgres extension.

Or take these ideas to a job in tech, the right kind of tech. The safe employers are companies whose product carries domain depth no lab will rebuild: an Atlassian, whose tools hold twenty years of how software teams work, or a marketplace like Booking, whose moat is the list of hotels and the relationships behind it. People will keep traveling no matter how smart the models get. A job applying AI inside a company like that beats being employee number forty at the tenth wrapper startup, and pays in salary, not lottery tickets.

And the fifth path, half a joke: go physical. Land, food and fitness are businesses where AI helps but can never be the product. The story that AI researchers are quitting to buy farms turns out to be a myth, but what they actually buy is more telling. Sam Altman keeps “guns, gold, potassium iodide... and a big patch of land in Big Sur I can fly to.” Ilya Sutskever promised his team a bunker before AGI. Geoffrey Hinton tells young people to become plumbers. Draw your own conclusions.

But Cursor got $60 billion

Fair objection: if wrappers are doomed, why did one just sell for $60 billion?

Because selling was the only way out. Cursor couldn’t win a price war with its own supplier, so it sold at the top, the smartest move on the board. Perplexity waited too long: Apple considered buying it for about $14 billion and never made an offer, then ChatGPT and Google copied its search product, then its $400 million Snap deal collapsed. The company still grows, but its best exit is gone.

So yes, you can build a wrapper to sell it. Just be honest that this is the plan: a race to get acquired before you get copied. The window is real, but it closes fast.

What if you still want to build one?

Then build it with open eyes. There are two honest plans.

Plan one is the Cursor play: take the obvious idea, run faster than everyone, and sell before the labs catch up. It can work; it just made a few hundred people very rich. But it’s a sprint against a closing window, so speed is the whole strategy, and falling in love with the product is how you miss the exit.

Plan two is to build outside the labs’ path, and you can test your idea with three questions. Does the next model upgrade make your product better or pointless? If GPT-6 improves you, good; if it replaces you, stop now. Is your customer buying intelligence, or something around it: a certification, an integration into their industry’s systems, someone to blame when things fail? Intelligence belongs to the labs; everything around it can be yours. And could a lab honestly sell your product? Anthropic can’t sell the dashboard that says Grok is better, and nobody in San Francisco wants to face your industry’s regulator.

Whatever you pick, don’t repeat Cursor’s structural mistake: if one lab is both your supplier and your competitor, your margins belong to them. Route between models and keep the open ones as your fallback. They’re four months behind and free, which makes them your insurance as much as your competition.

The idea was never the moat

If your idea can be a feature of a frontier model, some lab is already building it; ask Viktor. If it can be a free tool, it will be one within months.

What lasts is what the labs can’t or won’t touch: an industry whose workflow and trust you own, a service with your name on the result, tools the labs need to stay neutral. Or, if you take Hinton seriously, a plumbing license.

The product management role is having an identity crisis right now. And to be honest, it’s about time.

I’ve been watching this happen for over a decade. When I started as a PM, you had to know how to code - or at least be able to fake it convincingly enough. You had to design things in Photoshop or later Sketch. You had to figure out go-to-market on your own because there simply weren’t enough people in the company who knew how to do this stuff.

Then somewhere along the way, the role got complicated. It became specialized, professionalized. And in many companies, it became completely hollow - all theater, no substance.

Now AI is forcing things to change, and a lot of PMs aren’t going to like it.

How the PM Role Evolved (And Why It’s Evolving Back)

Let’s talk about what “product manager” actually means in 2026. The role has gone through an interesting evolution:

The Original PM was essentially the founder. You did everything because you had to - sales, marketing, design, code, strategy, all of it. Every startup still has this person, usually the CEO or CTO. Thirteen years ago when I started, this was the norm even in larger companies. You couldn’t just write a spec and throw it over the wall to someone else. You had to roll up your sleeves and make things happen.

Then as companies grew and products became more complex, the role specialized. The generalist PM split into different types:

The IT Project Manager emerged to manage projects, not products. They’re focused on scope, timeline, and budget, but the strategy comes from business stakeholders or external clients. They’re executing someone else’s vision, keeping the trains running on time.

The Product Owner showed up with Agile methodologies. They’re backlog managers who run all the ceremonies - standups, retros, planning sessions - but they don’t actually own the strategic direction of what they’re building. Marty Cagan calls this “product management theater,” and that’s exactly what it is. All the motions, zero substance.

The Big Tech PM is the one who actually owns strategy and direction. This is the “real” PM role that everyone aspires to. But here’s the thing - they’re pure strategists. They write documentation, align stakeholders, do research, pitch ideas. Then they hand everything off to specialists. Designers figure out how it should look, engineers figure out how it should work, marketers figure out how to launch it. They’re conducting the orchestra, not playing the instruments.

This specialization made sense when coding was hard and coordination was necessary. But now AI is changing the equation.

The Product Builder is the new type that’s showing up now - and it’s basically the original PM coming back, but with AI superpowers. They have a core skill (engineering, design, or PM) but they’ve picked up 2-3 adjacent skills that let them ship end-to-end without handoffs. An engineer who learned design and product thinking. A designer who can code and run growth experiments. A PM who can build prototypes and launch them.

They use AI tools to level up in areas outside their expertise, but they know enough to judge what good looks like. The key is they combine both: the strategic judgment to know what to build AND the tactical ability to build it themselves.

From Assembly Line to Rapid Validation

The old way of building products looked like an assembly line. UX Research would study users, hand off insights to Design. Design would create mockups, hand off to Product to write specs. Product would hand off to Engineering to build. Engineering would hand off to QA to test. Then maybe you’d iterate.

Each specialist did their part. Each handoff took time. The whole cycle could take months.

This made sense when coordination was cheap and building was expensive. You needed all those specialists because each domain - research, design, engineering, testing - required deep expertise that couldn’t coexist in one person.

Now you can build a working prototype in the time it used to take to finish the research phase. With tools like Lovable, Replit, and Cursor, one person can go from idea to functional software in hours. Test it with real users. Get feedback. Iterate. All before the traditional assembly line would have moved from UX Research to Design.

When you finally need to build the production version, you show the AI your validated prototype and user feedback. It figures out what to build from there.

I see this transition happening in real-time. A Principal Designer I talked to, isn’t waiting for the assembly line anymore. He’s not just designing interfaces—he’s vibe coding working prototypes. Electron-based desktop apps. Full features. He validates ideas in days, then hands engineers a working prototype instead of static mockups.

He’s one of the most productive persons on his Agile team. Why? Because unlike in the old world, when everyone was waiting for the previous specialist to finish their work, he is able to prototype, test with users and ship on his own. The bottleneck isn’t his speed anymore—it’s everyone else’s handoffs.

That’s the future showing up unevenly distributed. He is a Product Builder surrounded by specialists still playing their roles. And the specialists have to be moving slower because of the handoffs.

But here’s the trap: when building becomes this cheap, companies face a new problem. Abundance.

In an economy where production costs approach zero, everyone can build everything. More prototypes. More apps. More products. The constraint isn’t “can we build this?” anymore. It’s “should we build this?”

Amazon saw this coming. They built Cairo, a development environment that forces you to write testable specifications before generating any code. When Amazon - a company that profits from speed - decides the most valuable thing is to slow you down and make you define what you want clearly, that tells you everything.

The bottleneck moved from execution speed to clarity of intent. From “how do we build?” to “what deserves to be built?”

From Two Pizza Boxes to One

You’ve probably heard the old Agile saying - a perfect team is small enough to be fed with two pizza boxes. That traditionally meant 4-6 people: a PM, a designer, 2-3 engineers, maybe a researcher.

But with product builders who can design, code, and ship autonomously, you’re looking at 2-3 people doing the same amount of work. We’re moving into a one-pizza-box reality.

What I’m seeing in practice is that when everyone on a team is a full-stack product builder, most people work solo on their own pieces within small teams. Each person owns their area end-to-end. There are fewer dependencies, less passing the brick around between specialists, less waiting on other people to unblock you.

This completely changes team dynamics.

Dailies stop being about alignment and blocker detection. Because when people aren’t dependent on each other, there simply aren’t that many blockers to discuss. Instead, these meetings become more like social coffee breaks - early morning check-ins where people catch up.

And you know what? That’s actually fine. People need social connection at work. But let’s not pretend it’s still a coordination meeting.

But if teams are shrinking and everyone works solo on their own pieces... what happens to all the organizational structures we built around coordinating specialists?

Three Things Nobody’s Talking About

The PM Career Ladder Collapses

Most companies have a nice, clean progression: APM → PM → Senior PM → Principal PM → Group PM → VP of Product. It’s linear and predictable.

But if Product Builders are working solo and owning things end-to-end... what exactly are you promoting them into?

You can’t really “manage” 2-3 autonomous builders the same way you managed a 6-person team of specialists. There’s no coordination overhead to oversee anymore. No handoffs to optimize. No ceremonies to facilitate.

The economics prove this. Cursor generates $16 million per employee. Midjourney hit $200 million in revenue with 11 people. When one person with the right skills and agent infrastructure can produce what a 20-person team used to build, they capture most of the value that used to be distributed across that team. The middle management layer that used to coordinate all those specialists simply isn’t needed anymore.

The career ladder collapses into something much simpler: Builder → Senior Builder → Founder or Executive. The Group PM role coordinating other PMs stops making sense.

So if you’re a Senior PM today and you’re banking on that Group PM promotion in a couple years, you might want to start rethinking your career path.

Hiring Becomes Completely Different

How do you even interview for this new type of role?

Read more

Something’s shifting. Articles about new AI tools flood your feed weekly. Your developers are coding in English more than the actual code. Every meeting circles back to AI. The ground is moving.

AI lets you code 10x faster. But most products built with AI? Still shit.

Why? Because AI doesn’t teach you what to build. Only how to build faster. Coding faster is just one piece of the puzzle. Maybe not even the most important one.

The tech industry is changing. There are new tools, new processes, and new ways of thinking. There’s a formula emerging from the chaos. And a new role sitting at the center of it.

The Superheroes Who Taught Us

Everyone has a favorite superhero. Turns out we have the same one. It’s Batman - because he showed up without having supernatural powers (besides cool gadgets). He had a formula.

But the superheroes can be real. You’ve seen them - the coworker who turns chaos into customer love, the manager who ships impossible things. They have a formula too.

Piotr: The first superhero I met wasn’t fictional. His name was Jan. He was my manager. Whatever product that guy touched turned into customer love.

Watching him work was like watching someone play a different game than everyone else. Same company, same team, same people to deal with. Completely different outcomes.

He knew how to pick the right problems, ship fast, leverage technology to gain an edge, and work effectively with both technical and non-technical people. His formula created repeatable product success.

Wojtek: For me, my own Batcave was DocPlanner. Being part of a rag-tag crew of early employees who scaled a 25-person, garage-style operation into an international business with offices in Warsaw, Rome, Istanbul, and Barcelona was a defining experience.

Back then, we had no idea what we were doing. Everything we touched was new to us. There was no formula to lean on. We had to figure it out ourselves, through trial and error. But we did it.

What sets people like Jan and the early DocPlanner team apart from everyone else doing the same job?

Not talent. Not luck. A repeatable system.

Best vs Rest

What separates the best companies from the rest of the companies? We call it the Superhero Formula. When applied you can see the clear differences when it comes to the way they operate:

1. They solve different problems. Most teams pick problems their executives care about. Great product people pick problems that unlock 10x more value when solved.

2. They validate differently. Most teams ask customers what they want. Great product people watch what customers actually do. There’s a massive gap between the two.

3. They ship without the bullshit. Most teams drown in process - roadmaps that change weekly, stakeholder reviews that kill momentum. Great product people know which corners to cut and when.

4. They work with humans, not org charts. Most people struggle with “difficult” team members or “unreasonable” stakeholders. Great product people understand how different minds work and leverage that instead of fighting it.

5. They build cultures that build great products. Most teams depend on a few strong individuals to keep the bar high. Great product people create environments where high standards, ownership, and pace are shared by everyone.

6. They use AI without becoming lazy. Most people use AI to generate more slop, faster. Great product people use AI to learn faster and eliminate repetitive work - while keeping their judgment sharp.

The formula works because it’s not about doing more. It’s about doing things differently.

And here’s what separates good from great: knowing when to break the formula. Once you understand the system, you learn which rules to bend, which corners to cut, and how to be less predictable than your competition.

This formula matters more now than ever. Here’s why:

AI made a promise it only half-delivered

AI was supposed to make building 10x faster. And it did. Sort of.

Your engineer writes code quicker. Your designer iterates on mockups quicker. Your PM drafts specs quicker. English became the new programming language. People who thought coding was black magic can now build things.

But the building process is still slow as hell.

You’ve probably heard about the Build Measure Learn loop:

That’s not how product development works in real life. It’s research, ideation, design, coding, launch, marketing, and so much more. Each step answers a different question: What should we build? How should it work? How do we build it? How do people find it? How do we know it works?

And here’s the problem: engineers, designers, PMs, marketers still operate in silos. They wait. They coordinate. They waste time in meetings. They miscommunicate.

The bottleneck was never individual speed. It’s the gaps between people.

AI made everyone faster at their job while they wait for everyone else. That’s not 10x. That’s maybe 1.2x. And that’s a maybe - we’ve seen people working more slowly.

The real unlock isn’t faster individuals. It’s eliminating the handoffs.

The Solution: Product Builder

Marc Andreessen recently said something that nails it: the future belongs to generalists who can work across 6-8 fields. Deep expertise still matters, but broad knowledge plus AI tools beats narrow specialists in most areas.

We’re seeing something even more radical. AI isn’t just creating generalists. It’s letting specialists gain new superpowers and become T-Shapes or V-Shapes:

Marc Andreessen is essentially describing what we call Product Builders.

A Product Builder is someone with a core skill (engineering, design, or product management) who acquired 2-3 adjacent skills to eliminate handoffs and ship autonomously.

It’s like in Cyberpunk (sorry, we’re ex-CD Projekt guys). First, you’re selecting your class (sort of):

Then you’re selecting your auguments, which are skills needed to do the job end-to-end.

What does this look like in practice?

A PM who prototypes in Cursor instead of waiting for engineering.

An engineer who runs their own user interviews instead of waiting for research.

A designer who ships working code instead of handing off mockups.

They don’t wait for “the right person” to unlock the next step. They close the gaps themselves.

Small teams. Massive impact.

The market is already moving

LinkedIn just introduced a formal “Full Stack Builder” title. They want their entire product, design, and engineering org to move towards that model.

Figma’s 2025 research shows 72% of people say AI is driving role expansion, with 56% of non-designers now doing design work. Dylan Field puts it bluntly: “We’re all product builders, and some of us are specialized in our particular area.”

Linear built a $1.25B company with one PM. One. The rest? Engineers and designers who think about product and business, not specialists waiting for instructions.

The future doesn’t belong to people who can prompt AI to generate stuff. It belongs to product builders who know what good looks like. People who can judge what AI produces. People who know when the output is shit and when it’s gold.

That judgment doesn’t come from AI. It comes from experience. From pattern recognition. From building things that worked and things that didn’t.

From knowing the formula.

We’ve spent years applying this formula across companies. Teaching people the missing parts. Within 6-12 months, they’re doing the Batman thing. Solving impossible problems. Getting pulled into bigger rooms. Having other companies try to poach them.

The pattern held everywhere. Different companies, different markets, different team sizes. It wasn’t luck. The formula works.

Now we’re sharing it with the world. That’s why we built Superhero.tech.

What’s Superhero.tech?

Superhero.tech is the consulting and education platform we wished existed when we were in the messy middle of our careers.

It’s a newsletter, podcast, webinars, courses, and a community. Think Batman’s training program for creating more Batmans and more Robins. Or Avengers Academy if you’re more into Marvel.

We teach the Superhero Formula - the full product development lifecycle for the AI era.

We teach product people - PMs, designers, engineers, leaders, researchers, data people, founders - how to become product builders.

How to build products customers actually love and teams you’ll be proud of. When to use AI, and when to use your own judgment.

We teach from our successes and failures, not from books:

• 32+ years combined experience building global products at Booking, CD Projekt, Displate, GOG.com

• 42,000+ people trained through workshops and programs

• NPS 80 on our last cohort - 92% would recommend it to a colleague

Practice over theory. No AI slop.

Join Us

AI Product Heroes 2.0 is our flagship cohort program (currently for Polish speakers only). The training ground for Product Builders. You’ll learn and apply the complete formula with other like-minded makers.

Join AI Product Heroes 2.0

Not ready for the cohort? Subscribe to the newsletter. We’ll share the formula piece by piece - what works, what doesn’t, and why most product advice is garbage.

PS: Want to work with us on your AI and product muscles? Reach us at superhero@superhero.tech.