[fhd2]

Things are turning truly bonkers right now. I've hired two developers in the last couple of weeks:

1. One from my network, just announced it, someone I had worked with in the past reached out, quick chat, hired, great.

2. One with the usual approach of posting job ads and all that. We got an _insane_ amount of noise, even as an obscure, small company. I've hired hundreds of people, but most of those three years ago or earlier. Never seen such noise, most candidates barely meeting any of the requirements, weird auto generated cover letters and CVs. It was a bit exhausting, but I went through everyone manually to make sure I'm not accidentally filtering out a solid candidate. We found two in the end, one quickly backed out because they got another offer. But there was one good candidate left, and they accepted the offer. I don't remember this being so hard.

A few years ago I'd call people trying to automate screening or mainly hiring from their network lazy. In this time, I see the appeal.

The last thing I need is more bots spamming me on my LinkedIn account on top of all this madness.

[neilv]

I think some recruiters are already doing AI-slop, too. (Not just the LinkedIn-spamming templates that some sourcers were doing before.)

Hopefully we'll get a big backlash against disingenuous passing off AI-slop as a communication from a human.

Maybe sending someone an AI-slop message will become universally recognized as trashy behavior -- employed only by the corporate communications of companies that really don't care, but not by anyone respectable.

[notpushkin]

Yeah. Personally, I’ve also tried reaching out to people in companies I’m interested in, but automating that just feels dishonest somehow.

[fhd2]

The reaching out might work, I know some people actually appreciate it. I personally don't, it feels like an attempt to circumvent the system, and the system is in place for a reason: It attempts to make the insane volume of candidate messages manageable. It's not perfect, I wish more people would work on better systems rather than ways to circumvent them.

There is one notable exception to this, a guy I actually ended up hiring: The job ad link I posted stopped working after a few days, and he wrote a nice message to the company (!) email address telling us about that, and saying he's interested if the role is still open. That's proactiveness I really appreciate. It's very different from random people reaching out on my personal LinkedIn account or email, which I simply ignore.

I'd _love_ it if there wasn't so much noise, then I would probably think about it differently. But the volume does burn you out, and any attempts at circumventing systems put in place to try and manage this is not helpful.

And that's for manually written messages. Automating that - Jesus.

[tough]

you should always pre-vet any message you send and sign with your name.

it doesnt really matter where the message or how was created, only that you're 100% ok with its contents reaching its destinatatary.

that still puts some limits to how much you can -spam- and should make you not want to -spam- but only send real messages, but you can certainly use AI to help you research/generate those leads initial messages

i have not tried to do this, but i wouldn't see how its problematic?

[notpushkin]

That’s how I would use this (and is my approach to LLMs in general). I’m just afraid that others would take a more lazy approach and spoil it for everybody else: https://news.ycombinator.com/item?id=44274583

[dloku]

thanks man!

[tough]

no problem! gl with the job search

[dloku]

right, we don't automate...

[notpushkin]

Yeah, but (if I understand how your product works correctly!) I’m afraid the laziest of your users would just copy-paste the suggested message and send it as is, to as many people as possible.

This would quickly saturate the decision-makers’ inboxes, and even those who genuinely put some thought into the roles they apply to and rewrite the suggested openers would be lost in this cacophony.

Is there some way to prevent that?