Hi ,

Happy Sunday!

Whenever I conduct an interview, I try to turn it into a dialogue as quickly as possible. When you have been doing this long enough, you realize you do not always need a whiteboard to evaluate technical expertise. You can gauge a developer's depth simply by talking to them about seemingly random technical topics.

Right now, my absolute favorite "random" topic to bring up is the usage of AI tools like Codex, Claude, or Copilot.

The way a candidate answers this question tells me immediately if I am talking to a Junior, Mid-level, Senior, or Senior+ developer. The focus shifts entirely from "can you solve FizzBuzz" to how you evaluate risk, architecture, and team dynamics.

Here is what I see in these conversations and what your answer signals about your experience level.

Junior: Blind Trust When I talk to junior developers, they often fall into the bucket of "AI makes everything better!"

They are thrilled that an LLM can spit out a complete React component in seconds. The red flag here is uncritical enthusiasm. When I ask them how they verify the generated code or what security risks they consider, they usually can't answer in any meaningful way.

=> They are using AI to increase their output, but they lack the foundational knowledge to catch the AI's mistakes.

Mid-Level: The Skeptic Mid-level developers have usually been burned by AI at least once. They asked it to solve a complex problem, it hallucinated a non-existent API, and they wasted three hours debugging it.

Because of this, they often fall into the bucket of "I tried it once, and it is useless." They dismiss it entirely because it is not perfect.

=> They have enough experience to spot bad code, but they are lacking the adaptability to change their workflow. Refusing to use a tool because it requires supervision is a missed opportunity for growth.

Senior: The Pragmatist This is where the conversation gets interesting. A true senior developer sits firmly in the bucket of "I use it daily, and here are its exact limits."

In a dialogue, a senior will tell me: "I use Copilot to generate boilerplate for my unit tests, but I turn it off when writing domain-specific business logic because it just lacks the context of our monolith."

=> They understand that AI is just another tool. They know exactly when it saves time and when it introduces unacceptable risk. This is the exact critical thinking I want to see.

Senior+ / Staff: The Strategist When I talk to Staff or Principal engineers, the conversation naturally elevates beyond their personal IDE.

They do not just talk about how they use AI. They talk about how the team should use it. They bring up concerns about pushing proprietary code into cloud models. They suggest adding new steps to the PR review process to catch AI-generated bugs.

=> They are looking at the overall impact. They are thinking about architectural integrity, legal compliance, and team productivity as a whole.

The Takeaway In your next technical interview, try to move away from treating it like an exam. Treat it like a conversation with a future colleague.

When the topic of AI comes up, do not just say "it is cool." Talk about a time it failed you, what you learned from it, and how you adjusted your workflow. That is how you prove your seniority without writing a single line of code.

If you have an interview coming up and want to practice this exact type of technical dialogue, we can do a mock session together.
We will just talk tech and see where you naturally land.

Have a great start to your week,

Cheers
Andre