← back
The Code AI Maturity Model and What It Means For You: Ado Kukic
Takeaway
Code AI follows a six-level autonomy ladder from autocomplete to full SDLC ownership — most teams sit at L2-L3 today.
Summary
- Sourcegraph's Ado Kukic proposes six levels of code AI (0-5) mirroring SAE driving autonomy levels.
- L0: manual coding; L1: line/block completions trained on OSS; L2: codebase-aware completions (axios-specific vs node-fetch).
- L3: AI implements multi-component features from high-level requirements (e.g. add user auth); L4: AI proactively monitors and submits PRs for doc drift or bug reports.
- L5: full SDLC autonomy — AI designs, codes, deploys, iterates; developer only validates outputs.
- Quinn Slack's prediction: 99% of code written by AI in 5 years; 92% of developers already use code-AI tools (vs 1% a year prior).
code-generationautonomycody
Original description
The Code AI Maturity Model provides a framework for understanding and benchmarking the progress of AI in coding assistant tools, similar to how the SAE Vehicle Levels classify self-driving cars. In this talk, you'll learn about the various levels of the Code AI Maturity Model, where we're at with the available tools today, where we think the future lies, and what it means for you. Recorded & streamed live for the AI Engineer Summit 2023. See the full schedule of talks at https://ai.engineer/summit/schedule & join us at the AI Engineer World's Fair in 2024! Get your tickets today at https://ai.engineer/worlds-fair About Ado Ado is the Director of Developer Relations at Sourcegraph