Podcast Summary: Lenny’s Podcast: Product | Career | Growth

Episode: Getting paid to vibe code: Inside the new AI-era job
Guest: Lazar Jovanovitch (Professional Vibe Coder, Lovable)
Host: Lenny Rachitsky
Date: February 8, 2026

Overview

This episode explores the groundbreaking new role of “professional vibe coding” with Lazar Jovanovitch, Lovable’s first official Vibe Coding Engineer. Lenny and Lazar dive deep into what vibe coding is, why it’s fundamentally changing the nature of product and software development, and how anyone—even those with no technical background—can leverage AI tools to become high-impact builders. Throughout the conversation, Lazar shares actionable frameworks, mind-expanding anecdotes, and candid insights on excelling in an AI-first product landscape.

Key Discussion Points & Insights

1. What Is Vibe Coding? (05:00–06:43)

• Definition: Vibe coding is building products and tools by orchestrating AI agents, rather than writing code directly.
• Breadth: The role covers a huge surface area—internal tools, external products, integrations, and rapid prototyping.
• Quote:
  • “It’s an ideas role. A lot of people have great ideas, but they don’t know how to build them... That’s where I step in.” – Lazar (05:47)

2. From Non-Technical Background to Elite Builder (09:51–12:02)

• Lazar never wrote code before vibe coding—just “console logs.”
• Counter-intuitive Advantage:
  • Not knowing “what can’t be done” allows for more creative, bias-free problem-solving with AI tools.
• Delusional Optimism:
  • “We're just coming to this completely unbiased and very positively delusional, which I think you have to have when working with AI tools...” – Lazar (10:20)

3. Success with AI Tools: Clarity is the Ultimate Skill (12:36–18:07)

• The Core Problem: Not “how to code” but “how to be clear with AI.”
• Planning vs. Executing:
  • 80% of Lazar’s workflow is spent “in planning and chatting”—only 20% in execution.
• AI as Genie Analogy:
  • “I want to be taller.” The genie makes you 13ft tall—because you weren’t specific enough. This is how AI interprets unclear instructions.
• Actionable Skill:
  • The most important skill: learning how to ask AI the right questions and requests (“clarity in the ask”).

4. Investment in Taste, Judgment, and Exposure (18:07–22:21)

• Being “World-Class” now means moving far beyond “good enough”; AI will produce “good enough” by default for everyone.
• Judgment and Taste:
  • Developing good judgment comes from exposure to great products, relationships, and content.
  • “We will be rewarded in the world of AI for better judgment, not for faster raw output.” – Lazar (19:56)

5. Lazar’s Clarity Framework (“Parallel Vibe” Approach) (22:21–30:26)

• Run Multiple Attempts Simultaneously:
  1. Start with a vague idea—“brain dump” prompt.
  2. Iterate with more detail and references (screenshots, design inspiration).
  3. Find and attach relevant code snippets or templates.
  4. Compare 4–5 parallel builds to quickly clarify and choose the winner.
• Quote:
  • “I never built just one project at a time. I built five or six. I have six Lovable tabs and I just switch between them.” – Lazar (29:02)

6. Managing Context and Scaling with PRDs (30:26–46:32)

• Problem: LLM tools can’t remember everything; their “token window” is a resource constraint.
• Solution:
  • Develop clear, evolving project documentation (PRDs) that the agent can refer to, such as:
    1. Masterplan.md: High-level intent and who it’s for (44:43, 45:12)
    2. Implementation_Plan.md: Proposed sequencing/orders of building.
    3. Design_Guidelines.md: Specs on visual style and UX.
    4. User_Journey.md: Core flows for users.
    5. Tasks.md: Task breakdown for the agent.
    6. Rules.md or Agents.md: Behavioral guidelines for the AI agent (project settings).
• Execution:
  • Once documentation is solid, switch to “proceed with next task” prompts, letting the agent autonomously manage building, while you focus on review and judgment.
• Quote:
  • “The ceiling on the AI isn’t the model intelligence, it’s what the model sees before it acts.” – Lazar (43:59)

7. Debugging & Learning: Unblocking Yourself Framework (65:36–75:29)

• The 4x4 Debugging Flow:
  1. Ask the Agent to Fix: Use built-in “try to fix” buttons.
  2. Add Debugging & Console Logs: Guide AI to add logs; bring awareness to the problem.
  3. External Code Reviews: Use Codex, Claude, or ChatGPT with your codebase and logs for deeper diagnostics.
  4. Revert & Re-prompt: Use version control; reissue prompt after reflection or break.
• Learning:
  • After solving, ask the agent how you could have prompted better—then encode the learning into the project rules for next time.

8. The Emerging Skill Set: PMs and Designers Ascendant (55:25–64:54)

• Product Management Skills Dominate:
  • Clarity, defining “what to build,” and requirements are now the differentiator.
• Design:
  • The next “winner” skill: emotional intelligence and exquisite design and UX sense.
  • Learn modern design trends, terminology, and how to prompt for them.
    • “Learn what good design means. Learn how to prompt to get them is probably what I would optimize for at this stage.” – Lazar (59:36)
• Engineering:
  • Elite, infrastructure-focused engineers will always be needed for scaling and maintenance.
  • Writing original code is becoming like calligraphy—rare and valuable, but no longer the norm.
  • “Coding is going to be like calligraphy... That’s so amazing. It’s going to be so rare that it becomes an art.” – Lazar (63:27)

9. Evolution of Roles and Advice for the Future (82:52–87:57)

• Key Survival Skills:
  • Emotional intelligence, design, exceptional copywriting, and “human to human” abilities will future-proof your career.
  • Middle-management, routine translation, and deterministic “math” work will be most susceptible to AI automation.
• Distribution Skills:
  • In a saturated world of builders, getting attention and traction for your apps is more scarce than building itself.

10. How to Become a Professional Vibe Coder (88:49–94:18)

• Lazar’s Path:
  • Non-linear—blue collar jobs, social media, community management, then startups.
  • Became visible by building in public, sharing failures and learnings (especially on LinkedIn and YouTube).
  • “Build in public. Share your knowledge. There are no secrets whatsoever. If you’re sitting on a good concept, you’re missing out. Just share it immediately.” – Lazar (90:00)
• Actionable Advice:
  • Hackathons, posting real apps as job applications (not just resumes), direct outreach with demos.

Notable Quotes

• On Clarity:
  • “I’m optimizing 100% of my time today on good judgment, clarity, quality, taste, good fonts...” – Lazar (36:55)
• On Role Convergence:
  • “These Venn diagrams of engineer, designer, PM—they used to be very separate. Now they’re converging.” – Lenny (64:54)
• On AI as Amplifier:
  • “AI, regardless of your background, is an amplifier. If you don’t know what you’re doing, you’re just going to produce garbage faster.” – Lazar (09:29, 57:03)
• On Emotional Intelligence:
  • “Anything human to human is going to be a big thing to skill up on. Understand the dynamics...” – Lazar (83:33)
• On the Power of Building:
  • “Go in, build something. Build anything. And the fear goes away. You should only be afraid if you’re doing nothing.” – Lazar (96:45)

Timestamps for Important Segments

• [05:00] What is vibe coding? The role explained
• [09:51] Why a non-technical background can be a strength
• [12:36] The #1 skill: clarity in AI prompting
• [22:21] The “parallel builds” framework for idea refinement
• [30:26] Keeping context with PRDs and project structure
• [44:43] Lazar’s template for planning files (Masterplan.md, etc.)
• [65:36] Lazar’s 4x4 debugging and unblocking framework
• [82:52] The most valuable future skills in an AI-driven world
• [88:49] How Lazar entered the profession and how to follow
• [94:38] Switching from fear to excitement in the new builder era

Memorable Moments

• The Genie Analogy: For every user who prompts AI with “I want to be taller,” get ready to be 13ft tall if you’re not specific. (09:41, 12:36)
• Running 5 versions in parallel: Unconventional but effective—use the AI as “cheap labor” to clarify and explore directions rapidly. (22:21–30:26)
• “Vibe coding” will soon just be coding: The rapid absorption of “workarounds” into mainstream tools (79:32)
• Coding as calligraphy: Writing raw code will become rarified, bordering on art. (63:27)
• Build in public: Every major opportunity came from sharing visibly and generously. (90:00)

Actionable Frameworks

• Clarity Framework:
  • Start with a brain dump; iterate with more detail; provide references and code; build 4–5 options; select and plan with PRDs.
• Project Structure:
  • Masterplan.md → Implementation_Plan.md → Design_Guidelines.md → User_Journey.md → Tasks.md → Rules.md
• Debugging (4x4):
  1. Ask agent “try to fix;”
  2. Add debugging logs and rerun;
  3. External code review with Codex/Claude;
  4. Revert/re-prompt, then encode learnings in project docs.

Resources & Next Steps

• Lazar’s GPT Tools: Search “Lovable PRD generator” in ChatGPT GPT Store.
• Connect:
  • LinkedIn (preferred) / YouTube for Lazar’s content.
• Lovable Hiring: Open positions on their website for like-minded aspiring vibe coders.
• Final Advice:
  • “Set aside more time for learning than building. Read agent output, learn how it’s thinking. Exposure time—to great design, to great copy, to what magic looks like—is the best investment you can make.” – Lazar (97:15)

For those inspired, stop reading, start building. “Just go.”