Tangential to this. I wonder how technical interviews are going to evolve now that LLMs are so powerful and part of the skill set.
Right now I could imagine you wanting to test the higher order skills candidates have. Thinking about memory hierarchy and distributed systems.
The exciting part might be that they have to continue to evolve rapidly as model capabilities progress.
Referred a friend for a job
Now HR wants to interview me again 😭
Users approve of technical interviews that test debugging of AI-generated ML training code because they see the approach as interesting and valid for assessing higher-order skills.
No Digg Deeper questions have been answered for this story yet.
Leet smoketest
@code_star An interesting interview I did last cycle had me correct training code with bugs in it. I think this is a valid interview since it tests your ability to monitor generated code
@code_star An interesting interview I did last cycle had me correct training code with bugs in it. I think this is a valid interview since it tests your ability to monitor generated code