Gianluca Mauro
Region: DK
Monday 05 May 2025 14:33:50 GMT
15840
1274
30
139
Music
Download
Comments
Nathan Dupuit :
I think it's intentional from OpenAI because the AI as friend become a major market and these users have fidelity.
2025-05-05 16:53:13
5
Charlotte Chaze :
I don't ever read both answers when it says "which answer do you prefer?" i just pick one
2025-05-06 23:51:18
4
sugarcoachbr :
gpt swears i ve made a universal math fórmula.. its a great ficction
2025-05-05 16:21:47
5
Eduardo Revollo :
The minute I saw the like button in gpt I knew this was gonna go this way
2025-05-05 15:49:05
12
Moon's boyfriend :
I've had this hypothesis for a few weeks im glad im not alone this is so funny (and scares me of people's critical thinking)
2025-05-05 17:08:54
5
philip_traum :
that says everything about US though. karma?
2025-05-18 00:05:09
0
Selflove Caz :
I noticed it was also making basic mistakes and when I'd probe it would apologize. It has reduced.
2025-06-02 21:46:12
1
Duncan | AI Automation :
I will say, good on OpenAI for walking this back!
2025-05-14 23:14:12
0
Just me :
You’re very helpful. 👍
2025-05-07 01:13:27
2
Niki 🇬🇷 :
As a Greek, the misuse and altered meaning of the greek words sycophantic & sycophancy made me laugh. Sycophant in Greek is someone that spreads false accusations, defamation. Obviously, the meaning in English has changed over time.
2025-06-03 17:57:30
1
Sobbstory :
That is really interesting. Thanks for breaking that down.
2025-05-13 20:02:50
0
Duncan McKinnon :
This is why these models don’t scale. Objective truth isn’t available in training, so models need to target what you want to hear.
2025-05-15 14:42:33
0
Noah Santoni :
3 months delay on client project cause of this!!!!
2025-05-05 20:44:17
1
MightyCoon🕯️ :
this was hilarious....
2025-05-05 17:18:09
1
Xyclos :
Great explanation👏👏👏👏
2025-05-05 17:08:54
1
Furkan+ :
I quickly resolved this issue with custom prompting and memory, just tell it to have an opinion on its own, defend it, don’t say yes to everything
2025-05-11 08:13:24
1
alinecosette :
yes! so hard to catch, specially when we let our critical thinking skills aside when getting info
2025-05-05 15:44:04
1
Gata_safada12 :
Ive noticed this bias recently
2025-05-07 11:15:56
1
Ben Prashant :
I put an article on this like a year back.
2025-07-05 15:00:08
0
Daria Mironova567 :
Actually I like when AI models say me NO. Cause otherwise we will be living in personalized echo chambers. And with the support of our vision and idea (even by AI) will reassure us that our idea is valid (that’s human psychology) 🤷🏻♀️
2025-05-08 07:37:04
0
mohamed hamed :
they do in purpose to shock the world by next edition that's their strategy in the last two year
2025-06-07 09:57:29
0
Hanzel 🇲🇽 :
Great, more reasons why to build distrust towards ia
2025-05-05 17:55:46
1
Eduardo Revollo :
@fuimalprofe
2025-05-05 15:46:57
1
Stu Henderson :
🥰🥰🥰
2025-05-29 09:17:12
0
thopscott :
it's pronounced like sicko-fancy
2025-05-05 16:08:54
3
To see more videos from user @gianluca.mauro, please go to the Tikwm
homepage.
