the candidate experience angle is definitely worth its own post because i don't think it's black and white. plenty of companies get thousands of applications and are still dissatisfied with the talent quality (although i'm skeptical of this, i don't dismiss it entirely). others get thousands and simply can't process the volume, with 75% left unreviewed (ask me how i know).

if you work in big tech and you've ever referred an extremely qualified friend who's also in big tech and watched it get ignored despite 10 open support engineer or 20 account executive roles they'd be perfect for, that's exactly the incentives and infrastructure problem i'm describing.

AI might help with these problems for both recruiters and candidates though. pre-AI the pitch for many recruiting tools was about skills and keywords (garbage in, garbage out). but now we can use LLMs to reason positive signals related to qualification (compared against the job description, does resume/application show experience match based on years, sales quotas, industries, programming languages, etc.). the goal is surfacing more qualified candidates for human review, which matters a lot when a recruiter is literally waiting 5-10 seconds for one application to load in workday while every incentive they have pushes them to just source someone off linkedin instead.