Building on Ethereum
summary
Frames Ethereum as a coordination layer, uses Uniswap and overcollateralized lending to show that liquidation mechanics only work with explicit incentive design so anyone can hit the button and get paid, cites a concrete AI workflow: one model plans, a second critiques, then build a prototype and verify every result, and recommends Speedrun Ethereum and ETH Skills to learn by doing. Demonstrates Slop Computer, an on‑chain podcast on IPFS with multisigs that actually publishes via transactions, notes 50 interviews and a November 2025 AI kickoff.
AI-written summary · about
chapters (24)
- 0:00Introduction and Builder Mindset
- 2:10Ethereum as a Coordination Layer
- 3:20The Protocol Walk-Away Test
- 4:38Learning With Speedrun Ethereum
- 6:12Understanding Concepts Before Coding
- 7:05Liquidations and Incentive Design
- 8:09Why Liquidation Mechanics Matter
- 9:24Markets, Stablecoins, and ZK
- 11:08Using AI After Learning the Concepts
- 12:16Planning an AI-Built Prototype
- 12:56Having a Second Model Critique
- 14:57Rapid Prototypes and Venture Funding
- 15:37The Learn, Build, and Verify Loop
- 16:41Slop Computer Onchain Podcast
- 18:07Extracting Lessons From Experiments
- 19:18The New Solo Founder Opportunity
- 21:20Avoiding AI Flattery and Overconfidence
- 22:21Why Two LLMs Improve Critique
- 24:05Making Space for Human Thinking
- 27:23Avoiding Output-Driven AI Traps
- 28:22Spend Most Time on Real Utility
- 30:16Training Data and Verification
- 30:27Verification as the New Bottleneck
- 31:06Solve Problems for Real Businesses
related talks
53mBuilding on Ethereum in the AI Era
01h 00mBuilding on Ethereum Human vs AI — Who Wins
25mFireside Chat
29mBuilding Onchain in the Age of Agency With BuidlGuidl
21mKeynote: Ethereum and AI: Synergies and Potential Applications
01h 08mBuilding apps quickly on Ethereum
23mScaling Wars: The Race to Speed Up On-Chain Infrastructure
15mBuilding AI Agents the Web3 Way