The most prestigious law school admissions discussion board in the world.
|
 |
 
|
Can AI detect lies?
| https://imgur.com/a/o2g8xYK | 05/14/25 |
| https://imgur.com/a/o2g8xYK | 05/14/25 |
Poast new message in this thread
Date: May 14th, 2025 7:40 PM Author: https://imgur.com/a/o2g8xYK Serious question. Right now the AIs are trained to do the opposite. They are trained to help you amplify whatever you present to them as the truth. If I lie and say I'm a car salesman trying to sell a Toyota to a rich lawyer, the AI never even considers the possibility that I'm a rich lawyer shopping for a Toyota. If you asked the AI "did you detect any indicators in our discussion that I was lying" the AI would have to say no, wouldn't it? (http://www.autoadmit.com/thread.php?thread_id=5725071&forum_id=2...id.#48931410) |
Date: May 14th, 2025 7:44 PM Author: https://imgur.com/a/o2g8xYK What about this: suppose I make a 3D-printed bouquet of flowers, then photograph it with my AI phone. Suppose the AI has a "flower enhance" algorithm built into it. Won't the algorithm go nuts trying to enhance my "flowers?" Or does the AI have a bullshit detector that looks for anomalies before it triggers the other algorithms? Is that even possible? How would you train it? (http://www.autoadmit.com/thread.php?thread_id=5725071&forum_id=2...id.#48931430) |