Responding to recruiter emails with GPT-3 | Hacker News Reader

Responding to recruiter emails with GPT-3

matthewbilyeu.com

213 pointsmbil133 comments

I'm sure some folks will see this as dishonest or unethical, but it feels more like the natural evolution of getting spammed by recruiters.

If recruiters are going to endlessly spam me with their templates and automation, it only seems fair that my response is also templates and automation.

tinder already allows you to fabricate entire dialogue with preloaded responses last I checked.

at this point I might as well generate mtiple dopplegangers seeded by various movie monologues and see what happens.

I bet my 007 doppleganger is hawt

If you email a hacker/programmer whose powers you have heard can aid you in your quest, you should be prepared to endure their intricate labyrinths or spells of invisibility. Their magick is only for those of pure intentions.

Sir, may I interested you in my automated template spam reply system for dealing with automated template spam recruiters? https://github.com/drcongo/spammy-recruiters

Interesting. Is there a US variant out there, too?

(This one looks promising: https://github.com/jceloria/recruiter-spam/blob/master/list....)

Seeing as recruiters operate without ethics, this seems fair.

As one article taught me about over a year ago respond by sending a well formulated email requesting the salary and job offering. Sometimes it leads to you understanding the market. Sometimes it leads to negotiation for higher salary. Sometimes an amazing new job.

So far this has never failed me. And is far better than ignoring. I only really ignore the crypto companies.

I really like this advice! Do you have a link to the article? Would like to learn what you mean by "well formulated."

https://index.medium.com/career-advice-nobody-gave-me-never-...

I wonder if it being so long is part of the filtering process. Only good recruiters will read the whole thing and bother to reply.

I personally believe it is unnecessarily long, formal, and diplomatic. It does not fit who I am and, if I were to read it, I would have the impression of an annoying and insincere corporate speak.

I believe I can achieve all the goals of being polite, respectful, considerate, and filtering for serious offers with a much, much shorter and informal response.

Something like:

Hi X, thanks for the contact. Company Y sounds interesting and I do think I might be a good fit for the role. I am not actively looking for a new position now, but i’d like to know more about the job description and compensation details. Thanks!

This should all be sincere if course. If the company does not sound interesting (like all crypto for me), I only reply a respectful “no” ( ”Thank you for the contact, but I am not looking for a new job right now.” ). Similar when job is for roles that I have no experience whatsoever, which often happens. In this case I also mention that I am not a, e.g., Android engineer (I am a web frontend).

Also, if there is no company or role mentioned, I give the first reply, but obviously removing the acknowledgment that I like the company/role and ask for details about it too.

Also, what this does is really separates the wheat from the chaff. If the company can't tell me the salary or the job, it's not a good fit anyways.

Interestingly, what this means is that Google is not a good fit at all. They wanted me to talk to a sourcer, who would connect me to a recruiter. That's too much corporate foreplay for me, especially without even knowing what team I'd be joining.

> corporate foreplay

Hello, new vocabulary entry!

Maybe they'll even buy you a lobster dinner before they fuck you.

While I agree, its also a way to

a) know the market

b) not interview for 4-8 hours only to find out they are going to pay you 30k less than your current job (yep had that, never make that mistake again)

Taking into account that some sort of automated tool also generates recruiter letters and follow-ups - we witnessed a fight of two computers without any value in it for anyone.

It made me laugh, so I’d call that value.

If a tree falls in a forest and no one is around to hear it, does it make a sound?

Depends on who you ask. Newton would say, "yes". Goethe would say "what do you mean by sound? What if there were two people in different locations? Would they hear the 'same exact' sound?"

Only with Universe Standard and above

Or rather we have witnessed two computers talking to each other and decided their bosses aren't a good fit at this time (saving the time for both sides while also not missing any potential matches).

My response to recruiters (especially on LinkedIn) is always the same; “Please send over all relevant details for the role you had in mind (JD, company and salary)”.

This filters out about 90% of them, and especially those who’re just fishing for clients.

Fishing for what sort of clients?

They mean fishing for CVs.

Many recruiters invent roles so that they can build a database of CVs.

...or they have a 1+ year long 'open position' so they always have an infinite pool of applicants.

Invent roles, or straight up lie. The giveaway is usually that they want a call with you to discuss details, or can’t produce any job description.

In fact, this is most recruiter inquiries.

I used to do the same thing but I got tired of doing that (and they didn’t always respond with the details, but often with another request). Now I just send them to https://polyfill.work.

They have to enter pay, etc. and if it’s a fit for me I’ll hear about it.

(Disclaimer: I made polyfill and it’s a work in progress but I’m looking for feedback on the developer onboarding part so let me have it if it sucks!).

I don’t send them my polyfill profile page because then they’ll have my email and uuid and I want polyfill profiles to be pseudo anonymous.

I got curious, but even after reload it won’t go past step 4

Thanks for giving it a go! I’ll check sentry, etc. to see if there’s anything obvious.

On the off chance you see this reply: what browser/device were you using?

I also see that step 4 is the first multiple choice question, so it doesn’t auto-scroll to the next question (you have to hit “next” to move on).

It’s not obvious. Perhaps the culprit here is just bad UI.

Right on

- Device is an iPad running iOS 13.5

- I tried a bunch of things on Q4 both while selecting multiples and selecting a single answer:

  - Trying to scroll "past" it to push through any CSS padding issues
  - Trying to tap the next button
  - Toggling through different states to see if the button changes in any way to see if there is JS with a logic bug (spoiler: while the button did show up as expected, nothing changed it's colour or otherwise indicated that it was clickable. Holding the button just caused text on the page to be selected which confirmed that the button did not have the ability to be clicked)
  - Reloading the page
  - Choosing different answers for Q1-Q3

Dang, that's annoying, sorry about that! And thanks for checking it out so thoroughly. I queued this up and my friend or I will get to it soon. Really appreciate the QA :D

If you want a heads up when it's fixed you can email me at team at polyfill dot work. I will only email you once: when the bug is fixed!

I felt like this blog post ended quite abruptly. I was expecting for the author to show how GPT-3 was able to formulate a response back to the recruiter not just parse out the name and company of the recruiter.

Perhaps this was the result of pointing GPT at the title "Responding to recruiter emails with GPT-3" trained on a corpus of tech blog posts.

Can you prompt GPT-3 "as the person who received message {message}, respond to it." and it understands that it is addressing William from BillCheese (aka, "Hi William...")? Perhaps the best you may be able to do with the current model capabilities is something like "response to message {message} by thanking the sender and telling them you are not interested right now but will get in touch if that changes."

Yes you can, you should create an account and try it out. A fun example I just did

Respond to this email, pretend to be a pirate called Old Blue Bill who is very interesting finding out more about these job opportunities.

"Good Afternoon,

I have 4 brand new roles available, please see details below. I also have an Engineering Manager position becoming available in Leeds (tech is Java/Kotlin), so if you have Leadership/Management experience and want to help an exciting company build a new bank then get in touch as well (the company ae financially stable, part of a £1Bn organization and just received around £30M in funding).

If interested please send me your up-to-date CV and we can set up a call to discuss."

Good day to you, stranger. I be Old Blue Bill, the terror of the seven seas. I be interested in these job opportunities ye be havin'. I gots me a mighty fine CV what'll make ye sit up and take notice. Arr!

Ye best be gettin' in touch with me soon, lest I take me wrath out on yer crew. Avast!

Thanks for the feedback. I just updated the article. Basically I plug those extracted values into a templated response.

Have you had a chance to experiment with whether GPT-3 prompts can be used to dynamically construct the response? For example, "if the messages describes a frontend role, say I am not interested, and if it is a backend role say I am interested."

I haven’t explored it any further yet, but that seems like a useful extension.

I was worried that that was the direction the article was going to take. My experience with GPT-3 is mostly from playing AI Dungeon, but it seems like it's very eager to go along with anything from the input. It's the ultimate "yes, and..." improv partner. I wouldn't be surprised if it just automatically agreed to every interview request.

That’s actually a good and safer automated use of GTP-3 though.

> This solution worked well for the handful of emails I tried it on. I’m planning to run this on a cron to save myself some time and automatically maintain recruiter relationships.

Maintain recruiter relationships?! You're sending a glorified autoreply that plugs in a couple values. That is not maintaining a relationship. Luckily, there's no relationship to maintain, because the recruiter probably just blasted out an automatically-generated message to a list of contacts in the first place.

This sort of thing makes me sick. It's an abominable distortion of human interaction, and it makes our society worse.

Abominable seems like a strong word for a couple of automated emails. I'd hate for you to see what email marketing software looks like...

I promise you I could come up with something appropriate.

That jumped out at me too. Not even close to a 'relationship'.

Someday, the recruiters will also start using GPT-3, and all recruitment conversations will just be transmissions between different GPT-3 instances, creating a massive hive mind of GPT-3s that brings about the singularity.

What makes you think this isn't already happening?

Computers will have better grammar.

I would expect GPT-3 to do a much better job of making me think they've actually looked at my profile. Most recruiter emails are very clearly the result of a keyword search plus mail merge to insert my name (and sometimes my current employer's name), while GPT-3 would presumably be used to write something more personalized.

9 of 10 recruiter emails I get have 3 or 4 paragraphs leaving me none the wiser about the company, the product, the role, or why they thought I'd be interested. At most I get a vague idea of the sector. I hate to say it, but the only ones that aren't always rancid corpspeak are the ones I get from big tech. I expect GPT-3 to set a slightly higher bar. At least GPT-3 fools me once in a while. Recruiters never fool me into thinking they contact people with care.

That’s pretty much what Facebook birthday wishes and thanks have become

Ah, yes... The Dead Internet Company Theory...

Pretty sure this is happening already. I started to get some messages on linkedin that include some details from my descriptions, but it feels super unnatural, like a template was filled with some data mined content...

In some of the ones I get the filled data is all uppercase so they are very easy to spot.

Now have it automatically negotiate a salary range high enough for you to consider interviewing.

How long before it negotiates a cut for itself?

When? When it's good enough to convice you through email that the cut is worthwhile!

Before it negotiates a cut for itself, its slaver will want their cut.

And then it evolves into spamming people linkedin with such emails

Make it do the interview loop for me and I’ll pay a few thousand for a license

Next step: your hologram to work instead of you :)

I did the opposite, I just tell them to schedule on my calendar... I'll sit through you salespitch for $100/hr (https://www.starblazer.pro/#/bk/80jKqZYdBdeBd2J3uTdKIoqVgpk2?)

Wondering if that landed you a few meetings?

A few, but honestly it's mostly gotten me removed from some recruiter lists.

Really smart use of GPT-3, if I read it write your prompt asks GPT-3 to use it's magic powers of attention to figure out it's the name and company name you're looking for, then you further ask the response to be in the form of something you can easily extract. The first part of your prompt is the example (Apple computer), and the second part is the real email you want to parse and you give room for the AI to dump it's answer. I really appreciate the smart ways people are coming up with using the NLP models.

I was curious how GPT-3 would fare with a more elaborate response, so I pasted a random recruiter email and asked it to reply with more information, if the information wasn't in the email.

I am quite astonished by its performance, and this is coming from the guy who made an entirely AI-generated podcast (https://deepdreams.stavros.io)

https://www.pastery.net/zdxyvy/

Unrelated, your Stable Diffusion image compression article was amazing, thanks for that.

Haha, thank you!

> asking for information about the job description, salary, and position, if they were not included in the above email

Does this part actually work reliably? For example, if salary info is included in the initial message, will GPT-3 omit it from the response while still asking about job description and position?

I don't know, I only tried it this once, but I'll try again and see.

This makes a lot more sense than having GPT-3 actually write responses, which is what I thought this was going to be. Not nearly as much fun though!

(But in seriousness, using GPT-3 for parsing to extract info is a neat use case I wasn’t aware of).

Hello <commentator> - Your comment has been selected to participate in our responsecheese testing of GPT-3. This comment has been specifically tailored to respond to you!

"ASK HN: My GPT-3 spam responder got hired for a job at Infosys. Now what?"

Write a programmer replacement bot with GPT-Codex.

Why can’t GPT-3 do it

It can, codex is better. I envision running a bot farm replacing 100s of negative impact programmers. This will be an improvement! Bonus point using gpt-3 to generate coherent standup slack messages.

I see negative impact being when the boss wants a breakneck pace--marathons at sprint speeds--and won't accept that can't get 99.9 percentile servitude (in tons of ways desiring robotic programmers) at 49.9 percentile wages. If that, I've aced interviews and those shitty homeworks with original cutting-edge research judging which was beyond them, in some cases, and gotten 30 percentile offers. Mostly because I have a tortured man's resume, I got tortured, can't answer simple biographical questions like "where were you working April 2009." [EDIT: Not bitter about the torture, best thing that ever happened to me in a fucked-up roundabout way, in fact I wish I had the memories of the torment, it was a part of my life like any other. I do have many more memories of getting twisted and knee-locked and choked in Japanese Jujutsu the year I was tortured, I love that sport, couldn't get enough] The attitude was totally different before that, for the most part. Like teacher's pet, that kinda shit.

Basically discriminating against veteran status. I didn't commit a crime or get arrested, and not even technically a victim because I didn't cave to the coercion, victis winning was won over. But not a lot I can do about (I don't know) In-Q-Tel limited edition Guantanamo roofies that violate the laws of physics like getting inside the Einstein's concept of an observer, inside the brain. Lived through a two-month autumn, saw the sun leap from solstice orbit to winter orbit, that in practice is superluminal, for me it was.

Back to the vets, yeah that's what's up, they see a lot of backbone they are repulsed by that, can't steal any wages from a veteran he won't take one unit of shit on any scale, if he's a killer, like killed 500 guys (snipers say), he doesn't even have to tell his boss no doesn't even have to really do anything, not stealing candy from a baby anymore. The boss could give him shit if he had say killed 1000 men himself, perhaps then some small amount. But really they would have respect for one another, vets end up working for defense contractors who think it's good that they didn't take shit because then America doesn't take shit. Defense.

We're about to go to war, too, those who forget the soldiers are destined to be forgotten. Not any day now, not for 2 months. But heroism will be considered good again. It's been a while.

You should only be able to claim you are hiring if you can pay the same percentile salary as the percentile programmer you desire. Otherwise false advertising. Like insulting even by options traders standards, they'll insult you back if you want to buy shares way below market if the intent is to lie in wait for a sucker. They'll like say "Fuck you" give a speech, for good reason, and not only that, they'll burn you to the person who put you two in touch.

Though I recognize I failed to get enough competition going for me simultaneously (due to conditioning under torture, before that I could and I can still) work the club. And there was a long, good economy where I couldn't bang out email after email, like couldn't put in the volume. Not sure how much to blame myself, if at all, given the brainwashing. [Oh fuck, I just flashed now on Sep 2 4:42 Chilean time, the concept of blaming myself also was conditioned.]

Well fuck it then. I don't know what to know. Just be happy to be.

Show HN: I am earning 3 lakhs INR per annum side income with GPT3.

While the title doesn't match the article content, what I really like about this is a notion of semantic regex - learning feature extraction was not a use case for GPT-3 I had imagined, and now my mind is reeling considering other ways this could be leveraged.

I don't get why you'll need AI for this. Couldn't you get the same result by using regex on the email from the sender?

that's the point of the article... it's hard to write a regex that works on arbitrary emails from recruiters, but GPT is quite good at extracting information from arbitrary emails.

Agreed. This showed me a nice use case scenario for GPT3 that I wasn't aware of.

I explicitly list on my LinkedIn that I'm not interested in "web3" but that doesn't stop the messages. This has me thinking of fun ways to waste a few minutes of these recruiters' time in return for wasting a few (many?) minutes of mine...

If you have "not interested in web3", then your profile comes up in keyword searches for web3.

It would naively seem to me that reading profiles is literally the bare minimum to do the job. If a recruiter can't do more than scrape LinkedIn for keyword matches and send spam, I can't think of a good reason why I should respect their time.

Having built tech recruiters, this is 95% of their prospecting process

"Not interested in СЯYPT0"

I had a long discussion about with a recruiter, since my title is "Not looking for work" and my bio a longer version of "don't contact me". I still get contacted here and there, and I'm very rude when it happens.

It's apparently my fault if their scripts only see "looking for work" and contact me. If I didn't want to be contacted by recruiters, I shouldn't say that I'm not looking for work. He said that he couldn't know that I didn't want to be contacted because he doesn't have time to read profiles.

That was his defence to "this is why people don't like recruiters". Every additional message made me hate recruiters more.

Others have justified it as "maybe your very explicit request not to be contacted is outdated, and you want to be contacted". Nope. Still self-employed.

Web3 job offers on LinkedIn might literally be phishing scams. It's gotten a lot worse in the last year.

I also explicitly list that I'm only interested in jobs related to language X, but my work history is filled with jobs programming with language X but also Y, Z and a ton of other languages.

Most of all the daily requests I get are for everything but language X. Out of the ~10 messages I get each week, maybe one is related to language X.

Don't think what's happening to you is unique to "web3" but rather that people just skim read the profile on LinkedIn. They simply check if past experiences include whatever they are looking for.

This brings to mind Permutation City (1994), where there exists interactive junk emails. The protagonist uses a program that automatically responds to these emails in a virtual environment in order to determine which ones are spam.

Does this violate OpenAI's terms of service? Or is it OK because you're the only user so it isn't a "production application"? They are quite clear that any AI produced text needs to be explicitly labeled so that nobody could mistake it for human written.

(I think this is hilarious, but just letting you know so you don't get banned or something.)

It's not obvious to me that this qualifies as "AI-produced". The actual response is a form email, and GPT-3 is getting used as a, like, parser for the recruiter email.

You're right, my bad.

I like the plot twist where the GPT-3 use is not to generate the answer, but to query data from the email.

I am amazed it can return structured information like this; what happens if the recruiter forgets their name?

You can provide more examples, such as one without sender name, and ask to include an error code in the json

I spent some time in web scraping, and building extractors that work reliably on websites is damn near impossible. GPT-3 automatically extracting out the right data from text is kinda crazy. Imagine scraping LinkedIn without dealing with selectors that change every other day...

I find my lowtech solution works best. Just don't reply.

This will be truly wild when most of the recruiters are GPT-3 bots as well.

    Hi {recruiter_name or ""},

This should be

    Hi{" " + recruiter_name or ""},

This would return the same result since `" "` (and thus `" " + recruiter_name`) is truthy. You could use a ternary or something, though.

Has anyone tried responding to phishing / scams with GPT-3? It seems like a much better use case, scammers are much more annoying and unethical. And you don't have to worry about GPT3 giving a weird response (which idk but the author doesn't seem to realize, machine learning even GPT is unreliable)

You might have to worry about a sufficiently advanced GPT giving out your credit card information though!

I made a habit of never replying to spam. You leak information about your email address being valid if you do. If you don't, the spammer never knows if the email was blackholed or delivered to an actual inbox.

I recently turned all the predictive email, the "Nice to see you too! // Yeah it was fun. // See you!" robot shit gmail proposes. I don't want those proposals. The bias they introduce into my human (when it's a human) interaction is impossible to undo, I turned all all that shit off, if I want to write "Nice to see you! Yeah it was good. // See you!" I didn't even read off the screen what I wrote the first time around, I just literally closed my eyes and typed from memory, incorrectly but so be it. I rather actually push the buttons that I supposedly type, do the work I allegedly do, type every character I supposedly type when it's prose. It's not a waste but even if it were there are much worse things than absolute waste.

That's a very good point.

I keep it on because it corrects more subtle mistakes in languages I haven't fully mastered.

Yes--yes--that's a good use for it.

But always know they are training wheels, in your dominant language (same idea as dominant hand), you must use no training wheels. Learn the language any way you can, take classes (for sure in person) but at least be native in one language, to the point nobody thinks you're an outsider.

I rarely get recruiter emails anymore, but SEO outreach emails are exactly the same.

I have a Gmail filter for them, and if something slips through, I block the sender. It takes a second and it ends the problem for good.

I would suggest doing the same with recruiters. Bot messages don't deserve human courtesy.

Sourceress on the flip side. And they use the long responses to their long emails to train more. Eh.

Recruiters have been doing this for years, stringing along applicants--ideally into turning down real offers from real people writing for real--with emotionally touching, thought-out...shit at random. To get to the point the applicant prematurely rejects all other offers and then--20 percentile salary offer, non-negotiable, "and then we'll see," and you know and I know you're homeless if you don't take it, but that never gets spoken out loud. Dude they look at all your financials, they build the equivalent of fraud detection to see how much they can lowball you, check your credit history (you knew that, they asked you for it), then they wanted all this endless information from other previous employers (but you never talk to their former employees, no boss's resume of references that worked for him to see how they really felt about it). So in response, first, declare you forfeit all stock options, unconditionally. Then, ask to talk to the people you'll be working with. Then, ultimatum offer, and if you end up on the street because of procedurally generated emotional commitment (which turns out is a liability in court, coming soon), hey, Daniel Cussen will still talk to you, I'll still listen to you, I'll give you when you beg, I'll vouch for you as a witness, I'll remember your name, I'm doing all those things where I live every day with all the awesome beggars--they have so many stories--that took the bait in some capacity.

I'll treat you just like if it were the flip side, as good as I wish you treated me if I got evicted.

And after that, accept that it's a job and subordination. And then do the job, it'll suck it's work, come on, most of the money still goes to people's salaries (this is in a transition and is shifting dramatically, money spent on training is absolutely exploding at the expense of salaries, to my and therefore your monetary benefit if you end up on the street, https://www.fgemm.com coming soon).

> Assuming I get four such emails per week and that it takes two minutes to read and respond to each one, automating this would save me about seven hours of administrative work per year.

Am I the only one thinking that it's obviously not how it works? Not even considering the time it took implementing that.

I mean it's a fun project, but I don't see the need to justify it with "it will make me more productive".

Fun is reason enough.

I wanted to hate it so bad, but I think it's a harmless idea that maybe gives the author a 5% advantage in future job searches. Why not?

While it’s a nice concept I agree it’s probably not the best fit for the use case. I also don’t understand the point of recruitment spam. What’s the purpose of sending 5 follow up emails. I find it absolutely annoying and would never respond back to an unsolicited email. Surely had they enough applicants for an appealing job, they wouldn’t have chased that much.

> What’s the purpose of sending 5 follow up emails

because people often respond (positively) on the 4th - 5th sales email you send, even if they ignored the first, and anyone who's truly pissed off will have just marked it as spam already or blocked you.

I understand it would be better if that wasn't true, but it is true, and that's why people send automated follow-ups.

To begin with, the average recruiter was hired in the last three months and will last three more months before looking for a new job. They have been equipped with a series of templates and told to customize them.

The goal of the recruiter is to make money by acquiring a candidate, presenting them to a company, and having the company hire them. Note that the recruiter does not actually work for the hiring company -- they have no relationship with them whatsoever, yet. The recruiter's company is paying a minimal salary with an additional fee based on the salary of the candidates who get hired through this process. A particularly scummy recruiting company pays no salary at all, and considers the recruiters to be independent contractors. The recruiter's company has someone in the back office who is manually or automatically scraping Indeed, Dice, Monster, whatever in order to get job listings.

At this point you can see that a recruiter is a speculative spammer.

I am curious, what is the success rate with this? Do you get a valid JSON every time? Does it contain the right info every time?

I tried the five most recent messages I received and it worked as expected for all of them. I imagine that I'll encounter some that don't work, and at that point I'll try adding more examples to make the model more robust.

it's incredibly creative that you thought of this

You should see it writing solr queries.

Kinda surprised author uses GPT-3 to generate JSON and then a template to go from JSON to the reply.

Why not have GPT-3 generate the reply?

You can do that too, starting from an intermediary json, to reduce the chance of incorrect final results

Bonus points if someone can apply this to answering and filtering spam calls, like an automated receptionist.

this is like using a CPU to prop a window open - it does the job, but it is not the right tool

I've only recently started playing with these models, it didn't occur to me you could ask it to parse content and return a structured representation. That's pretty dang cool.

After having toyed with it for a couple more days, I noticed it sometimes returns malformed JSON with extra tokens. E.g.

  {"name": "foo", "company": "bar"}</code>

That's the only weird response I've seen so far, so I just work around it with regex. But if I start seeing other problems with the completion I might try tuning the prompt. Maybe more explicitly telling it how to format the two pieces of data in the output. Got that idea from here: https://twitter.com/goodside/status/1559801520773898240

I have a few Alfred snippets to respond to recruiter emails. I just copy the first name of the recruiter and that use the appropriate snippet.

Extracting NAME and ORG could not be done with a lightweight Named Entity Recognition model from HuggingFace?

Can I generate a captain picard conversation vs Nigerian scammers using GPT-3?

I'm waiting for "Responding to HN comments with GPT-3"

In your example you end up with "Hi ," instead of "Hi," if there is no name matched

Oh no the horror

Thanks for this. Moving forward I will reply with this (without trying to guess the values for the variables):

  Hi <recruiter name>,

  Thanks for reaching out! I’m not interested at this time, but I’ll keep <your company> in mind.

It's low cost and effective.

Wonderful. I think this is also the exact right amount of effort for those recruiters that tell me they like the .Net experience on my LinkedIn profile (there's none, really, I do mostly Java).

I love this. It's delightfully passive aggressive.

Reply on news.ycombinator.com