Ronny Chieng at Harvard: AI Makes Mediocre People Dumber, He Says
Abhishek Gautam9 min readQuick summary
Comedian Ronny Chieng told Harvard Class of 2026 on May 27, 2026 to resist AI shortcuts. Quote: AI will make mediocre people dumber. Substance vs shallow knowledge.
Read next
- Gemini 3.1 vs Claude Sonnet 4.6 vs GPT-5.3 Codex: Developer Benchmark Comparison March 2026Gemini 3.1 Pro, Claude Sonnet 4.6, and GPT-5.3 Codex all dropped within weeks of each other in early 2026. Here's how they actually compare on coding benchmarks, context windows, API pricing, and which model to use for what — a developer-first breakdown with real numbers.
- NVIDIA Nemotron 3 Super: 60% SWE-bench, Best Open Model for CodeNVIDIA Nemotron 3 Super hits 60.47% on SWE-bench — highest open-weight score ever. 120B total, 12B active, 1M context, 5x throughput vs GPT-OSS. Already in CodeRabbit and Greptile.
Ronny Chieng told Harvard's Class of 2026 on May 27, 2026 that "AI is just gonna end up making mediocre people dumber" — a profanity-laced Class Day counter-speech to the spring 2026 trend of commencement addresses urging graduates to master AI.
The Daily Show comedian and Emmy winner said the real fight is substance vs shallow knowledge, not humans vs robots "for at least two months."
What Did Ronny Chieng Say at Harvard?
Chieng's Tercentenary Theatre address (Harvard Class Day, one day before Commencement) went viral in early June 2026 as clips spread on YouTube, Boston Globe, and Harvard Crimson.
Core lines (paraphrased from transcript and Harvard Gazette):
- "F* AI"** (repeated) — crowd cheered.
- "The mission of your generation is to destroy AI" — delivered as comedy, then narrowed to misuse, not banning medical research AI.
- "AI is just gonna end up making mediocre people dumber" — aimed at people bragging that AI reads email, summarizes, and drafts replies they could do themselves.
- "You know who else can do that? Me." — punchline on automating low-skill office work.
- "Minimum escape velocity of intelligence" — only top talent gains real advantage; others accumulate cognitive debt (he cited a 2025 MIT study on LLM overuse).
- "People with substance versus people with shallow knowledge" — mastery vs faking it.
He praised AI for physics and medicine breakthroughs — attacking LLM shortcuts on reading, writing, and creative craft.
Why It Hit Now (June 2026 News Cycle)
Harvard President Alan M. Garber's Baccalaureate address May 26 took a measured AI tone — Chieng's May 27 speech was the sharp counterpoint.
The quote spread as:
- Graduation-season fatigue with generic "embrace AI" advice.
- Developer meme fuel — overlaps with vibe-coding backlash posts.
- Global audience — Chieng's Singapore upbringing resonates with abhs.in's APAC traffic.
This is culture/news, not a product launch — but it maps to engineering debates on Copilot dependence the same week GitHub switched to token billing (Copilot AI Credits).
What It Means for Developers
Chieng's argument rhymes with senior-engineer critiques:
| Chieng frame | Engineering parallel |
|---|---|
| Cognitive debt | Juniors who cannot debug without Copilot |
| Shallow knowledge | PRs that pass CI but fail architecture review |
| "Creating is the fun part" | Losing design skill to agent-first workflows |
| Escape velocity | Only strong baselines benefit from frontier models |
Counter-view: Token-priced tools (June 2026 Copilot change) reward skilled users who ship faster — the risk is unskilled users masking gaps.
Check your own exposure: Will AI Replace Me. For model economics: LLM API Pricing.
Key Takeaways
- May 27, 2026: Ronny Chieng Harvard Class Day speech — anti-AI-shortcut, pro-mastery
- Viral quote: "AI is just gonna end up making mediocre people dumber"
- Distinguishes research AI (medicine/physics) vs LLM email/slides laziness
- Frames battle as substance vs shallow knowledge, not humans vs AI yet
- For developers: cultural signal aligning with skills-first hiring; pair tools with fundamentals
Sources
FAQ
Frequently Asked Questions
What did Ronny Chieng say about AI at Harvard?
At Harvard Class Day on May 27, 2026, Ronny Chieng told graduates that excessive reliance on large language models creates cognitive debt and that AI will make mediocre people dumber if used to skip reading, writing, and creative work. He encouraged mastery and substance over AI shortcuts.
When was Ronny Chieng's Harvard Class Day speech?
Chieng delivered the Harvard College Class Day address on Wednesday, May 27, 2026, at Tercentenary Theatre, one day before Commencement.
Did Ronny Chieng tell Harvard students to ban all AI?
No. He said AI for breakthroughs in medicine and physics is not the problem. He attacked using AI to automate basic email and creative shortcuts, arguing that the fun and skill-building parts of work should stay human.
Why did Ronny Chieng's AI speech go viral in June 2026?
The speech contrasted with other 2026 commencement addresses urging students to master AI. His profanity, humor, and quote about mediocre people getting dumber spread through Harvard Gazette, Crimson, Boston Globe, and social video clips in early June.
Free Weekly Briefing
The AI & Dev Briefing
One honest email a week — what actually matters in AI and software engineering. No noise, no sponsored content. Read by developers across 30+ countries.
No spam. Unsubscribe anytime.
More on AI Models
All posts →Gemini 3.1 vs Claude Sonnet 4.6 vs GPT-5.3 Codex: Developer Benchmark Comparison March 2026
Gemini 3.1 Pro, Claude Sonnet 4.6, and GPT-5.3 Codex all dropped within weeks of each other in early 2026. Here's how they actually compare on coding benchmarks, context windows, API pricing, and which model to use for what — a developer-first breakdown with real numbers.
NVIDIA Nemotron 3 Super: 60% SWE-bench, Best Open Model for Code
NVIDIA Nemotron 3 Super hits 60.47% on SWE-bench — highest open-weight score ever. 120B total, 12B active, 1M context, 5x throughput vs GPT-OSS. Already in CodeRabbit and Greptile.
OpenAI GPT-5.5 Released: Agentic Coding and Multi-Step Reasoning Upgrade
OpenAI released GPT-5.5 on April 23-24 2026. Stronger agentic coding, multi-step reasoning chains. Rolling to ChatGPT Plus, Pro, Enterprise. API access coming soon.
Google Invests $40B in Anthropic: $350B Valuation, 5GW Compute Deal
Google committed $40B to Anthropic in April 2026 — $10B immediate, $30B conditional on milestones. Valuation stays $350B. 5GW compute over 5 years for Claude training.
Free Tool
Will AI replace your job?
4 questions. Get a personalised developer risk score based on your stack, role, and what you actually build day to day.
Check Your AI Risk Score →Written by
Software Engineer based in Delhi, India. Writes about AI models, semiconductor supply chains, and tech geopolitics — covering the intersection of infrastructure and global events. 952+ posts cited by ChatGPT, Perplexity, and Gemini. Read in 167 countries.