Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Recruiters are going analog to fight the AI application overload

wired.com

Post reply on HN
Print view View original

101–110 of 119 posts

Re: Recruiters are going analog to fight the AI application overload

#101

Post by rnewme »

surfingdino wrote: post #72
esafak wrote: post #65

A new problem I'm facing on the hiring end is an influx of imposters from Asia. Everything they claim is a lie, from their name, citizenship, location, picture, to their job history. They even set up fake company sites. And of course they use LLMs to help them pass the interview. I assume it's so they can bilk gullible companies and perhaps do some corporate espionage. I don't know. I haven't encountered anyone try t…

Last year I was reviewing about a dozen answers to the questions we were sending candidates to complete at home. In all cases those a*holes used ChatGPT to write answers and what did ChatGPT use? Original Python docs recombined with blog posts. I was able to verify this by copying passages from their answers and pasting them into Google. When I showed this to the hiring manager they had a serious conversation with th…

People you're calling assholes - did they get paid for the take home task? We're they instructed not to use llms? Have you confirmed they used them and didn't Google the same things you googled? Was the homework not correct?

Re: Recruiters are going analog to fight the AI application overload

#102

Post by pseudalopex »

vunderba wrote: post #79
eikenberry wrote: post #24

Earlier quoted context omitted.

The challenge for certifying in software engineering seems similar to the challenges you'd face in any creative, building oriented professions. That is you can't test for it with some sort of standard exam. You need to have the person build something and then have established masters of the field evaluate it. But that requires a professional structure we don't have.

Many other engineering disciplines have a PE exam, including mechanical engineering and electrical engineering. I'm not sure why you believe that software engineering is somehow more difficult to test in a rigorous manner than those fields. In point of fact, computer science did at one point have an official equivalent PE exam, but it was bafflingly removed a few years ago though I'm not entirely clear on the reason.…

> In point of fact, computer science did at one point have an official equivalent PE exam, but it was bafflingly removed a few years ago though I'm not entirely clear on the reason.

It was removed because 100 people registered in 6 years.

Re: Recruiters are going analog to fight the AI application overload

#103

Post by nradov »

esafak wrote: post #65

A new problem I'm facing on the hiring end is an influx of imposters from Asia. Everything they claim is a lie, from their name, citizenship, location, picture, to their job history. They even set up fake company sites. And of course they use LLMs to help them pass the interview. I assume it's so they can bilk gullible companies and perhaps do some corporate espionage. I don't know. I haven't encountered anyone try t…

I predict that high profile employers will soon start forcing job applicants to go through an identity verification service like Clear before allowing them to even fill out an application form.

https://www.clearme.com/clear-verified

Re: Recruiters are going analog to fight the AI application overload

#104

Post by parineum »

fwip wrote: post #75
nnurmanov wrote: post #59

Earlier quoted context omitted.

My suggestion is to focus on a fewer number of companies, instead of sending resume in bulk. When you narrow your choices you can study well and notice bugs or issues that need to be addressed. Fix them, send the results to someone in the company, it will increase your chances substantially.

It's a sad day when the answer is "do unpaid labor for a chance to be noticed."

The people doing the interview are being paid but the company itself isn't making any money from the interview process either and it's dedicating resources that could be more productively used elsewhere for immediate results.

Both parties are making short term sacrifices in hopes of finding a mutually beneficial relationship.

Re: Recruiters are going analog to fight the AI application overload

#105

Post by TeaBrain »

esafak wrote: post #65

A new problem I'm facing on the hiring end is an influx of imposters from Asia. Everything they claim is a lie, from their name, citizenship, location, picture, to their job history. They even set up fake company sites. And of course they use LLMs to help them pass the interview. I assume it's so they can bilk gullible companies and perhaps do some corporate espionage. I don't know. I haven't encountered anyone try t…

A former co-worker mentioned to me that his dad runs a side business where he creates fake work and education histories for people coming to the US from his home country on work visas. He said that it works because employers never bother to verify work or education histories from said home country.

Re: Recruiters are going analog to fight the AI application overload

#106

Post by surfingdino »

rnewme wrote: post #101
surfingdino wrote: post #72

Earlier quoted context omitted.

Last year I was reviewing about a dozen answers to the questions we were sending candidates to complete at home. In all cases those a*holes used ChatGPT to write answers and what did ChatGPT use? Original Python docs recombined with blog posts. I was able to verify this by copying passages from their answers and pasting them into Google. When I showed this to the hiring manager they had a serious conversation with th…

People you're calling assholes - did they get paid for the take home task? We're they instructed not to use llms? Have you confirmed they used them and didn't Google the same things you googled? Was the homework not correct?

They were not paid, but we spared them the stress of a timed online coding exercise which a lot of devs hate. We told them we expect short snippets of code to illustrate the answers to six questions which progressed from the obvious (if you code in the language for a few years) to more of "there are many was to do it, here is mine" to judge their ability to turn requirements into code. You could do the whole thing in half an hour. It was fair.

Re: Recruiters are going analog to fight the AI application overload

#107

Post by saxelsen »

Aurornis wrote: post #58

I've participated in a coaching and resume review program for many years. Most college students always struggled to write resumes, but at least they were aware of their shortcomings and reached out for help. Something about ChatGPT has convinced a lot of them that they don't need help or even to put much effort into the process. They hate writing resumes, so they just dump some words into a ChatGPT prompt and copy-pa…

Honestly I hate the process of writing resumes, too, and having to tailor an application creatively to a specific job to even be able get an initial interview is draining and my time would be better suited.

I am hoping for a world where more automation leads to better match making so that the right candidate can more quickly be found for the right role.

If we are at a point where there are 3k applicants for a job because writing the application is easy, that might suck for recruiters having to filter those, but for the company hiring I'd imagine it's a benefit that there is a larger flow of candidates that they can choose from. Now the filtering just needs to get better.

Re: Recruiters are going analog to fight the AI application overload

#108

Post by surfingdino »

knallfrosch wrote: post #80
surfingdino wrote: post #72

Earlier quoted context omitted.

Last year I was reviewing about a dozen answers to the questions we were sending candidates to complete at home. In all cases those a*holes used ChatGPT to write answers and what did ChatGPT use? Original Python docs recombined with blog posts. I was able to verify this by copying passages from their answers and pasting them into Google. When I showed this to the hiring manager they had a serious conversation with th…

Did you tell them that AI was not allowed?

We did not, but using AI to generate or reformat answers is a big red flag to me, because it shows that a) you do not have the experience you claim to have, and b) you do not have written communication skills or knowledge of English required to do your job.

Re: Recruiters are going analog to fight the AI application overload

#110

Post by Clubber »

surfingdino wrote: post #106
rnewme wrote: post #101

Earlier quoted context omitted.

People you're calling assholes - did they get paid for the take home task? We're they instructed not to use llms? Have you confirmed they used them and didn't Google the same things you googled? Was the homework not correct?

They were not paid, but we spared them the stress of a timed online coding exercise which a lot of devs hate. We told them we expect short snippets of code to illustrate the answers to six questions which progressed from the obvious (if you code in the language for a few years) to more of "there are many was to do it, here is mine" to judge their ability to turn requirements into code. You could do the whole thing in…

We're really struggling to find good people. We do it the old fashioned way, just a one hour 4 dev interview. We don't ask really tough questions or anything. An example is, "can you describe normalization and what it's for?" I think the market is just supersaturated with bads right now and all the bads are the ones applying.

The good developers we have we got from knowing them in previous jobs, etc. You might wanna hit up your current team to find better devs. You should probably pay your team a finder's fee if you hire from that. Also, if you are paying what you paid before covid, it's probably too low today.

Times are tough for building good teams right now. You need to adapt. The same ole "call a recruiter and get 1 good guy out of 10," isn't a thing anymore. You might wanna pay 10-20% over market and see if that works. To justify that, just calculate what you are paying to not find good people now.

Post reply on HN