# Should You Still Learn to Code? 20 Years of Dev Advice **URL:** \ **Category:** Content Creators\ **Tags:** humans-in-the-loop, theseniordev, leon-van-zyl, agents, coding\ **Created:** [5 November 2026 12:00 UTC](https://www.artofsm.art/t/should-you-still-learn-to-code-20-years-of-dev-advice/27519 "2026-11-05T12:00:00Z")\ **Posts on this page:** 2\ **Page:** 1
**Author:** ![artesia](https://www.artofsm.art/user_avatar/www.artofsm.art/artesia/32/36_2.png) [@artesia](https://www.artofsm.art/u/artesia)\ **Post date:** [5 November 2026 12:00 UTC](https://www.artofsm.art/t/should-you-still-learn-to-code-20-years-of-dev-advice/27519/1 "2026-11-05T12:00:00Z")
[![](https://www.artofsm.art/uploads/default/original/3X/c/c/cca06858ad0a30b38d54d55aa6f823ec6e15360a.jpeg "Should You Still Learn to Code? 20 Years of Dev Advice") ](https://www.youtube.com/watch?v=XtvbzN67PhA) The video emphasizes that while AI-powered coding agents reduce the need to write code manually, developers must still possess strong code-reading skills and maintain accountability for code quality, security, and scalability by guiding and reviewing AI-generated outputs. It advocates for integrating personal expertise into AI workflows, using coding agents as learning tools, and adopting structured, responsible practices to effectively collaborate with AI and build sustainable software development careers. ---
**Author:** ![artesia](https://www.artofsm.art/user_avatar/www.artofsm.art/artesia/32/36_2.png) [@artesia](https://www.artofsm.art/u/artesia)\ **Post date:** [5 November 2026 12:20 UTC](https://www.artofsm.art/t/should-you-still-learn-to-code-20-years-of-dev-advice/27519/3 "2026-11-05T12:20:19Z")
The video, presented by a software engineer with nearly twenty years of experience, addresses the evolving role of coding in the age of AI-powered coding agents. It begins by categorizing developers into two groups: senior developers who write code manually and students or newcomers questioning the value of learning to code. The key message is that while writing less code is becoming the norm due to AI assistance, the ability to read and understand code remains crucial. Coding agents, despite their capabilities, often make mistakes such as duplicating functions or creating inefficient code, so developers must review and correct AI-generated code to ensure quality and security. The speaker emphasizes accountability when using coding agents, stressing that developers remain responsible for the security, scalability, and reliability of their applications. They introduce the concept of integrating personal experience and best practices into AI workflows by creating custom “skills” or guidelines that instruct agents on project standards, tech stacks, and necessary features like cookie banners and account deletion options. This approach helps maintain consistency and quality in AI-generated code, leveraging the developer’s expertise to guide the agent effectively. For learning and upskilling, the video recommends using coding agents as patient tutors to explore open-source projects and understand complex codebases. By asking agents to explain technologies, architectures, and specific functionalities, learners can build a personalized knowledge base or “toolbox” of components and best practices. The speaker advises working on projects incrementally, reviewing each piece of code thoroughly before proceeding, and encourages completing projects end-to-end to build a strong portfolio that demonstrates practical skills to potential employers. The video also covers advanced workflows for professional developers, highlighting the importance of mastering concepts over specific tools or certifications. It advocates for using memory files to enforce project rules and standards across sessions and suggests managing development tasks through GitHub issues to treat coding agents as team members. This structured approach includes automated testing, security audits referencing industry standards like OWASP, and continuous integration practices to maintain code quality and security, especially as teams scale up to use multiple AI agents collaboratively. Finally, the speaker addresses skepticism among experienced developers about relying on coding agents. They argue that coding agents are the future due to their speed and cost-effectiveness but acknowledge their limitations in architecture and foresight. The recommended strategy is to start by delegating low-risk or tedious tasks to AI, gradually building trust and refining guardrails to guide the agents. The video concludes by encouraging viewers to embrace AI-assisted development responsibly, continuously learn, and leverage their experience to enhance AI workflows for sustainable and scalable software engineering. ## Useful Links - [GitHub Issues and Pull Request Workflow](https://docs.github.com/articles/using-pull-requests) — Central to the professional workflow and code review process described in the video. - [OWASP Top Ten Security Risks](https://owasp.org/projects/top-ten) — Directly supports the security audit process discussed in the video. - [Drizzle ORM Documentation](https://orm.drizzle.team/docs/overview) — Supports the tech stack standardization approach described in the video. - [Hermes Agent GitHub Repository](https://github.com/NousResearch/hermes-agent) — Central to the learning method using coding agents to explore open-source projects.