We've figured out technical interviews in an agent-first world.
The 2 most critical skills are problem finding and problem solving. We have tests for each that are truly impossible to hack or cheat.
DM me if you want to join our team and try it out.
Users are excited about new technical interview tests for AI agent-era roles because they see solid success stories and call the approaches impressive.
No Digg Deeper questions have been answered for this story yet.
I actually now feel more hope for post-RSI job security after giving these sorts of interviews.
One surprising result is that candidates still have drastically different performances on our interviews. This means that some humans are still better than other humans for our engineering needs, even if agents are already superhuman at many aspects of engineering.
More recently, engineering feels more like an allocation problem of how to allocate our time and tokens to problems. I think human capabilities will continue to co-evolve with those of models so that humans will just be able to allocate increasingly-capable agents to increasingly-difficult problems.
Some humans are better than others at that, so therefore those humans will continue to have jobs.
We've figured out technical interviews in an agent-first world.
The 2 most critical skills are problem finding and problem solving. We have tests for each that are truly impossible to hack or cheat.
DM me if you want to join our team and try it out.
@philhchen hard pivot to recruiting???
@johnrtian solid success stories 💪
@philhchen so cracked