Summary
Highlights
Despite advancements in AI, learning to code is essential in 2026 due to numerous emerging opportunities. The speaker emphasizes building foundational knowledge over just memorizing syntax, as AI can handle boilerplate code. The core skill to develop is logical thinking, which was traditionally tested with pen-and-paper coding rounds, and remains crucial for identifying AI's shortcomings.
The speaker recommends starting with Python due to its ease of learning, English-like syntax, and broad career opportunities. Python's simplicity helps reduce procrastination and makes the learning curve gentler compared to more complex languages like C++ or Java, which involve intricate concepts like pointers or object-oriented programming earlier on. Python is currently ranked the easiest and most relevant language by TIOBE.
After learning a language, the next step is to master basic data structures and algorithms (DSA). This improves logic building. The speaker suggests solving a few easy, medium, and hard problems for each data structure to establish a strong foundation. Once this is in place, learners can then move on to development, machine learning, or other specialized fields.
In 2026, syntax is less important than understanding the underlying logic and system design. While AI can generate code for individual functions (like a login API), true engineering involves designing how various components interact and anticipating potential issues, such as scalability and data consistency. The speaker uses the example of a food delivery app to illustrate the complexity of system design, which AI cannot fully handle. Interviews today often test these system design skills with the help of AI assistants.
Version control, particularly Git and GitHub, is a critical skill for any developer. It allows tracking changes, collaborating with teams, and reverting to previous versions when issues arise. The speaker regrets not learning this effectively in college, noting its importance in professional environments and how accelerated AI-driven development makes it even more crucial for managing frequent code changes.
Logic building is the most vital skill for an engineer. It involves knowing how to structure decisions (if-else), loops, variable definitions, and optimizing operations like network calls. The speaker emphasizes that language is just a medium; the core skill is computational thinking—understanding how to build and optimize logic. Mathematical concepts, particularly Boolean logic and basic algebra, can fast-track this learning.
AI should be treated as a powerful assistant, not a replacement for learning. It can help by generating roadmaps for new topics, creating practice tests, and explaining complex concepts in simpler terms (using techniques like 'EL10' - Explain Like I'm 10). Familiarity with AI tools like Claude, Cursor, and Codex, as well as platforms like Hugging Face, is essential for staying current and running models locally.
The most effective way to learn is by actively implementing knowledge through projects, competitive programming, and hackathons. The speaker advises building three types of projects: one that personally excites you, one that enhances your resume by demonstrating complex skills (e.g., multiplayer games, compilers), and one that challenges you beyond existing tutorials, requiring you to learn from documentation or experiment with advanced AI models.
Contributing to open-source projects is a pro tip for gaining real-world coding experience and opening doors to programs like GSoC (Google Summer of Code). The speaker concludes by reiterating his recommended learning path: start with pen and paper, then learn engineering foundations, followed by logic development, and finally, build projects. He emphasizes the critical role of ideas and execution in the current coding landscape, especially with AI handling much of the boilerplate.