4 ms·

OP here. It's a question to easily detect for people who use chatgpt to fill the form. Anyone who enters alpha-gal syndrome has probably used chatgpt. That isn'

OP here. It's a question to easily detect for people who use chatgpt to fill the form. Anyone who enters alpha-gal syndrome has probably used chatgpt. That isn't an automatic reject though (very generic template answers are). Guess we'll have to find another question soon now this is out here for LLMs to slurp up! The reason for this questionnaire is culture fit. It's also the very practical matter of weeding through 2500 applications when AI has made all straightforward code-based questions very pointless

Chaosvex 2y ago
It's not about a single question, it's about all of them that don't do anything to further the process. They're a waste of time for people who have options - in other words, the people you want to hire. If a company wants me to jump through hoops just for the shot at an interview, I'm not doing it because I can go afford to go elsewhere. I'm sure I'm far from alone in that.
MattGaiser 2y ago
Anyone who has done a Google search of "meat allergy" will also put alpha-gal syndrome. For a question like this to work, there should be a much wider variety of possible answers.
mystraline 2y ago
Simply put, you can complain about "nobody wants to apply to my jobs". But you ask terribly invasive questions, questions that are just frankly garbage, and your AI landmine questions. And this is 100x worse applications than other jobs. And you admit that if you happen to know about alpha-gal (tick related disease prevalent in the USA) then you probably are using an LLM?! Seriously? If this is how you treat candidates, how do you treat employees? How many garbage 'tests' do you put them through?
- ChaitanyaSai 2y ago
  Here's the problem we are trying to solve for. Filter from 2500+ applicants. What would you suggest. And we do not want to rely on credentialing and leetcode filters. What non-garbage tests would you suggest? Please note that we are talking about an initial test that's generous enough in error, but not so much so that we end up with 1000 people to seriously interview.
  - throw-qqqqq 2y ago
    I honestly suggest YOU put in the work to filter the applications instead of trying to get the applicants to do it. You will lose confident candidates otherwise. I sure as hell would quit halfway through and find other options. IMHO you risk getting more “desperate” candidates by requiring them to answer all these questions. I understand your motivation 100%. My company also gets at least 10x as many applications now compared to before ChatGPT, and by far most are garbage. This is OUR problem though, not the applicants'!
  - mystraline 2y ago
    There is an answer, but I 100% believe you wont like it, nor implement. Do video interviews, or better yet, in-person interviews. Ask questions there. Request sharing desktop, and share if you ask LLMs and your strategy in doing so. Academics are dealing with the same problem, with "do they know the material or is it an LLM?". Universities also have access to a potent tool: oral examination. Its also the hardest to implement due to time and 1:1 labor. The answer's right there.
  - dogleash 2y ago
    > What would you suggest. Hard work. From you.
tw19837 2y ago
The first Google result for "Someone who eats meat often has an allergic reaction" is alpha-gal syndrome. There are two kinds of people who would know this without searching: Medical professionals or people who obsessively follow every health scare in the press, often the yellow press. This question is useless, unless you want to hear "I googled the question" or "I don't know". Anyway, most people don't care what you want to hear, hence the low application numbers.

Chaosvex 2y ago

It's not about a single question, it's about all of them that don't do anything to further the process. They're a waste of time for people who have options - in other words, the people you want to hire. If a company wants me to jump through hoops just for the shot at an interview, I'm not doing it because I can go afford to go elsewhere. I'm sure I'm far from alone in that.

MattGaiser 2y ago

Anyone who has done a Google search of "meat allergy" will also put alpha-gal syndrome. For a question like this to work, there should be a much wider variety of possible answers.

mystraline 2y ago

Simply put, you can complain about "nobody wants to apply to my jobs". But you ask terribly invasive questions, questions that are just frankly garbage, and your AI landmine questions. And this is 100x worse applications than other jobs. And you admit that if you happen to know about alpha-gal (tick related disease prevalent in the USA) then you probably are using an LLM?! Seriously? If this is how you treat candidates, how do you treat employees? How many garbage 'tests' do you put them through?

ChaitanyaSai 2y ago
Here's the problem we are trying to solve for. Filter from 2500+ applicants. What would you suggest. And we do not want to rely on credentialing and leetcode filters. What non-garbage tests would you suggest? Please note that we are talking about an initial test that's generous enough in error, but not so much so that we end up with 1000 people to seriously interview.
- throw-qqqqq 2y ago
  I honestly suggest YOU put in the work to filter the applications instead of trying to get the applicants to do it. You will lose confident candidates otherwise. I sure as hell would quit halfway through and find other options. IMHO you risk getting more “desperate” candidates by requiring them to answer all these questions. I understand your motivation 100%. My company also gets at least 10x as many applications now compared to before ChatGPT, and by far most are garbage. This is OUR problem though, not the applicants'!
- mystraline 2y ago
  There is an answer, but I 100% believe you wont like it, nor implement. Do video interviews, or better yet, in-person interviews. Ask questions there. Request sharing desktop, and share if you ask LLMs and your strategy in doing so. Academics are dealing with the same problem, with "do they know the material or is it an LLM?". Universities also have access to a potent tool: oral examination. Its also the hardest to implement due to time and 1:1 labor. The answer's right there.
- dogleash 2y ago
  > What would you suggest. Hard work. From you.

tw19837 2y ago

The first Google result for "Someone who eats meat often has an allergic reaction" is alpha-gal syndrome. There are two kinds of people who would know this without searching: Medical professionals or people who obsessively follow every health scare in the press, often the yellow press. This question is useless, unless you want to hear "I googled the question" or "I don't know". Anyway, most people don't care what you want to hear, hence the low application numbers.