Live data from Hacker News

Interviewing a software engineer who prepared with AI

kapwing.com

261–270 of 731 posts

Re: Interviewing a software engineer who prepared with AI

#261

Earlier quoted context omitted.

> Some people think it's perfectly normal to stretch the truth on a resume, and to lie in an interview. So marketing works in the company's favor, and not the candidates? Its a tough pill to swallow, but bending the truth and lying seems to be the way folks get jobs now. Perhaps not lying... But I've thought about the 1pt font white on white mega-tech-list attached to Workday resumes to get past THEIR ai-slop filters…

Lying isn't marketing. If you lie in marketing, people can sue you. It is one thing to frame your experiences in ways that are relevant to what the job is looking for: it is not only unethical to fabricate experiences, it is counter-productive. I will be checking references, and if their reports of the role you played on a project don't match yours I will not be hiring you. If you don't have references who can speak…

Completely agree with the distinction you're making: framing is fine, fabrication is a deal-breaker. It's frustrating how often people conflate "putting your best foot forward" with just making stuff up, especially when they underestimate how easily it can fall apart during reference checks or follow-ups.

Re: Interviewing a software engineer who prepared with AI

#262
post #257

Earlier quoted context omitted.

> People have been lying about their experience for literally centuries. Millenia. Just ask Nanni what happened when he trusted Ea-nāṣir. https://en.wikipedia.org/wiki/Complaint_tablet_to_Ea-n%C4%81...

Man, how annoyed do you have to be with someone to carve your complaint into stone!

Pressing an edged tip into damp clay to write a complaint or ledger is almost as fast as quill and ink writing, the clay takes longer to dry solid than ink though.

Re: Interviewing a software engineer who prepared with AI

#263
I don't think this has anything to do with using AI for prep. 20 years ago I was interviewing candidates who had somewhat lied on their resume, knew some of the things that they'd written about, but had everything fall apart under a little more questioning of what exactly they'd done and why.

Re: Interviewing a software engineer who prepared with AI

#264

I keep coming back to this phrase used in this post: "it was scary". Yeah, hiring is scary. Hiring is insanely expensive on all fronts. Firing people is difficult, it's expensive and legally exposing. Hiring the wrong person, allowing them access your systems and potentially exfiltrate your IP to them is a hazardous but necessary venture. The thing is, none of these things really changed with AI. People have been lyi…

In most highly developed countries, there is a probation period for 90+ days for new hires. During this period, you can be fired for any reason. It is not an expensive as people think to fire someone who deceived your hiring process. However, institution inertia is real.

That said, I very much agree with your last paragraph. In the late 1990s and early 2000s, a lot of hiring was done this way in the US.

Re: Interviewing a software engineer who prepared with AI

#265

I keep coming back to this phrase used in this post: "it was scary". Yeah, hiring is scary. Hiring is insanely expensive on all fronts. Firing people is difficult, it's expensive and legally exposing. Hiring the wrong person, allowing them access your systems and potentially exfiltrate your IP to them is a hazardous but necessary venture. The thing is, none of these things really changed with AI. People have been lyi…

wrote something similar a while back: https://www.softwaredesign.ing/blog/ai-is-the-reason-intervi...

Re: Interviewing a software engineer who prepared with AI

#266
post #263

I don't think this has anything to do with using AI for prep. 20 years ago I was interviewing candidates who had somewhat lied on their resume, knew some of the things that they'd written about, but had everything fall apart under a little more questioning of what exactly they'd done and why.

I think the difference is that you used to need a certain knowledge to be able to bullshit. You could still do it, but it would mainly be to embellish stuff you already somewhat know. With LLMs, it's easy to make it write a whole page of interview prep you can use to hide your tracks, without any prior knowledge. My guess is they saw that kapwing wanted experience in X,Y,Z and made an LLM create projects that sounds real in a way you otherwise wouldn't be able to do as easily.

Re: Interviewing a software engineer who prepared with AI

#267
post #263

I don't think this has anything to do with using AI for prep. 20 years ago I was interviewing candidates who had somewhat lied on their resume, knew some of the things that they'd written about, but had everything fall apart under a little more questioning of what exactly they'd done and why.

I think the difference is that you used to need a certain knowledge to be able to bullshit. You could still do it, but it would mainly be to embellish stuff you already somewhat know. With LLMs, it's easy to make it write a whole page of interview prep you can use to hide your tracks, without any prior knowledge. My guess is they saw that kapwing wanted experience in X,Y,Z and made an LLM create projects that sounds…

From the article:

> but it had been some time ago, and they never worked on any of the features

It appears that the candidate might have actually worked on the daycare app, but not on what they said they worked - i.e., the ratelimiting and pagination. It appears that they might have been working on the frontend, and took the liberty of "expanding" their role - this used to be extremely common in a big sample of the resumes, and I'm guessing it still is. They might have used AI to prep - they used to use google earlier, but the prep was (and is) still inadequate if you've not actually worked on and implemented it. I don't think it was an entirely LLM created project...

Re: Interviewing a software engineer who prepared with AI

#269
post #254

I keep coming back to this phrase used in this post: "it was scary". Yeah, hiring is scary. Hiring is insanely expensive on all fronts. Firing people is difficult, it's expensive and legally exposing. Hiring the wrong person, allowing them access your systems and potentially exfiltrate your IP to them is a hazardous but necessary venture. The thing is, none of these things really changed with AI. People have been lyi…

I get where you're coming from, especially on the cost of bad hires: it really is one of the riskiest bets a company makes. But I'm not convinced going back to the "fly them out and grill them on a whiteboard" era is the right answer either.

There's an opportunity for wework for hiring - rent out a conference room for a couple hours and have a third party be present during the interview. The first one to figure this out and not go bankrupt a year or two later wins. Probably not a unicorn business, though.

Re: Interviewing a software engineer who prepared with AI

#270

I keep coming back to this phrase used in this post: "it was scary". Yeah, hiring is scary. Hiring is insanely expensive on all fronts. Firing people is difficult, it's expensive and legally exposing. Hiring the wrong person, allowing them access your systems and potentially exfiltrate your IP to them is a hazardous but necessary venture. The thing is, none of these things really changed with AI. People have been lyi…

> - awkward and/or neurodivergent people are going to have to learn social skills again It's not like these are skills that they haven't learned, these are things that they have a hard time with. Expecting them to be 'normal' is like asking a person of medium stature to be taller. They could mask them but ultimately it's not who they are and expecting everyone to be the same is a fools errand.

It's not expectation of sameness, it's that they will be working with 'normal' people and need to meet some standard to not be a net negative for any team they're in.
Post reply on HN