Do AI-era product teams still need dedicated product managers?

Luke Harries · May 23, 2025

If you hire great engineers, they don't need PM support — and you can screen for that by testing engineers on the full product loop from feature selection to wireframes to system architecture

An engineer screened for strong coding and then tested through the full product loop can be trusted to take features from idea to launch

51:39 20VC: ElevenLabs Head of Growth on Why You Do Not Need PMs | The 7-Part Launch Playbook That Gets 700K+ Views Per Product | The Truth About CAC, Payback & Performance Marketing in AI with Luke Harries

Jason James · Aug 29, 2025

Product, design and engineering roles are not fundamentally distinct — design is a set of decisions about a product that an engineer, designer, or PM can make; roles exist only to fill skill gaps on a team

Historically UX and then product emerged as separate fields only because some engineers couldn't design and someone had to own business requirements, but some engineers and designers could always do both; his own move from design to product showed the work isn't really different

7:20 20Product: Why Most CPOs are Bad | Why You Do Not Need PMs in a World of AI | Why the Design Stage is Dead and How to Use Vibe Coding to Replace It | The Three Roles All Founders End Up Firing on Repeat with Jason James @ Tezi

Zach Lloyd · Oct 17, 2025

Product roles will not collapse into a single superhuman builder; if one role were to dominate it would be the product-minded senior engineer

PMs and designers can make cool prototypes but there's a real cap on what they can build with these tools, and getting to production requires someone more sophisticated than the agent

Scope: caps described as 'right now'

20:05 20VC: The Startup Adding $1M ARR Every Week | Competing Against OpenAI's Codex and Claude Code: Who Wins | Why Gemini is Failing and GPT-5 Is Winning | Do Margins Matter in a World of AI | The Ugly Truth About AI Coding with Zach Lloyd, Warp

Alexander Embiricos · Feb 21, 2026

You don't need many product managers until a team is really large, because everything a PM does can be done by a strong engineering lead or a product-minded designer

The PM role is explicitly undefined — it's adapting to whatever the team needs (looking around corners, working with go-to-market, being cheerleader and quality bar), and all of those functions can be absorbed by a strong eng lead or designer

Scope: it is often still useful to have product managers; framed partly as his own joke about his role

7:21 20VC: Codex vs Claude Code vs Cursor: Who Wins, Who Loses | Will All Coding Be Automated - Do We Need PMs | The Real Bottleneck to AGI | The Three Phases of Agents and What You Need to Know with Alex Embiricos, Head of Codex at OpenAI

Luke Harries · May 23, 2025

Instead of separate engineering, product and marketing functions, companies should merge product with marketing into a 'growth' function, pairing an engineering lead per product with a growth lead

The engineering lead owns overall product quality while the growth lead owns awareness and acquisition, and they collaborate on activation and retention

Scope: ElevenLabs' growth team is mostly ex-PMs

49:59 20VC: ElevenLabs Head of Growth on Why You Do Not Need PMs | The 7-Part Launch Playbook That Gets 700K+ Views Per Product | The Truth About CAC, Payback & Performance Marketing in AI with Luke Harries

Luke Harries · May 23, 2025

PMs as a distinct role will shrink, with PMs moving either into growth (PM combined with marketing) or into product engineering

PMs are often somewhat technical already and can upskill with tools like Cursor or Lovable to ship end-to-end products themselves

Scope: how much PM work remains depends on company size and engineer strength

52:56 20VC: ElevenLabs Head of Growth on Why You Do Not Need PMs | The 7-Part Launch Playbook That Gets 700K+ Views Per Product | The Truth About CAC, Payback & Performance Marketing in AI with Luke Harries

Varun Mohan · Jun 2, 2025

PMs will still have a role but the expectation shifts from specifying work to building it themselves; technical PMs who can code become far more powerful because they can prove an idea without a team of ten

AI tools give PMs the capability to build, so spending time writing docs to persuade the organization and playing Game of Thrones politics makes no sense when you can just build a version and prove the point

Scope: requires much more agency from PMs

33:54 20VC: Windsurf Founder on Will Model Companies Own the App Layer | Why Moats Do Not Exist in a World of AI | Why the Notion of Single Person $BN Companies is BS | Lovable vs Bolt & Cursor vs Windsurf: How Does it All End with Varun Mohan

Cem Kansu · Jun 20, 2025

PMs will have to become far more independent, building prototypes themselves with AI tools instead of relying on other functions to visualize an idea

AI tools will replace the human-heavy machinery currently needed just to visualize and prototype an idea

71:19 20Product: How Duolingo Build Product 10x Faster with AI | Duolingo's Biggest Lessons on Paywalls, Push Notifications and In-App Purchases | Why Small Teams are the Future of Product | Why PMs Will Become Extinct with Cem Kansu, CPO @ Duolingo

Jason James · Aug 29, 2025

PM should be a hire of last resort rather than a default early hire

Leaving room for engineers and designers to lean into the business creates ownership and deeper employee engagement and makes them think like founders, versus 'stay in your lane' specialization

9:38 Hire pm only as a last resort once execution suffers

Jason James · Aug 29, 2025 · hedged

PMs' time will shift away from long PRDs toward gap-filling, context-building and prioritization, with output compressed into one-pagers instead of twenty-pagers

Teams deep in building can't zoom out to choose the three things to do next among fifty, or trade off current customers vs. desired customers vs. fixing what's built vs. new capabilities — that requires context that takes time to build

10:35 Pm role shifts to context building and prioritization not long prds

Mike Krieger · Mar 3, 2025

Driving alignment on what to build — product strategy and solving real user problems — is the remaining bottleneck in product development and is at least three years from being solved by models

Claude already helps with PRDs, coding, and synthesizing product conversations, but deciding what to build is still best resolved by people in a room or exploring in Figma; in any dynamic system, optimizing one step makes another the critical path

47:54 Deciding what to build remains unsolved by models for years

Cem Kansu · Jun 20, 2025

AI tools level up product managers in particular, relieving the imposter syndrome of being neither technical enough for engineers nor creative enough for designers

The tools put everyone on a more similar footing with strong capabilities, making PMs more productive and independent

25:05 Ai tools level up pms specifically closing the technical creative gap

Cem Kansu · Jun 20, 2025 · hedged

The PM job will look entirely different within twelve to twenty-four months, not the five-to-ten years he previously believed

He previously assumed AI adoption in product would be gradual; he now thinks people say 'five years' because it feels safe

72:11 Pm role change is imminent within two years not five to ten

Alexander Embiricos · Feb 21, 2026

Product managers are still necessary, but they must be a perfect fit — a non-fit product person does more harm than good

The bar is such that a poor-fit PM is net negative, which forces far more selective hiring than in other roles

49:31 Pms are necessary but only perfect fit ones add value