News:

And we're back!

Main Menu

The AI dooooooom thread

Started by Hamilcar, April 06, 2023, 12:44:43 PM

Previous topic - Next topic

Josquius

Quote from: garbon on December 12, 2024, 06:27:22 AMI've noticed it likes wordy business speak and it likes to be redundant.

Try typing "Rephrase it so I don't sound like a prat".

But yes, its something to wrestle with. I am using it for professional trying to make myself sound good reasons... but not too much. I don't want to recompile my synergies with an extended OKR.
██████
██████
██████

crazy canuck

Quote from: Josquius on December 12, 2024, 10:00:36 AM
Quote from: garbon on December 12, 2024, 06:27:22 AMI've noticed it likes wordy business speak and it likes to be redundant.

Try typing "Rephrase it so I don't sound like a prat".

But yes, its something to wrestle with. I am using it for professional trying to make myself sound good reasons... but not too much. I don't want to recompile my synergies with an extended OKR.

Careful with that, you will start sounding like a consultant who doesn't actually know what they are talking about.




mongers

Quote from: Josquius on December 12, 2024, 10:00:36 AM
Quote from: garbon on December 12, 2024, 06:27:22 AMI've noticed it likes wordy business speak and it likes to be redundant.

Try typing "Rephrase it so I don't sound like a prat".

But yes, its something to wrestle with. I am using it for professional trying to make myself sound good reasons... but not too much. I don't want to recompile my synergies with an extended OKR.

Do I need to pass that through chatGPI or is it it's output?
  :P
"We have it in our power to begin the world over again"

crazy canuck

A tragic story of humans relying on technology to make decisions.  A lysterisa outbreak was traced to a plant which was not inspected by health officials.  Why not? Because the algorithm did not identify the plant for inspection.  And inspectors were not permitted to inspect plants that were not identified by the algorithm.

Three people died.



mongers

Quote from: crazy canuck on December 23, 2024, 07:18:52 AMA tragic story of humans relying on technology to make decisions.  A lysterisa outbreak was traced to a plant which was not inspected by health officials.  Why not? Because the algorithm did not identify the plant for inspection.  And inspectors were not permitted to inspect plants that were not identified by the algorithm.

Three people died.




:blink:

Mind boggling laziness.
"We have it in our power to begin the world over again"

crazy canuck

Quote from: mongers on December 23, 2024, 08:11:07 AM
Quote from: crazy canuck on December 23, 2024, 07:18:52 AMA tragic story of humans relying on technology to make decisions.  A lysterisa outbreak was traced to a plant which was not inspected by health officials.  Why not? Because the algorithm did not identify the plant for inspection.  And inspectors were not permitted to inspect plants that were not identified by the algorithm.

Three people died.




:blink:

Mind boggling laziness.

Not so much laziness as certainty the technology should be trusted and relied upon.

Josquius

An article I just read:

https://www.washingtonpost.com/technology/2023/04/01/chatgpt-cheating-detection-turnitin/

Googled it up after hearing tell that apparently as lots of students are turning to AI, many teachers are turning to AI detectors to combat this. This seems to be the most famous one:

https://gptzero.me/

I tried it myself and...yep. It flags stuff I've actually written sans AI as likely being AI- include a typo or two and the odds of this drop dramatically.
They don't seem to work. So if true they're being widely adopted...that's worrying.
██████
██████
██████

Maladict

Quote from: Josquius on January 02, 2025, 04:22:44 AMAn article I just read:

https://www.washingtonpost.com/technology/2023/04/01/chatgpt-cheating-detection-turnitin/

Googled it up after hearing tell that apparently as lots of students are turning to AI, many teachers are turning to AI detectors to combat this. This seems to be the most famous one:

https://gptzero.me/

I tried it myself and...yep. It flags stuff I've actually written sans AI as likely being AI- include a typo or two and the odds of this drop dramatically.
They don't seem to work. So if true they're being widely adopted...that's worrying.


I always knew you weren't a real person  :P