Develop Yourself #230 - AI Will Replace Most Coding: What Chief Scientist Laly Bar-Ilan Says Developers Should Learn Next
Apr 14, 2025
Laly Bar-Ilan, Chief Scientist at Bit, has been pioneering AI and natural language processing for over 20 years. In this engaging discussion, she highlights how AI is set to revolutionize coding jobs, urging developers to embrace evolving skills like retrieval-augmented generation and prompt engineering. Laly emphasizes the importance of continuous learning as traditional coding methods become obsolete. She also explores the future of composable software and how AI can streamline development, making adaptability key in the shifting tech landscape.
AI Snips
Chapters
Transcript
Episode notes
AI Replacing Developer Tasks
- AI is poised to replace most repetitive tasks developers currently perform, including architecture and platform building.
- This transition is already underway at a rapid pace with AI capabilities advancing by the minute.
Must Learn Foundational and New AI Skills
- Junior developers must learn to code to evaluate AI-generated code and understand architectural principles.
- New skills such as RAG, model fine-tuning, and guardrails are essential to integrate AI into deterministic systems.
Brian's AI Coding Experience
- Brian writes 80 to 90% of his code with AI but finds AI often wrong on the first try, especially on complex tasks.
- He validates and fixes AI-generated code, emphasizing that knowing the code remains essential.