How do you get hired in 2026? The rules have changed in the past few months.
In a world where anyone can go from idea to working app rapidly, “what you built” matters far less than “how deeply you understand it.” I just read Sakana AI’s unofficial guide on research hiring, and while it targets researchers, it’s spot on and the lessons apply to every field navigating the post-AI shift. I highly recommend it.
Here is the new hiring meta:
1. The Interview is a Debate, Not a Presentation. No one is going to be impressed with volume (e.g. code, papers). Interviewers are easily bored and they have seen the standard solutions dozens of times. Instead, turn the interview into an interesting discussion. If this doesn’t come naturally to you, treat it as a skill to practice with friends, and even an AI. Be ready to articulate why you made specific choices and defend them against alternatives. What were your “best” failures and what did you learn from them?
2. Depth > Breadth. It is easy in 2026 to be creative with AI tools. To stand out, you can’t just tweak existing paradigms. Find a “good rabbit hole.” Interviewers want to see you think deeply about small interesting things that have potential rather than throw shallow experiments at the wall.
3. AI Accelerates, You Steer. The expectation is now that you will use AI tools to be more productive. But there is a catch: You must understand every line the AI produces. Because AI creates high-level abstractions, you need to also review your basics more than ever. Do you actually remember how an Adam Optimizer works?
4. Be Your Own Harshest Critic. Before you walk in, critique your work like a skeptical reviewer. Know your limitations explicitly. If you don’t know an answer, just say so, then reason through it. Skilled interviewers can smell BS instantly, it’s a huge turn off.
The differentiator in 2026 is clarity of thought and a deep level of understanding.