Should Interviews Move From "Show Me How You Code" To "Show Me How You Prompt"?
How should the technical interview process be in an AI-native world?
Welcome to Infinite Curiosity, a newsletter that explores the intersection of Artificial Intelligence and Startups. Tech enthusiasts across 200 countries have been reading what I write. Subscribe to this newsletter for free to directly receive it in your inbox:
Technical interviews are stuck in the past. Companies still expect engineers to manually churn out code like it’s 2010. It reminds me a bit of the time when there was a “no calculator in the class” rule. And how it quickly became obsolete.
AI-assisted coding tools like Cursor, Copilot, and ChatGPT are everywhere. Garry Tan (President and CEO of Y Combinator) posted the following on X about the latest Y Combinator batch:
Engineers are no longer writing code. They’re getting LLMs to do it. And then they orchestrate it. The best ones know how to guide AI to generate solutions, not just type faster than the next person. So how should the hiring process catch up?
AI Is Changing How Products Are Built
Software engineering is no longer about debugging by hand. Developers used to spend hours fixing syntax errors and optimizing algorithms. Now, a well-formed prompt can generate working code in seconds.
The skill is no longer just writing the perfect function from scratch. It’s knowing how to prompt AI to do it right. Debugging isn’t about manual fixes anymore. It’s about refining AI-generated outputs until they work. The best engineers of the future won’t be human compilers. They will be experts at commanding AI to get the job done.
Why Prompting Skills Matter?
Good prompting separates real engineers from copy-paste engineers. If your idea of debugging is typing “pls fix this” into Cursor over and over again, you’re not going to make it. Strong prompting is about breaking down problems, setting constraints, and iterating on AI outputs.
It’s the same logical thinking and problem-solving ability that has always defined great developers. A well-crafted prompt shows you understand the problem at its core. A weak prompt shows you’re just hoping AI will do all the work for you. Companies need to start filtering for this distinction.
A Better Approach to Technical Interviews
Most technical interviews test outdated skills. They should be testing how well candidates work with AI instead. Instead of throwing a blank coding challenge at someone, interviewers should hand them an AI coding assistant and see what they do with it.
Can they refine prompts to improve outputs? Can they debug AI-generated code methodically? Can they optimize AI’s output instead of rewriting everything from scratch? Strong engineers will refine, iterate, and improve their AI interactions. Weak ones will just take whatever the AI spits out and move on.
Long-Term Implications for Hiring and Engineering Culture
Companies that don’t adapt will fall behind. Hiring engineers who know how to work with AI will lead to better teams, faster development, and higher productivity. Prompt engineering ability isn’t a side skill. It’s the core skill of modern software development.
Teams that embrace this will stop wasting time on manual coding and start building software at AI speed. Engineers who can steer AI effectively will outpace those who refuse to evolve.
A Call to Action for Companies
The old way of hiring engineers is dead. The future of software engineering isn’t about typing fast or memorizing algorithms. It’s about commanding AI to do the work efficiently.
Companies need to start evaluating engineers based on how well they prompt, refine, and debug AI-generated solutions. The ones that do will build better teams and ship faster. The ones that don’t will hire engineers for skills that AI has already automated.
If you're a founder or an investor who has been thinking about this, I'd love to hear from you.
If you are getting value from this newsletter, consider subscribing for free and sharing it with 1 friend who’s curious about AI: