A new software engineering paradigm – Blog

kept by eddie jul 22

Using formal verification, AI agents can generate code and machine‑checked proofs, removing the need for human code review and overcoming the AI‑generated‑code bottleneck.

Karpathy's Autoresearch found a 3-year-old bug in our query engine (and improved performance by 11%) - PostHog
PostHog used a Karpathy-style AI agent to find a 3-year-old ClickHouse bug where `toTimeZone()` disabled primary key usage, and the fix cut query time by 37%
also on autoresearch, Andrej Karpathy, #ai-coding, #software-engineering, #developer-tools
'The Karpathy Loop': 700 experiments, 2 days, and a glimpse of where AI is heading
Andrej Karpathy's 'autoresearch' AI agent ran 700 experiments in 2 days, discovering 20 optimizations that sped up language model training by 11%.
also on autoresearch, Andrej Karpathy, #ai-coding
A Type System Is a Search Oracle
Stronger type systems like Rust and Lean improve AI-generated code quality by providing a verifiable compiler feedback loop that catches errors.
also on Lean, #rust, #ai-coding, #software-engineering, #developer-tools
Test, don't just verify
Testing is argued to be more practical than formal verification, but commenters are divided on AI’s role in making verification mainstream.
also on Lean, #rust, #ai-coding, #software-engineering, #developer-tools
LLM Wiki v2 — extending Karpathy's LLM Wiki pattern with lessons from building agentmemory
Extends Karpathy's LLM Wiki pattern with memory lifecycle, knowledge graphs, hybrid search, and automated quality controls to keep personal wikis healthy as they scale.
also on Andrej Karpathy, #ai-coding, #software-engineering, #developer-tools