To vibe code or not to vibe code: Are product managers strategic builders or just distracted?
AI has shattered the execution barrier for product managers. With the rise of "vibe coding" – using natural language and LLMs to generate functional software without deep manual coding – product managers now have the power to build what they used to only describe. While this shift promises a new era of "maker" productivity, it also threatens to transform product teams into the ultimate organizational bottleneck.
If engineering teams are shipping 10x faster thanks to AI, but product managers are busy "nitty-gritty" fine-tuning code instead of validating problems, does the entire system stall? Here, we hear the arguments for both sides, from:
Malte Scholz, Head of Product and co-founder at airfocus by Lucid
Teresa Torres, Product Discovery Coach at Product Talk
Chris Butler, Director of Product Operations at GitHub
Lauren Kearney, Director of Health & Platform Products at Twitch
What is vibe coding in product management?
Vibe coding for product managers means using natural language and LLMs to generate functional software without deep manual coding. It breaks the traditional execution barrier, allowing PMs to move beyond just describing requirements to actually building low-cost prototypes and exploring technical "plumbing" themselves. This enables product managers to become "makers" who can rapidly translate a high-level insight into a tangible, testable solution in a matter of hours.
can support your team.
Book a demo
1. The pros of vibe coding for product managers
Some argue that when a product manager becomes a builder, the distance between "insight" and "solution" shrinks. This is not about replacing engineers, but about augmenting the product manager’s ability to communicate and de-risk.
Building technical empathy
“I'm just blown away by how accessible building has become," says Teresa Torres. She views diving into the "plumbing" of AI products as more than just a hobby, as a form of professional development. By experimenting with tools including Claude Code, product people can see exactly where a model gets stuck or how a context window fills up. “This has been the best way for me to build my builder toolkit,” she concludes.
Malte Scholz also appreciates the “maker” component of a product manager’s role: "I like product managers going out there creating prototypes... in order to understand what it really takes to build AI products.”
Faster conceptual de-risking
In the past, a prototype could take weeks. Today, it can take hours, allowing for a much tighter loop between an idea and a tangible discussion point. In Malte’s words, “You can test solutions, ideas, and experiments much faster with these new prototyping tools.”
"Product managers should be using prototyping for the early conceptual parts of the world. It’s helpful to have a discussion with a cross-functional team," agrees Chris Butler.
Teresa uses her own experience building with AI as an example: “I started prototyping in a Claude project [...] I was just blown away by how much progress I was able to make just in a few hours". Her goal wasn’t to immediately build a product then, simply “to see how good this thing can be.”
2. The cons of vibe coding for product managers
The danger of vibe coding is how fun it can be. It feels productive to ship code, but for a product manager, is this a distraction from the “real work” of decision-making, strategic alignment and customer discovery?
Avoiding “uncomfortable” work
Strategic thinking is often "tapping in the dark" and can be deeply uncomfortable. Tangible tasks that provide immediate gratification, like tweaking a UI or fine-tuning a prompt, always seem more tempting.
Malte warns against this trap: "People tend to pick work that makes them feel busy and productive, and to avoid the uncomfortable thinking work. I think it's just built into humans. They love to spin up code and fine-tune all the nitty-gritty stuff. But they have not thought heavily about what problem they are actually solving.”
Teams need to know what to build
Malte is concerned that product managers who are vibe coding are neglecting the "problem space”, which is the only thing that justifies the engineering team's increased velocity. ”Development gets faster, and product management becomes the bottleneck,” he says. "This is critical time that they won't spend on the problem space, the discovery, the critical thinking, the alignment work they have to do. If product management fails to move at an accelerated pace... they still need help knowing what to build.”
The expectation for vibe coding in product management
A vibe-coded prototype is rarely production-ready. Chris Butler warns that relying on these prototypes for actual delivery is the perfect recipe for technical debt: "This leadership expectation that product managers will vibe code something and then it will go into production and be scaled to a million users... that's the biggest way that every vibe-coded startup fails.”
3. The middle ground: Prototyping for insight, not execution
The most effective product managers in 2026 use vibe coding as a tool, not a primary job function. They recognize that their unique value lies in human judgment, storytelling, and prioritization – areas where AI still falters.
Defining the "new" technical product manager
Being a "technical product manager" doesn’t mean writing Python, it means understanding systems well enough to sit in a room of engineers and notice when something feels off.
"A product manager who is vibe coding is not a technical product manager,” reflects Lauren Kearney. “What makes a technical product manager is that they actually understand the systems and the technologies of their products."
Teresa Torres agrees. "I don't think you have to learn how to write Python or Node.js, but I think we need to be proficient in things like how LLMs work." Product teams need to get acquainted with the “plumbing” of AI:
LLMs: “Learning how transformer models work and why they’re different from past machine learning.”
APIs: “Getting exposed to APIs and learning how to use them; how to design them for different contexts.”
Context management: “Understanding the context window and how to manage it, and how to provide just enough information at the right time.”
Upskilling is urgent, but it’s also easier than ever: “Now that we live in a world where literally an LLM can walk you through it while you do it, all this stuff is very learnable,” Teresa notes.
Guarding the quality standards
In a world of infinite AI-generated ideas, product teams’ role shifts from "creator" to "curator" and "judge" of quality.
"Quality is going to be a really important piece of what a product manager does," says Lauren. “The question for a product manager for 2026 is really, ‘What are the guardrails for quality and can I contribute to those?’"
That’s why evals are so important moving forward. “We can put evals in place as guardrails to tell our models, ‘You cannot show this to a customer if it doesn't match these criteria,'” says Teresa.
It’s critical that we don’t lower our standards, as she sometimes notices. “I see people write on LinkedIn, ‘How are we supposed to train our customers to expect mediocre responses from LLMs?’ That’s not your job,” she says. “Your job is to get good responses from the LLM.”
Teresa is vocal about her quality standards, and she lives by them when building with AI. “When I started to build the Interview Coach, I wasn’t going to compromise on that,” she says. “I wasn’t going to release it if I couldn’t get it to a good place.”
Customers remain at the core of our products
No matter how fast a product manager can code, they must not forget that their primary goal is to build the right products.
For those concerned that they’re losing their jobs to AI, Lauren has some reassurance and advice: “When in doubt, be the voice of the customer.” Technology will never have the human experience, and it is the human population that is using the products.”
Teresa supports this: “The better we know our customers, the more likely we're going to differentiate from our competitors in the market.”
Vibe coding for product managers: The dos and don’ts
| Strategic practices | Distractions |
| Building a "throwaway" prototype to explain a complex user flow to your team. | Spending 6 hours fine-tuning a CSS layout instead of doing a customer interview. |
| Tinkering with an IDE to understand the limits of your product's API. | Believing your "vibe code" is ready for a million-user production environment. |
| Using AI to "challenge" your strategy by presenting "doomsday scenarios". | Micromanaging engineers by getting caught in the minutia of their work. |
The verdict for 2026? The product team’s role is to ensure the team isn't just shipping fast, but shipping right. And if vibe coding helps you get there, then it’s the right tool in your toolbox.
Francisca Berger Cabral
Read also
Experience the new way of doing product management
Experience the new way of doing product management