I read an article in The Atlantic that said young people are using ChatGPT to write their applications and HR teams are using AI to read them. And nobody is getting hired.

It sounds exaggerated, but it is kind of true. But honestly, I do not think AI is the problem. I think our hiring process just has not been updated for the world we are living in.

We experimented with a new system at Canvas AI, and it has worked surprisingly well for us.

AI Empathy: Why Machines Score Higher Than Doctors

Meanwhile, a new study says AI is being rated almost 10x more empathetic than human doctors. Not because AI is "too good," but because the reason behind it is far more uncomfortable, both for humans and for the future we are building.

These are not just interesting data points. They are signals that the relationship between humans and AI is fundamentally shifting, and we need to think carefully about what we are building and why.

This ties directly into the broader AI privacy conversation: as AI becomes more embedded in personal interactions, the line between helpful and invasive gets thinner.

As we explored in the 60% revolution, a huge portion of knowledge work (including screening applications) is grunt work that AI can handle. The question is whether we are applying AI to the right parts of the hiring process, or just adding AI to a system that was already broken.

Frequently Asked Questions

Why is AI making the hiring process worse?

Candidates use AI to generate polished but generic applications, while HR teams use AI to screen them. The result is two AI systems talking past each other, filtering out genuine candidates and letting through optimized-for-algorithms applications.

Can AI actually be more empathetic than humans?

Studies show AI is rated as more empathetic in specific contexts because it has unlimited patience and gives consistent, thorough responses. However, this reflects more about the time constraints on human professionals than about AI having genuine empathy.