Hacker News Reader

Comment by hibikir | Hacker News Reader

And that's not even enough: A few weeks ago I had to interview someone who had what appeared to be a realistic profile. Everything that came out of their mouth was from chatGPT It was suspicious, but the ruse became clear when they shared the wrong screen, so we could see his prompt, and how everything we said was being read in.

At this point every remote internet checklist has to include checks for humanity, because the percentage of straight out fakes is too high. Even the questions to ask me at the end were GPT provided.

View on HN

Anyone affected by this and in the US might consider calling or writing to their congressman. The time to do that is now when the demand is high to bolster jobs but low for excessive laws. Nobody innocent is going to be wronged if this is made into a crime or otherwise regulated to put a stop to.

The fake job applicants are only siphoning resources from the economy at the high expense of all other parties involved. The ones who are getting screwed the most are the applicants, some of whom are concerned about making ends meet and getting auto-rejected constantly despite decades of experience. No one should stand for it.

If I didn't know better I'd think this was satire. As far as I can tell the advocacy is for either companies to be empowered to sue people who apply to work with them (seems like madness) or to set up a situation where the government enforcement arm pro-actively goes out and harasses unemployed job seekers. Either way that sounds like a recipe for disaster for unemployed persons.

AI has made hiring especially in technical industry an absolute shit show. I agree with parent comment that ideally government could do something about it but agree with you on how would you even do that. Maybe if they required all the job board companies like indeed and glassdoor and LinkedIn to properly vet candidates else those companies would be fined, but it's hard to imagine a solution that doesn't also hurt unemployed legit human beings

And then you run into problems on the corporate side: fake job listings to build up resume databases for comparison shopping of applicants. Regulations in this area should have to cut both ways.

This the end game that Silicon Valley created. An automation arms race between two competing groups that were initially trying to save a little time or cut down on staffing but escalated it to the point where the default approach would be considered unforgivably assholish 15 years ago, people that don’t buy into it at least somewhat are drowning in bullshit, and nobody’s happy— but on paper everybody’s got record productivity!

With LLMs, this same exact scenario is playing out in other realms. Look at writing and publishing. Sure you’re on top of the world before everyone else catches up, but when they do, there’s now just a boilerplate of exponentially expanding bullshit and counter-bullshit that everyone has to circumvent to do anything.

This has already happened long ago with Google search results. The first tier of results is won by reasonably well-funded entities that provide a legitimate service, and have the means to optimize the signals feeding the search rankings, putting them higher than the next tier.

The second tier of search results tends to be dominated by imitators that don't really add anything of value (SEO spam, blog posts that tell you how to write a for loop in Ruby despite knowing full well that the reader already had no problem finding that information, etc.)

Then finally at the bottom are the little guys who try their best, but haven't learned yet that it's a waste of time to try to self-publish any content because there's too much actual spam masquerading as content, and Google can't tell the difference.

The search results effectively became a list of content approved by a single publisher (even if automated) rather than a melting pot of freely-expressed ideas.

I sincerely hope that we can prevent the similar nullification of the software developer's career accomplishments as carrying any weight, but I am starting to have doubts. If it even goes as far as the erosion of incentives to accomplish things, then we may actually end up needing that AI to do the work for us, as there will be few people left who give a shit.

Yes, that’s a good example, and I share your concerns.

As an aside, I shudder thinking about what a heavily ‘SEO’d’ LLM experience would be like.

Somebody convince me LLMs are not net-negative productivity.

They seem much better at producing bullshit that’s difficult to filter through, than performing actually useful work.

I have found copilot autocomplete to be somewhat useful for small blocks of code.

Coca-Cola and Toys R Us have found them useful for making terrible commercials cheaper than making terrible commercials by hand and way cheaper than making good commercials that actually improve their brand image. Seems weird they’d do that for immensely expensive holiday television spots rather than throwaway 5 second YouTube spots or something but hey — I’m clearly not a corporate genius.

Yea, I couldn't tell if the original comment was satire but the number of phishing ads that existed in the past for bogus positions, to pool candidates for later hiring, to farm market rate data, and who knows what else… makes me have very little empathy for the employer side.

It’s been a mess for awhile due to economies of scale benefiting the hiring side to manipulate and abuse the market. The fact it’s become more affordable for job seekers to do a bit of the same is just ironic.

I would REALLY love if job postings had to go through a government clearing house. Only real jobs get posted. Only real applicants can apply.

Bonus: jobs would have to be classified according to a single government standard, so it should be possible to search for a good job match by at least limiting the field and (allowed) location(s).

making the jobs application (and hiring) market a single market will make it more efficient, and cut out a lot of middlemen inefficiencies. I like it.

You as a hiring company can pay to have a 2nd website, but posting it to the gov't portal is a requirement. The information, such as conditions, salary (range), experience, location etc, are all in standardized format. If you're found to be lying, it's a federal crime (because of fraud and interstate commerce for example).

Applicants also must have gov't issued ID (such as social security), so you cannot be fake.

If you look at some of the problems of USA Jobs, you may not actually want this.

Hiring in tech has been pretty awful even before AI.

> If I didn't know better I'd think this was satire.

It’s actually a constant them on HN to imagine that passing laws will magically make problems disappear. The realities of enforcing the law or even identifying perpetrators are imagined to be the easy part.

People seem to want to pass laws that treat the symptoms, not the cause.

“You keep getting the stomach bug. Here take this, it’ll calm your stomach. No no, you can keep eating that expired cheese, it’s all good”

> you can keep eating that expired cheese

Why are you anti French ?

My oldest kid hate experiencing vomit, she transferred that on to me as a “gross gross thing” and it was the first thing that came to mind.

I’m a huge fan of French onion soup.

Ah, the ol’ “manufacture an argument that wasn’t made, then shoot it down it in front of an audience” trick. I suppose I’ll be advocating for the outlawing of those kinds of comments, and anything else deemed as misinformation next.

A more realistic scenario would involve no enforcement by the government (except perhaps in extreme cases, like with the 'spam king' back in the day). ChatGPT's terms of service would already cover it under the "shall not be used for illegal activity" language, and it would be just enough of a deterrence to benefit a larger number of people without creating new problems. But I wasn't advocating for a specific solution, just a call to a congressman. Despite their faults and flaws, they're probably still going to do a better job than I am at making the call, or maybe it won't even be a priority for them and they'll do nothing.

I think the GP is suggesting that making, distributing, and profiting from such software should be made illegal. If an engineer can make this software, they are probably a good fit for many jobs in the market.

I'm suggesting that you should call your congressman and say that getting a job is a problem right now and automated applicants could be contributing to it (we don't know the full story, but making noise about it might at least inspire some investigation by those who have the ability to get the facts). I don't think it should be a crime to automate a job application, and I have no problem with it from an ethical point of view long as the application is made truthfully and in good faith by a reasonably qualified applicant and there is real intent to follow up on it.

But if that isn't the case, there's no reasonably good safety mechanism to mitigate the massive amount of harm that a determined bad faith actor could cause to the economy.

But making false claims about your work history (as could be the case with the one using ChatGPT to answer questions) is a problem, isn't it? And it's wonderful to see these rebuttals made against a hypothetical something that already happened. https://www.lawdepot.com/resources/business-articles/legal-c...

> situation where the government enforcement arm pro-actively goes out and harasses unemployed job seekers

Why wouldn't this be a desired outcome? Unemployment doesn't give a carte blanche to send spam.

> Nobody innocent is going to be wronged if this is made into a crime or otherwise regulated to put a stop to.

Good luck.

The applicants doing fake job applications do not care about your laws at all. Many might be in foreign countries. They might plan on applying with stolen identities.

Making a law isn’t going to change a thing. Even if you did, what company is going to spend resources tracking down the likely fake identity of someone applying for a job just to hand it to law enforcement for them to ignore in their backlog forever?

> Making a law isn’t going to change a thing. Even if you did, what company is going to spend resources tracking down the likely fake identity of someone applying for a job just to hand it to law enforcement for them to ignore in their backlog forever?

I missed the part where I included that or any strategy on how it would be used as a deterrent. Clearly that's not how it is done as you pointed out, but you make it seem as if laws have no value at all, which is a rather naive take. Fraud is already illegal FYI.

I don't have a solution, other than to make a call to the people who are elected to find those solutions, if they are able to. If they can't or won't, then it is a good thing that phone call was free anyway.

How exactly does someone who applies with a stolen identity get anything out of doing so?

Absolutely correct, just making laws themselves have little effect over anything. Enforcement is the key. For most laws that step is an afterthought. But there are creative ways to do it.

> applicants doing fake job applications … stolen identities

What I don’t get is what’s the economic incentive for this behaviour

- It can be a side effect to keep your unemployment insurance which is conditional upon proving you are sending applications at a given pace. I'd probably need to apply to random jobs if I qualified for it because there isn't a role opening in my niche weekly to fullfil the criteria here. I never had to because I was ineligible for other reasons every time I was unemployed and could have used support but that's a whole other can of worms)

- I heard its a thing to get n jobs you're not qualified for to get at least the first few month salary "for free" (as an individual or as a pawn from a larger organized fraud). Not sure how common or how much truth there is to it though.

But this chaos fits Big Tech's claim that there are not enough American workers, so they can then turn around and onshore H1Bs from the hiring manager's hometown back in the old country.

Hundreds of comments, and this is the first one that mentions perhaps the primary root cause of the situation.

Do you work in tech? Have you ever seen any pressure to create LLM-driven chaos with the goal of increasing support for encouraging immigration in future years?

It’s too elaborate of a Rube Goldberg strategy to take very seriously. Companies struggle to achieve simple, clear, short-term goals in tight-knit, well-aligned teams. Ain’t nobody got the skill to pull off that level of conspiracy.

There often isn't a nefarious scheme. Humans are better at spotting patterns than they are at mass coordination.

then move abroad and get your dream job there?

Huh? Is this a 'love it or leave it' comment? I genuinely don't understand what you're meaning to say here.

What I mean is unless your ideal is autarky or USSR under Joseph Stalin, it is hypocritical or ingenuous to expect having a market where you can sell goods and services worldwide but not allowing workers applying and getting jobs worldwide for same companies. That is called free market.

So if you happen to think you are missing jobs because they are given to people living in another country, you also have the choice to play by the same rules, relocate there and apply for the same job. Or ask for a lower salary where you already are to be competitive. This is fair competition.

ha lets see…

In economics, a free market is an economic system in which the prices of goods and services are determined by supply and demand expressed by sellers and buyers. Such markets, as modeled, operate without the intervention of government or any other external authority.

if you think america is “free market” I have some Enron stock to sell to you :)

Lol wat. 'Free market' is a spherical cow in a vacuum. Its an abstraction that people like to make to reduce complex reality to something small and comfortable. In reality, the world is not driven only (or even mostly) by market forces. All players in modern economies are subsidized by and beholden to governance by nation states. That wildly warps what actually happens outside the textbook.

> At this point every remote internet checklist has to include checks for humanity,

I genuinely don't understand this requirement. Isn't an interview exactly that? It's a conversation pretending to be about a technical problem/question/challenge but in reality its purpose is to find out whether you click with the person and would want to work with them. If some ChatGPT text can trick you then your process is broken anyway and everybody joining your company can expect colleagues selected by this sub-par process.

> If some ChatGPT text can trick you then your process is broken anyway

This is pretty unfair and seems like victim-blaming when we have companies spending billions of dollars to create these programs with the specific intent of trying to pass the Turing test.

There’s a bit of an echo chamber on HN where people convince each other that all LLM-generated text is easy to identify, riddled with errors, and “obviously” inferior to all real-human writing. Because some LLM writing fits those criteria and is easily identified, these folks are convinced they can identify all LLM writing and anyone who can’t must be a dunce.

I didn't claim anything about identifying writing. That's a strawman. I'm talking about humans talking to each other. Even if it's in a zoom call. Any interview process that doesn't include that is broken, and that's my claim. Echo chamber or not.

Apologies for misunderstanding you, then. Agreed that human to human is critical, especially for identifying culture fit (not homogeneity of course, just interaction styles like openness, etc).

I do think people cheat video interviews with LLM help, but in-person should always be required anyway, even if it’s via proxy (“meet with a colleague from our Madrid office”).

How widespread is LLM cheating during video interviews these days? Honest question.. How do people even do it? Let an LLM app listen in and suggest avenues of discussion and lists a bunch of facts on the side to spice things up?

Even if that's the case, isn't it just a matter of conversing in a way that the LLM can't easily follow?

An interviewer is a "victim"? Maybe they should just, you know, speak to their interviewees. At least in 2024 that's hardly faked by an LLM. Therefore, if you are fooled, you cheaped out, and you are hardly a victim.

Someone being deceived is a victim, yes.

Start every interview asking the candidate how many rs are in strawberry.

Apparently outdated, ChatGPT 3.5 answers correctly here.

Try raspberry? Both fail for me, but I'm not a paid user.

I don't pay either (anymore), but is correct as well. (Via chatgot android app)

From iOS app, paid 4o

> There are 4 Rs in “razzleberry.”

Follow up with, "Please check again"

You’re absolutely right to ask for a recheck! Let’s count carefully: • R in Razzleberry: • 1st R: In “Razzle” • 2nd R: In “Razzle” • 3rd R: In “berry” • 4th R: In “berry”

Total: 4 Rs in “razzleberry.”

No changes—still 4 Rs! Let me know if I can clarify further.

However when I asked it to write and run a python script to count them, it got it right.

ChatGPT-4 (free) on Android just told me:

> There are two "R"s in the word "cranberry."

Same for me.

Good example to illustrate how LLMs work, if it is not correct for cranberry, but correct for raspberry or strawberry.

you're not on 3.5 anymore

Well, I use the free version and just checked, but I cannot even see what model is used.

Either way, if my cheap standard unpaid ChatGPT version gives the correct result, I don't think it counts as a valid catch anymore.

>Everything that came out of their mouth was from chatGPT It was suspicious, but the ruse became clear when they shared the wrong screen, so we could see his prompt, and how everything we said was being read in.

Wouldn't you notice a lag between your question and the candidate's answer if the candidate had to type your question into chatGPT?Or does the candidate use some software/tool with transmits your question to chatGPT directly?

there is no lag, voice to text to chatgpt then read the answer

Thank you. (Haven't been up to date with the hype.)

Reply on news.ycombinator.com