Short talks: Past, Present and Future of AI
summary
A rapid history of abstract interpretation that traces concrete claims and engineering outcomes: presents the formal method used to analyze Airbus C code with thousands of floating variables and hundreds of thousands of lines, then describes turning research into an industrial tool that analyzes 15 million lines of code and ships under license. Details include the intermediate trace semantics trick, European projects that funded prototypes, and a deployed static analyzer that successfully finds runtime and safety issues.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 2:15Personal journey with abstract interpretation
- 11:2540 years of AI research
- 23:47Patrick Cousot’s lectures
- 32:10Tribute to Patrick Cousot
- 57:55Abstract interpretation’s impact
- 1:12:54First encounter with abstract interpretation
- 1:20:56Personal history with abstract interpretation
- 1:29:45Closing remarks
related talks
01h 22mSoftware Abstraction in the Age of AI-Generated Code
26mAl and Software: The Good, The Bad, The Scary
56mFundamentals of Software Engineering In the age of AI
47mAI and the Future of Software Development
15mUsing AI effectively in large Java codebases
59mIt AI n't What You Think!
14mWhat 80 years of coding history says about AI tools today
26mThe Developer’s New Teammate: AI in Embedded and IoT Environments