Gleaning information isn't the goal; whittling down deluge of applicants is. For the company, candidate time is free and manager time is massively expensive. The AI tools are cheaper than hiring more HR staff, so companies buy them lest they be haunted by the ghost of Milton Friedman.

Anybody who has been on the hiring side post-GPT knows why these AI tools are getting built: people and/or their bots are blind-applying to every job everywhere regardless of their skillset. The last mid-level Python dev job I posted had 300 applicants in the first hour, with 1/4 of them being from acupuncturists and restaurant servers who have never written a line of code. Sure, they're easy to screen out, but there are thousands to sift through.

Having said that, I don't like AI interview tools and will not be using them. I do understand why others do, though.

> The last mid-level Python dev job I posted had 300 applicants in the first hour, with 1/4 of them being from acupuncturists and restaurant servers who have never written a line of code.

That has to be due to policy failure of forcing people on benefits to apply for jobs to get benefits, even if they already have applied to all suitable jobs there are right now?

> candidate time is free and manager time is massively expensive

This is a naive view of the proceedings. Why not hire literally the first person that applies? That would reduce your cost even further.

The point is to figure out who would be good at making you money. The question is, does an ai chatbot wasting your prospective candidates time make you more, or less, likely to find people good at that? Perhaps it reduces the amount of cost reviewing applications, but I imagine it also drives away a good number of the better candidates, those that have more options, away. If you're cutting corners and cost this much, why are you even hiring? surely the point of the exercise is looking towards future growth.

Naturally, there is also a limit to that line of thinking also - spending weeks reviewing each one of the ten thousand applications to your junior developer role wouldn't be the most efficient way to grow. But surely there are better filtering methods you can think of than this, which is imo the equivalent of planning on reducing the number of candidates by lining them up in a room for hours in sweltering heat and hurling verbal abuse at them until only a couple of the wretched ones without a shred of dignity are left

[dead]