Practical AI Podcast: Agentic Coding and the Economics of Open Source

Date: April 2, 2026
Host: Chris Benson (Principal AI Research Engineer)
Guest: Dr. Miklós Koren (Professor of Economics, Central European University, Vienna)

Episode Overview

This episode dives into the intersection of artificial intelligence, agentic coding ("vibe coding"), and the future of open source software from an economic perspective. Dr. Miklós Koren discusses his provocative research, "Vibe Coding Kills Open Source," exploring how AI-driven software development reshapes incentives, attention, and sustainability within the open source ecosystem. The conversation illuminates both empirical research and broader implications for developers, companies, and the labor market.

Key Discussion Points and Insights

1. Origins: Economics and Open Source in the Age of AI

• Dr. Koren’s Background (02:04):
  • Interest in competitiveness—of companies and countries.
  • Accidental software developer, blending economics and computational research.
  • Initial research into localization of open source collaborations and why developers contribute.
• The "Vibe Coding" Trigger:
  • Observed explosion of AI-generated, personalized apps on social media (late 2025).
  • Realization: "Why download someone else's app when I can make my own, exactly as I want, with AI tools?" (04:49)
  • Connection: Personalized, on-demand app generation by AI has fundamental economic implications for open source.

2. "Vibe Coding Kills Open Source": What Does It Mean?

• Provocative Title & Its Rationale (06:06):
  • Not just clickbait: "It’s not clickbait if it’s true." (07:38) C
  • AI lowers friction and cost for producing software and open source packages.
  • Human attention—especially to open source projects and contributors—becomes a scarce resource.
  • Main economic pillars discussed:
    • People respond to incentives (monetary, reputation, satisfaction).
    • The economy is a closed system (scarcity and tradeoffs).
    • Scarce resources increase in value.

Open Source Incentives in Flux

• AI reduces the developer cost, increasing potential open source contributions, but...
• Attention Scarcity: Human attention doesn’t scale with code volume, making it harder for individual projects to get noticed.
• Open Source Success: "The more users they have, the better... For open source you have to have like millions of users to be successful because kind of the margins are so thin." (17:19-17:57) C

3. Empirical Evidence: AI’s Impact on Package Usage & Human Interaction

• Case Study: Tailwind CSS (21:22):

  • AI models (like Claude Code) increasingly recommend certain libraries (e.g., Tailwind).
  • Downloads surge (driven by AI agents); human engagement (e.g., GitHub stars, website visits) decline.
  • Implication: Business models relying on human web visits (e.g., upsells to premium, community support) are threatened.

• Controlled Experiment (22:00-26:01):

  • Researchers gave 100 popular website requirements (without any tech specifics) to various AI models.
  • AIs chose dependencies and libraries autonomously; data captured both NPM download stats and GitHub stars.
  • Findings: "Packages that become very, very vibe coding friendly... you divert attention from humans towards the machines. And so the machines are downloading, but the machines are not interacting with the developer on GitHub." (25:45) C

4. Should We Even Need Libraries? The "Future of Build vs. Reuse" Question

• Host’s Challenge: With powerful models, why use existing libraries at all? Couldn’t AI just build everything from scratch each time? (26:01-28:09) B
• Dr. Koren’s Perspective:
  • AI as a "fast, very capable coworker," not just a tool.
  • Human-friendly structures, function names, and abstractions help align AI outputs to developer intent (28:09-32:30) C.
  • Good coding practices are now more important for collaborating with AI models.
  • Rise of "throwaway code" and iterative, on-the-fly problem solving: "You start with a very concrete implementation and then you kind of generalize from there, which is very new." (32:46) C
  • Most open source remains essential for "core" tools; as agentic automation rises, fewer but higher-quality, well-maintained libraries may persist.
  • Open Source Health Warning: "If we froze everything today... and nobody contributed to open source anymore, it wouldn’t mean we are staying stuck, it would mean that it declined." (36:05) C

5. Labor Market & Educational Implications

• On Organizational Change (36:35-44:25):

  • Rapid shifts: layoffs, shift to senior-only engineering staff.
  • The "job" of software engineering splits into:
    • Understanding and translating user needs.
    • Designing systems and components.
    • Writing actual code (the last is close to fully automated).
  • Still vital: "The thinking part you cannot really get rid of... their primary job is to think." (41:29) C
  • Challenge for education: How to teach "computational thinking" when the new programming language is English?

• Comparative Advantage and Human Role:

  • "Even if AI can do everything better, we can kind of exploit [our] comparative advantage... the thinking part would always be a human comparative advantage." (42:55) C
  • AI amplifies knowledge worker productivity but shouldn’t wholly replace the creative, design, and critical thinking components.

6. Speculation & Vision: The Road Ahead

• Dr. Koren’s Forecast (45:18-47:39):
  • Cheap, locally available intelligence may counteract centralized platform power (Google/Facebook), greatly impacting digital economy structures.
  • "When everybody has kind of knowledge locally available to them... it could completely rewrite what we now understand about software, digital economy, knowledge industries, basically the entire economy and society." (47:15) C

Notable Quotes & Memorable Moments

• On the Paper’s Title:
  "It’s not clickbait if it’s true."
  — Dr. Miklós Koren (07:38)

• On AI & Open Source Incentive:
  "Human attention... is a very limited resource. So if you turn this towards AI, you have to take it away from something else."
  — Dr. Miklós Koren (09:20)

• On Package Popularity & Human Interaction:
  "For packages that kind of become very, very vibe coding friendly... the machines are downloading, but the machines are not interacting with the developer on GitHub."
  — Dr. Miklós Koren (25:45)

• On Changing Engineering Roles:
  "I think the thinking part you cannot really get rid of, and I think in particular the interaction with the users and figuring out what they really need."
  — Dr. Miklós Koren (41:29)

• On the Local AI Revolution:
  "Intelligence becoming very, very cheap and even locally reproducible is a force that goes the other way. And that could be just completely rewriting what we now understand about software, digital economy, knowledge industries, basically the entire economy and society."
  — Dr. Miklós Koren (47:15)

Timestamps for Key Segments

• [02:04] – Dr. Koren’s background and origin of research theme
• [06:06] – Meaning and ambition of "Vibe Coding Kills Open Source"
• [15:49] – Human attention as the limiting factor in the new paradigm
• [21:22] – Tailwind case study; data-driven analysis and findings
• [26:01] – The future relevance of libraries vs. AI "from scratch" construction
• [36:35] – Organizational and labor market impacts; the enduring value of "thinking"
• [45:18] – Speculative vision: decentralization and the changing digital landscape

Summary Takeaways

• AI-driven agentic coding reduces barrier to software creation, threatening traditional open source attention/reward systems.
• Primary value for developers shifts toward design, user-needs analysis, and critical thinking. Code generation is increasingly commoditized.
• Open source success will depend less on code volume and more on quality, maintenance, and human engagement.
• Broader social and economic arrangements—from education to digital platforms—are poised for significant realignment as AI’s role expands.

For more ongoing research and insights, listeners are encouraged to connect with the Practical AI Podcast team and follow Dr. Miklós Koren’s work.