Venkat Subramaniam Interview

The Better You Are at Programming, the Worse AI Looks - Venkat Subramaniam | The Marco Show

The Better You Are at Programming, the Worse AI Looks - Venkat Subramaniam | The Marco Show

Points touched upon

• [[Tech/Pioneers/Venkat Subramaniam]]’s views on AI and how it is changing software development.
• A fun conversation, including his funny but tense story of his computer dying minutes before a conference talk, and how he still managed to give it.
• That talk is here: [[Tech/Programming/Know your Java]]
• Venkat Subramaniam's perspective on AI in software development, its unpredictable nature, its strengths in identifying bugs and risks, and why generating code without proper guidance can be dangerous.
• The importance of developer experience when using AI, how experienced programmers can use it for rapid prototyping and why evaluating, verifying and maintaining generated code requires strong technical knowledge.
• Why junior developers must continue learning programming fundamentals, practising independent problem solving and critical thinking, rather than relying on AI and development tools for quick answers.
• The idea of AI second instead of AI first, allowing time for independent thought, discussions with colleagues and combining human and AI perspectives to arrive at better solutions.
• The distinction between programmer productivity and business agility, the problems with measuring AI usage through tokens and why organizations should encourage collaboration instead of forcing adoption.
• Venkat's approach to continuous learning through experimentation, conversations, books and conferences, his distinction between studying and learning, and why curiosity and practical understanding matter more than grades.
• How Venkat prepares technical talks and keynotes by collecting ideas over months or years, developing stories from personal experiences and delivering presentations without rehearsing, including his experience with a failed laptop during a live coding keynote.
• The evolution of Java since Java 8, the importance of lambdas and streams, the benefits of its release cycle, the approach to asynchronous programming and the future need for generics with value types to improve performance.