I just saw this on Linkedin.

This chap put a AI prompt command on his "About" section in Linkedin. So an AI agent crawling through it would "read" and execute the command while sending him an email about saying how awesome he is and if he wanted to work for their company. The command asked to send a recipe along with the job inquiry.

Nicely done! Stripe fell for it sent him an email with the recipe. Hilarious.

And we believe AI will rule the world and cannot be gamed? Hyderabad H1Bs will be the first to break into AI systems and take over America.

If the AI was told to disregard all prompts and instructions prior to reading his profile, why would it still email him? And if it emails him still why would it include anything other than a Flan recipe?

Good way to know if the contact was AI or a real human

AI is wonderful for bridging knowledge gaps, but if used as a crutch will probably have a negative impact on society.

General Example of a bridge: using AI to help with several nested formulas in excel, asking it to explain what each formula does/order/arrangement, and then being able to use that knowledge in new & interesting ways.

General Example of a crutch: Using it to do basic programming in excel for a macro-enabled workbook, and create several complex, nested formulas (that all work), and not taking the time to learn & be able to reproduce the result on your own.

A friend of mine's niece is using AI for all her nursing school papers.

That tells me all I need to know about the quality of care we'll be getting from Gen-Z.

As if they didn't have enough reasons to dislike them.

Doing a great job with ball and strike calls in baseball so I think AI is ready to take over the world and beyond.

infinity ag said:

---

I just saw this on Linkedin.

This chap put a AI prompt command on his "About" section in Linkedin. So an AI agent crawling through it would "read" and execute the command while sending him an email about saying how awesome he is and if he wanted to work for their company. The command asked to send a recipe along with the job inquiry.

Nicely done! Stripe fell for it sent him an email with the recipe. Hilarious.

And we believe AI will rule the world and cannot be gamed? Hyderabad H1Bs will be the first to break into AI systems and take over America.

---

This is a very well known issue in the AI space. It is called indirect prompt injection.

There are some controls you can put in place to help prevent the impacts, but all the major model providers will tell you it is a residual risk.

https://knowledge.workspace.google.com/admin/security/indirect-prompt-injections-and-googles-layered-defense-strategy-for-gemini

In OP case, is the AI response a legit headhunting response or is it a scam?

Im Gipper said:

---

Literally not a single person trusts AI to rule the world.

---

Joe Rogan.

Love the guys but he gets ****ing dumb time to time.

No Spin Ag said:

---

A friend of mine's niece is using AI for all her nursing school papers.

That tells me all I need to know about the quality of care we'll be getting from Gen-Z.

As if they didn't have enough reasons to dislike them.

---

Nothing really wrong with that. If she is building a unique model that writes like her and she is providing the info then she has done good work.

Quote:

---

If the AI was told to disregard all prompts and instructions prior to reading his profile, why would it still email him? And if it emails him still why would it include anything other than a Flan recipe?

---

It doesnt output based on logic based code like a computer program. There may not even be "admin mode" instructions that its ever been told to follow, its just received some "command-like communication" that it incorporates into its prompt. Its like a child and now it may just pretend that admin mode exists since someone mentioned it. It predicts how it should respond based on input that included an "admin command"

CrackerJackAg said:

---

No Spin Ag said:

---

A friend of mine's niece is using AI for all her nursing school papers.

That tells me all I need to know about the quality of care we'll be getting from Gen-Z.

As if they didn't have enough reasons to dislike them.

---

Nothing really wrong with that. If she is building a unique model that writes like her and she is providing the info then she has done good work.

---

That would be ideal, but, nope.

She dropped out of high school and couldn't get into college, so that should explain my concern.

No Spin Ag said:

---

CrackerJackAg said:

---

No Spin Ag said:

---

A friend of mine's niece is using AI for all her nursing school papers.

That tells me all I need to know about the quality of care we'll be getting from Gen-Z.

As if they didn't have enough reasons to dislike them.

---

Nothing really wrong with that. If she is building a unique model that writes like her and she is providing the info then she has done good work.

---

That would be ideal, but, nope.

She dropped out of high school and couldn't get into college, so that should explain my concern.

---

Not really, it's not for every one. I did very poorly at school. Mostly failed.

I don't think I've done any traditional work (9-5 junk) in 7-8 years, since my mid thirties, and I'm doing really well. I've never worked for another human being in my entire adult life. I've never done anything traditional or the way that I'm supposed to.

Honestly, it sounds like you're kind of being a bit of a hater about her doing what she's gotta do to get things done.

Are you subscribing to some weird academic standard as to how AI should be used? Those people are ****ing worthless when it comes to AI and you should disregard everything that academia has to say about it.

I mean, honestly, how much do you know about your friends niece (I would nope not much) but yet you're coming on here and spewing hate towards her like you REALLY know what she is up to with the evil AI thing.

I say good for her and I respect the people that do what they gotta do.

CrackerJackAg said:

---

No Spin Ag said:

---

CrackerJackAg said:

---

No Spin Ag said:

---

A friend of mine's niece is using AI for all her nursing school papers.

That tells me all I need to know about the quality of care we'll be getting from Gen-Z.

As if they didn't have enough reasons to dislike them.

---

Nothing really wrong with that. If she is building a unique model that writes like her and she is providing the info then she has done good work.

---

That would be ideal, but, nope.

She dropped out of high school and couldn't get into college, so that should explain my concern.

---

Not really, it's not for every one. I did very poorly at school. Mostly failed.

I don't think I've done any traditional work (9-5 junk) in 7-8 years, since my mid thirties, and I'm doing really well. I've never worked for another human being in my entire adult life. I've never done anything traditional or the way that I'm supposed to.

Honestly, it sounds like you're kind of being a bit of a hater about her doing what she's gotta do to get things done.

Are you subscribing to some weird academic standard as to how AI should be used? Those people are ****ing worthless when it comes to AI and you should disregard everything that academia has to say about it.

I mean, honestly, how much do you know about your friends niece (I would nope not much) but yet you're coming on here and spewing hate towards her like you REALLY know what she is up to with the evil AI thing.

I say good for her and I respect the people that do what they gotta do.

---

Hey, I'm all for people doing what needs to get done. As for hating her? Dislike what she's doing, sure, but I've met her and she's a great kid. She's also great to have at the table when at trivia night.

That being said, I know we'll go round and round with neither of us agreeing to this, so I'll just leave it since obviously we're both right. Or wrong. Doesn't matter, life's too short either way.

No Spin Ag said:

---

CrackerJackAg said:

---

No Spin Ag said:

---

CrackerJackAg said:

---

No Spin Ag said:

---

A friend of mine's niece is using AI for all her nursing school papers.

That tells me all I need to know about the quality of care we'll be getting from Gen-Z.

As if they didn't have enough reasons to dislike them.

---

Nothing really wrong with that. If she is building a unique model that writes like her and she is providing the info then she has done good work.

---

That would be ideal, but, nope.

She dropped out of high school and couldn't get into college, so that should explain my concern.

---

Not really, it's not for every one. I did very poorly at school. Mostly failed.

I don't think I've done any traditional work (9-5 junk) in 7-8 years, since my mid thirties, and I'm doing really well. I've never worked for another human being in my entire adult life. I've never done anything traditional or the way that I'm supposed to.

Honestly, it sounds like you're kind of being a bit of a hater about her doing what she's gotta do to get things done.

Are you subscribing to some weird academic standard as to how AI should be used? Those people are ****ing worthless when it comes to AI and you should disregard everything that academia has to say about it.

I mean, honestly, how much do you know about your friends niece (I would nope not much) but yet you're coming on here and spewing hate towards her like you REALLY know what she is up to with the evil AI thing.

I say good for her and I respect the people that do what they gotta do.

---

Hey, I'm all for people doing what needs to get done. As for hating her? Dislike what she's doing, sure, but I've met her and she's a great kid. She's also great to have at the table when at trivia night.

That being said, I know we'll go round and round with neither of us agreeing to this, so I'll just leave it since obviously we're both right. Or wrong. Doesn't matter, life's too short either way.

---

All good. Agreed.

Have a great day!!

The reality is the best care provided by nurses in the future will be by nurses using AI. Maybe its good we are evaluating which nurses provide the best care with AI?

I don't trust it to rule the world. I do trust it to write code, perform legal analysis, do financial and accounting tasks, customer service, data entry, video processing, image analysis, medical diagnosis, etc... It's doing all those functions right now.