Opening up a white collar crime consultancy firm

@StaleDonut @RustyCrab if you did want to try fraud like that though now is probably the best time, even if you did get caught by the auditors that didn't get fired if you donate to anyone political beforehand they'll just pardon you

@StaleDonut @RustyCrab Someone would have to pay the taxes and file to the IRS and whatever state they are in on behalf of the fake person. I'm no tax auditor so I have no idea how they could actually either do that or get away with it but I assume unless you're the size of Deloitte the IRS will find and pound you into oblivion for both the PR and the obvious fraud.

I doubt it would be possible without the company actively fudging things when it's probably much easier to do an H1B.

@StaleDonut @RustyCrab Would that not raise a massive red flag in the IRS or something similar?
Good thing all the auditors keep getting fired/laid off to prevent anything like that from being found out if it was even possible.

@RustyCrab @PNS @rw @soapone LinkedIn is owned by Microsoft and yes, it is sometimes a hard requirement on applications.
A lot of those "easy apply" applications are done by playwright or equivalent scripts and are not actually eligible whatsoever (anecdotes on places like the recruitinghell subreddit mention those jobs get thousands of applicants sometimes where 99% of them are either not qualified or not legally allowed to work in the country of the posting [Indian]).

@waifu

what if you applied for a job you're massively over qualified for and they just send this back

@Junes suicide

nuking my entire vscode install some keybindings didn't work the way they were supposed to despite removing all of them so I decided it would have been better to delete everything

@fixpoint yes, it only took them 3 weeks

I am now professionally unemployed 🎉

@mischievoustomato @pernia For that, I was using ollama and open-webui.

@mischievoustomato @pernia Well the 3n I just downloaded doesn't support image recognition unless I'm doing something wrong

@mischievoustomato @pernia Thank you, I will try that one and see if it's less horrid, though horrid is funny

@mischievoustomato @pernia deepseek doesn't support image input so you can't go @gunk is this real

@mischievoustomato @pernia It's the best model that takes images and runs on the 8 GB GPU in my server. I'm trying to find the least bad model to use as a fedi bot because funne

@waifu @zero @phnt @Junes @mischievoustomato @sendpaws the software is doing what it was intended to do, block scrapers. Is it perfect? Hell no, but I can attest to it working on improving the uptime of git.froth.zone

@phnt @Junes @waifu @sendpaws @zero @mischievoustomato Another good point in the default configuration. Someone who was smart would change the bot policy to be more aggressive and PoW everything, which you can do. Someone being DDoSed by bad actors would do that (or just do the HAProxy linked earlier, both valid choices)
https://anubis.techaro.lol/docs/admin/policies

@mischievoustomato @Junes @phnt @waifu @zero If you try hard enough you can bypass nearly anything, but is it worth it for what little you probably get in return? Probably not.

@mischievoustomato @Junes @phnt @waifu @zero they could, yes. They could also make their user agent something absurd, but they don’t. Why? Because it would look suspicious as fuck if they didn’t. Does that matter to some random git repository or open source wiki? Fuck no. Even the local LLM I’ve been messing with says that:

AI scrapers use normal browser user agents to:

• Evade anti-bot detection.
• Render dynamic content properly.
• Ensure compatibility with APIs and servers.
• Mimic real user behavior and reduce the risk of blocking.

However, this is just one part of a broader strategy (e.g., proxy rotation, request throttling, and behavioral mimicking) to avoid detection.