Writing a Vibe Coding Synthesis Paper: From AI Prompt to Polished Argument

In early 2023, Andrej Karpathy — a co-founder of OpenAI and former AI lead at Tesla — coined the phrase 'vibe coding' to describe a new way of building software: describe what you want to an AI in plain language, iterate on what it produces, and ship without writing every line yourself. A synthesis paper about vibe coding is not a summary of what it is. It is an argument: What has this shift meant for who gets to build software? For code quality? For the future of programming as a profession? This lesson teaches you how to build that argument.

By the end of this lesson you will be able to: 1. Explain what distinguishes a synthesis paper from a summary or a research report. 2. Identify credible sources for a paper on vibe coding and AI-assisted development. 3. Construct a thesis statement that makes a defensible claim. 4. Outline a five-section synthesis paper and draft a complete introduction.

A synthesis paper does three things a summary does not: 1. **Draws on multiple sources** and shows how they connect, contradict, or build on each other. 2. **Makes an original argument** — not just 'here is what I found' but 'here is what it means.' 3. **Evaluates evidence** — not all sources are equal; a synthesis paper weighs them and takes a position. For a vibe coding synthesis you might bring together: - Karpathy's original posts and interviews about AI-assisted development. - GitHub's Copilot research showing a 55% speed increase on certain coding tasks. - Security research finding that AI-generated code contains vulnerabilities at higher rates than human-written code. - Industry case studies on what ships faster versus what still requires deep engineering expertise. Your job is to read across those sources and arrive at a position that none of them alone spells out — that is your contribution as author.

Weak thesis: 'Vibe coding is a new way to write software using AI.' Why it is weak: It describes. Any reader responds: 'OK, and?' Stronger thesis: 'While vibe coding dramatically lowers the barrier to software creation, it also introduces a structural risk — builders who cannot read their own code output cannot detect its failures, creating a generation of products that work until they suddenly do not.' This stronger thesis: - Makes a specific **claim** (lowers barriers — a real benefit) - **Concedes** the counter-position (this is genuinely valuable) - **Pivots** to a deeper original argument (a structural risk others overlook) - Is **falsifiable** — someone could argue against it with real evidence Your thesis for this paper should do all four of these things.

**Section 1 — Introduction (1-2 paragraphs):** Hook + context + thesis. End with a road map: 'This paper will first examine X, then Y, and ultimately argue Z.' **Section 2 — Background (2-3 paragraphs):** What is vibe coding? When and how did it emerge? Who is using it and why? Synthesize 2-3 sources — do not summarize them one by one. **Section 3 — The Case For (2-3 paragraphs):** Present the strongest evidence for one major benefit — faster prototyping, lower barriers, democratization of software. Use real data: GitHub Copilot studies, Stack Overflow developer surveys. **Section 4 — Complications and Counter-Evidence (2-3 paragraphs):** Present the strongest counter-evidence — security vulnerabilities, debugging opacity, technical debt. Use data here too. **Section 5 — Synthesis and Conclusion (2-3 paragraphs):** Do not merely say both sides have merit. Take a stand. Show how your sources, read together, lead to your specific conclusion. End with implications: what should change?