It looks fishy to me because they claim to be “under a massive NDA”, while leaking info. In that situation, you’d probably avoid mentioning any fluff that would help the business to identify you. I’m not sure though; it’s also possible they genuinely give no fucks about it.
It’s possible. I’d argue it’s less effective than it looks like, but you’re right, at least some would do it.
I’d certainly do the same if I were in that position.
Speaking on that. It’s a bit off-topic, but PSA:
In case anyone here is ever in a position that you need to publish something anonymously, pay close attention to your own writing style. This is essential to hide your identity; we leak a lot of info about ourselves not just by what we write, but how we do it: vocab, spelling, grammar, even discourse structure. Here are some good albeit slightly sensationalised examples of that.
Also, keep it as short as reasonably possible; the more you write, the more you tell others who you are.
Yes, prompting an LLM to phrase it is a good idea. Just be careful if the LLM you’re using isn’t “dialling home”, I’d definitively not trust OpenAI/Anthropic/Google/etc.
“I put in my two weeks yesterday” = “I submitted my resignation letter in the 1st of January”. Any delivery business that fits what they said, and concerned about NDA violations, could use this info to narrow down the list of potential suspects.
It could easily be aabout any food app anywhere in the world.
There are plenty clues around the text that he’s from USA. You could even narrow it down further to general area + age range.
I’d love to dig further into this due to linguistic curiosity, but in the chance the text is genuine, I’d rather not, I’d be potentially helping a corpo against a worker.
It looks fishy to me because they claim to be “under a massive NDA”, while leaking info. In that situation, you’d probably avoid mentioning any fluff that would help the business to identify you. I’m not sure though; it’s also possible they genuinely give no fucks about it.
The fluff could also be intentionally placed red herrings for anyone trying to identify them. I’d certainly do the same if I were in that position.
It’s possible. I’d argue it’s less effective than it looks like, but you’re right, at least some would do it.
Speaking on that. It’s a bit off-topic, but PSA:
In case anyone here is ever in a position that you need to publish something anonymously, pay close attention to your own writing style. This is essential to hide your identity; we leak a lot of info about ourselves not just by what we write, but how we do it: vocab, spelling, grammar, even discourse structure. Here are some good albeit slightly sensationalised examples of that.
Also, keep it as short as reasonably possible; the more you write, the more you tell others who you are.
This is a proper time to use an LLM. You want their generic sound.
Yes, prompting an LLM to phrase it is a good idea. Just be careful if the LLM you’re using isn’t “dialling home”, I’d definitively not trust OpenAI/Anthropic/Google/etc.
What info given here would pinpoint the guy? It could easily be aabout any food app anywhere in the world.
“I put in my two weeks yesterday” = “I submitted my resignation letter in the 1st of January”. Any delivery business that fits what they said, and concerned about NDA violations, could use this info to narrow down the list of potential suspects.
There are plenty clues around the text that he’s from USA. You could even narrow it down further to general area + age range.
I’d love to dig further into this due to linguistic curiosity, but in the chance the text is genuine, I’d rather not, I’d be potentially helping a corpo against a worker.