A year ago, the debate over whether candidates should be allowed to use AI tools during take-home assignments was all the rage. Now, most engineering teams have moved on. The question isn't whether you use AI, but whether you can use it well. That's a big difference.
The shift has had a profound impact on technical hiring. Some candidates are getting a leg up, while others are struggling to keep up. To succeed, you need to understand what's going on. "Vibe coding" - directing AI tools with natural language to produce working code, then refining and integrating the output - is no longer a novelty, but a baseline expectation at many companies. Senior engineers now evaluate candidates on more than just their ability to write code from scratch. They want to see if you can write precise prompts, critically review AI-generated code, and know when not to use AI.
For instance, when I was interviewing for a role at a mid-sized startup, I was asked to write a prompt that would generate a specific function using AI tools. The interviewer didn't just want to see the output, but also how I arrived at the prompt and what I would do if the output wasn't what I expected. It's not just about using AI, but about using it intelligently.
