will probably do on a future iteration! The current job categories where jus the ones could come up with and then used the openAi api to check for conformance (or not) of each post to it :)
simonw is the best kind of spammer- he brings his creation into the discussion- but he does so in a way that enriches the value of what is being discussed.
His tool [Datasette] is of course such that it often is immensely handy at dissecting data in useful ways so is often exactly on point for a discussion on HN...
But sir, now that mega corps have monopolized attention on the internet and charge enormous fees in my humbles you should really rethink what it is you are trying to say?
Personally I cant find anything anymore, google is trolling me, people are afraid to link to the homepage of their subject matter. People foolish enough to make websites have zero visitors because they fear linking to their own shit. LOL
I cant name anything you've made and you can't name anything of mine. If you didn't mention it I wouldn't have known the tool was that shameful self promotion - now forbidden on the internet?
Ill be quiet now or else it will turn into an article rather than a low effort monologue and who would want to read that?
What's the AI usage here? From the raw JSON data, it seems you wrote a prompt to an LLM to extract structured data from the Who is Hiring comments, although I am not sure if that counts as an "AI filter" since the filtering criteria are explicitly defined beforehand.
right, I'm feeding each post to several queries to the openAi API. I guess I put "AI filters" so people knew this is actually curated by post content and not just a contains() filter so you get posts with the text "we don't do remote!" when you select the remote checkbox
Not knocking the approach, but how do you do quality control on the posts? Are you just spot checking? How often have you found bad data?
I've thought about doing something similar (using ChatGPT to structure and categorize unstructured data) for a different project in a completely different space and I'm worried about ChatGPT hallucinating things, especially when it comes to numbers.
The quality control is a good question, and one that can probably be addressed using evaluation as taught by some of the deeplearning.ai short courses (1).
I made an interactive resume ai bot on my personal website and there is an instance where I can ask it "tell me about your intel experience" and it added in C++ as one of the languages, but that is untrue. I had done C++ at a different company.
I posted the short courses just as answer to how to address quality control. I'm not selling anything, and those courses are free anyway.
deeplearning.ai was cofounded by Andrew Ng, who is probably the most well known for his work on teaching machine learning through deeplearning.ai, Coursera, Stanford, etc. He has taught and influenced millions.
In regards to "evaluation", I think these is what those short courses will cover:
Self-Evaluation with the LLM: The idea is to use the language model to generate an answer and then use the same or a different model to evaluate that answer. The evaluation could involve asking the model to rate the answer's accuracy, coherence, relevance, or any other desired metric. This self-evaluation process can be automated and scaled, although it's important to be aware of the limitations, as the model might inherit biases or blind spots from its training data.
LangChain for Structured Evaluation: LangChain can be used to structure this self-evaluation process. It can orchestrate the flow where the LLM first generates an answer and then follows a series of steps to evaluate it. This might include breaking down the evaluation into specific questions or tasks that the LLM must perform to assess its initial response.
Well to be fair, the original who is hiring post doesn't do much quality control.
Then, the other apps do neither.
Honestly, this whole thing came out just of my frustration using one of those and filtering for Remote, reading the text and finding out it wasn't remote at all.
As for quality control, there's a step for categorization that returns some tags. Posts that don't match any are rejected, that's kind of filters for relevancy.
I think it might be worthwhile adding a little info box on the website explaining that the data was gathered this way. I was wondering the same about AI usage and only just found the answer after reading hrough the whole discussion.
Btw. nice tool!
When providing a huge list of technologies, structure them somehow. Alphabetically ordered, for example - I shouldn't have to Ctrl-F to find my preferred programming language.
you're absolutely right, will get it on a next iteration. Just for now, know you can filter by your tech on the technologies list. Most people are looking for either remote or a specific location, and after choosing one of those honestly there are not that many posts left to sort through.
I think that's what is being noted. That there's a huge list of technologies, and it would be nice if they could be sorted by some criteria (alphabetical, old->recent tech, grouped (all these are Javascript based), categories (compiled, interpolated, data science, graphics/imagery, ...)
Also, is there a way to put in prior "Who is Hiring?" dates? If people keep putting out the same listing again and again it would be nice to have a way to find. Totally in the Nice to Have category.
Gosh now I get it! thanks for the clarification, looks like I didn't get the point. Totally right.
The search bar does filter the filters list. So if looking for "rust": 1. expand the technologies list 2. type rust in the search bar 3. select rust option.
But yes, should probably sort it though.
As for old posts, I just overwrite the old data and modify the month label just out of laziness. I Will keep posts history in future iterations!
It's built to be pretty fast by not pulling data it does not need. Since it operates in multiple passes on stored data, it would be easy to modify/add a pass to get what you want. Feel free to use parts you like.
A couple of things I think would help:
- sorted attributes (too hard to go through a hundred computer technologies to find Rust)
- multiple geographic entries for the same name (multiple entries for Germany, USA, UK, Europe)
- ability to select a month
- if you are showing a static pull of the data, the ability to refresh some month would be helpful
Tangential, but this is why I still love Ruby's Enumerable module; it has `#select` and `#reject`, as well as the `#filter` method that people from other languages are used to.
No, I can’t. Why do you ask? My point was that it's much nicer to have method names that do what they say instead of the programmatic version of the word "biweekly."
It would be really nice to add salaries. I remember there was a past effort where people would actually downvote listings that didn't show salaries. Now it's required in California (though plenty don't, including Mistral....). Honestly, it just saves everyone time and helps keep the market open. Maybe we should also restart the pressure? The template should be updated to reflect this (I swore it used to exist). And then allow us to filter by salaries.
I commented the same elsewhere. It’s a 10K fine per listing and makes you liable for civil action in certain circumstances. I’m surprised HN even allows posts without salaries, as you’re in violation if anyone from CA could potentially fill the role (e.g. a US-based remote position).
> Companies hit by wave of suits over job ads lacking pay info
> A batch of proposed class actions filed in Washington state against major companies such as adidas AG, Home Depot Inc., and Marriott International Inc., will help employers gauge the litigation risk they face under a recent nationwide wave of pay transparency laws requiring salary ranges in job ads.
The wheels are starting to turn. I think there are more class actions happening than individual cases but IANAL.
>Key legal questions will include whether the job applicants can show they suffered any injury to merit an award of damages
Further down, the article links to another about some actual enforcement in Colorado, but the fines are an absolute joke; clearly so insignificant that the Bloomberg writer avoids giving a number for any single suit.
You’d probably have to do a public records request with the Department of Industrial Relations to get an account of all the enforcement actions.
While I agree with your point about organizing and striking, California has some of the country’s most powerful unions and labor protections, even for non-union industries. For example, most software developers in CA can’t be classified as exempt if they’re paid under 115K and can sue for quite extensive damages in the event of misclassification:
This amount is also indexed to inflation and increases every year. This does not, however, prevent employers from posting jobs on HN below this amount, in contradiction to the law, which I’ve seen from time to time.
Is the a minimum employment size? Because Mistral doesn't seem to have them listed. But they are small in terms of employees. But they are listed as hybrid paris/london/SF. I don't think that allows them to get away. I even checked using my VPN in case weird shit was going on. I like how they are being open source but not being open salary seems counter to that ideology... There's a few other AI companies I've seen that have this issue too so I assume there's exceptions or no enforcement and people gambling.
(I'm only calling out Mistral because they are "big" and the open source messaging they push)
I also want to know who is truly remote. So many times a listing says remote from anywhere but then click through to the JD and it’s remote from cali or remote from USA.
Why would a listing say 'Remote from anywhere' if it's really USA? Is it just showing 'remote' and only the clickthrough claims where?
True 'remote' from anywhere in the world is tough, not just for working in a team but for taxes and payroll. If it's not a F500 or billion dollar company, you can't really expect them to support working anywhere in the world.
Would be nice to get a link directly to a post. F.e. links are a bit malformed due to formatting and wanting to see an original post, I had to ctrl+f over the whole thread with parts of the sentence.
57 comments
[ 2.2 ms ] story [ 131 ms ] threadNot everyone here is a developer!
This JSON file has the annotated data: https://github.com/bernawil/hn-who-is-hiring/blob/main/src/H...
Since it's JSON on GitHub you can explore using Datasette Lite like this:
https://lite.datasette.io/?install=datasette-pretty-json&jso...
Here's an example of a custom SQL query: https://lite.datasette.io/?install=datasette-pretty-json&jso...
His tool [Datasette] is of course such that it often is immensely handy at dissecting data in useful ways so is often exactly on point for a discussion on HN...
Personally I cant find anything anymore, google is trolling me, people are afraid to link to the homepage of their subject matter. People foolish enough to make websites have zero visitors because they fear linking to their own shit. LOL
I cant name anything you've made and you can't name anything of mine. If you didn't mention it I wouldn't have known the tool was that shameful self promotion - now forbidden on the internet?
Ill be quiet now or else it will turn into an article rather than a low effort monologue and who would want to read that?
Thanks
https://californiapayroll.com/blog/californias-new-pay-trans...
Better idea #2: Unionize and strike because historically nothing else has ever resulted in enforced universal workers' rights.
I've thought about doing something similar (using ChatGPT to structure and categorize unstructured data) for a different project in a completely different space and I'm worried about ChatGPT hallucinating things, especially when it comes to numbers.
I made an interactive resume ai bot on my personal website and there is an instance where I can ask it "tell me about your intel experience" and it added in C++ as one of the languages, but that is untrue. I had done C++ at a different company.
1. https://www.deeplearning.ai/short-courses/
https://en.wikipedia.org/wiki/Andrew_Ng
In regards to "evaluation", I think these is what those short courses will cover:
Self-Evaluation with the LLM: The idea is to use the language model to generate an answer and then use the same or a different model to evaluate that answer. The evaluation could involve asking the model to rate the answer's accuracy, coherence, relevance, or any other desired metric. This self-evaluation process can be automated and scaled, although it's important to be aware of the limitations, as the model might inherit biases or blind spots from its training data.
LangChain for Structured Evaluation: LangChain can be used to structure this self-evaluation process. It can orchestrate the flow where the LLM first generates an answer and then follows a series of steps to evaluate it. This might include breaking down the evaluation into specific questions or tasks that the LLM must perform to assess its initial response.
As for quality control, there's a step for categorization that returns some tags. Posts that don't match any are rejected, that's kind of filters for relevancy.
Great idea, just needs some UX love.
Also, is there a way to put in prior "Who is Hiring?" dates? If people keep putting out the same listing again and again it would be nice to have a way to find. Totally in the Nice to Have category.
The search bar does filter the filters list. So if looking for "rust": 1. expand the technologies list 2. type rust in the search bar 3. select rust option.
But yes, should probably sort it though.
As for old posts, I just overwrite the old data and modify the month label just out of laziness. I Will keep posts history in future iterations!
https://github.com/hughdbrown/who-is-hiring
It's built to be pretty fast by not pulling data it does not need. Since it operates in multiple passes on stored data, it would be easy to modify/add a pass to get what you want. Feel free to use parts you like.
A couple of things I think would help:
- sorted attributes (too hard to go through a hundred computer technologies to find Rust)
- multiple geographic entries for the same name (multiple entries for Germany, USA, UK, Europe)
- ability to select a month
- if you are showing a static pull of the data, the ability to refresh some month would be helpful
Edit: This is not snark, I simply do not wish to work in the AI sector, which seemed over-represented in the most recent Who's Hiring thread.
Is it? Has even a single party been charged? And if so, what was the enforcement process?
Without unions that strike (key word: strike) workers have no power.
Edit: the $10k is a maximum fine, not a minimum fine. In other words, it's a complete joke.
> A batch of proposed class actions filed in Washington state against major companies such as adidas AG, Home Depot Inc., and Marriott International Inc., will help employers gauge the litigation risk they face under a recent nationwide wave of pay transparency laws requiring salary ranges in job ads.
The wheels are starting to turn. I think there are more class actions happening than individual cases but IANAL.
https://news.bloomberglaw.com/daily-labor-report/adidas-home...
>Key legal questions will include whether the job applicants can show they suffered any injury to merit an award of damages
Further down, the article links to another about some actual enforcement in Colorado, but the fines are an absolute joke; clearly so insignificant that the Bloomberg writer avoids giving a number for any single suit.
Organize or perish.
https://news.bloomberglaw.com/daily-labor-report/tesla-schwa...
While I agree with your point about organizing and striking, California has some of the country’s most powerful unions and labor protections, even for non-union industries. For example, most software developers in CA can’t be classified as exempt if they’re paid under 115K and can sue for quite extensive damages in the event of misclassification:
https://www.dir.ca.gov/oprl/ComputerSoftware.htm
This amount is also indexed to inflation and increases every year. This does not, however, prevent employers from posting jobs on HN below this amount, in contradiction to the law, which I’ve seen from time to time.
(I'm only calling out Mistral because they are "big" and the open source messaging they push)
https://jobs.lever.co/mistral/675b7f06-a76b-4144-af0c-4dd328...
True 'remote' from anywhere in the world is tough, not just for working in a team but for taxes and payroll. If it's not a F500 or billion dollar company, you can't really expect them to support working anywhere in the world.
I guess because people assume there are only seniors on Hackernews?
I could imagine that being the reason and it would be pretty silly...
> USA (4)
> US (3)
> US (1)
> UNITED_STATES (1)
> Europe (5)
> Europe (18)
> EU (1)
> Canada (8)
> Canada (1)
The filters need a bit of consolidation.