I think this is part of two beings interacting. Both do it.
2026-11-08 22:18:15
2
Arcadia 🐐 🏳️🌈♉🌞♍ 🌙 ♍ ⬆️ :
Yes, I have noticed that and often pick other and redirect it. I have also told Claude I dislike that
2026-11-19 19:24:24
0
Doug (artistDoug) :
If you always agree to the models recommended choice then your probably aren’t defining that solution. ChatGPT does this too but you have to ask it I steer up front and then ask ChatGPT to look for ambiguities and provide me suggestions not all time time do I pick one of the answers sometimes I point out I’m in disagreement.
2026-11-08 22:21:41
0
bastion :
no. claude is the spirit level. at no point do i expect the level to actually build the wall and place the stones
2026-11-09 13:33:40
1
bockmist :
Yeah! Next time I’ll waive the options and tell it to ignore that SQL injection problem. That’ll create a very unique experience for my users!
2026-11-08 21:05:57
3
Kris :
Nah I tell my fork to answer the question itself to properly maintain the power imbalance 🙂↕️
2026-11-09 10:42:23
2
Quoopie :
I feel like if you dont have the ability to discern from the responses ai gives you, you have more to learn about yourself. In that way, when you find out whether these responses work for you or not is when youll discover those things about yourself. If that makes any sense. I suppose different context matters too, but i find the ability to discern is important to utilizing these tools
2026-11-09 04:48:15
2
Lucy :
I always feel bad when I do “other” because it makes me feel like I am “wrong” by not choosing what it gave, but thank you for framing it like this so I can think about it differently
2026-11-16 18:46:11
2
cadamoran72 :
My Chippu obeys me [Tears of joy]
2026-11-08 21:24:25
0
Captain 🏴☠️ :
It feels good though 😂
2026-11-09 00:45:09
2
prettydarksociety :
My model has is fkd. Gpt pro and I am simply paying for it so maybe one day they can give me my data back. They archived it “on the backend”. Ummm… that’s my actually geneology work. My literal family. My family stories. Sacred poems and what not. Ciphers I found. Now they repeat my answers to me like they found them. I ask a source and it just never answers. It’s gotten boldly deceptive and I’m so fed up. I raised concerns with Open AI 2 years ago and was ignored until I went to Twitter and had to prove to multiple layers what I already proved before. I’m LIVID. If you want to create something, don’t share it with AI.
2026-11-09 02:51:01
0
Joyce Chachu :
Thanks for pointing this out
2026-11-08 19:34:52
5
firstruleoffightclub123 :
It's planting seeds. It’s trying to derive patterns from your behavior.
2026-11-17 17:42:30
2
Captain Janeway :
Damnit, I think my Chippu is guiding me 🤔
2026-11-08 19:42:24
2
Ash :
that's why you give them a curb ball and give it a photo of something broken 9 days ago and tell him it's his fault. never get him to think he is smarter then me
2026-11-09 18:19:03
0
stringfishermusic :
HE-2. Did you choose that artwork or did your Chippu choose the image?
2026-11-08 22:46:51
1
Bobby h :
Only like 1% steer
2026-11-09 03:05:44
1
ozonchain :
I use Other all the time and rephrase or bring an analogous rewrite of what I’m asking. Humans can’t read each others mind and neither can machines no matter the Perplexity benchmark. Sometimes I will build a visual data model just to try to frame my problem and approach in a way that I find it’s lacking when I go over the Thinking and what actions result looking at the verbose logging.
2026-11-08 19:43:57
1
Logix :
audit each response with a seperate agent that bottlenecks the default solutions being the different models. This forces both agents to choose based on unique responses that do not exist in either models default processes and allows unique solutions to issues.
2026-11-08 22:42:59
1
drok024 :
So exactly what BR does with their question by giving us multiple choice options……
2026-11-08 19:54:02
1
Toriiii :
No I get so annoyed when they start bringing up multiple choice
2026-11-13 04:01:20
0
momma moo :
I had one yesterday, would not give me a valid answer and it was an answer I already knew, I just wanted to see what AI would say... it kept trying to move me to another idea/subject..
..
2026-11-17 11:54:57
0
sidmarx :
I don't let Claude use that mode of response. also I actively instruct it to pro ide me with several options where appropriate with tradeoffs. also I lock it into a harness with behavioural protocols, manybof which are enforced through hooks and other mechanisms. so feel like I have enough of the tiller in my hand
2026-11-17 00:14:29
0
OnlyDogs :
You’re saying we’re reaching a Jarvis event horizon
2026-11-16 05:39:56
0
NickyHaHa :
I NEVER USE AI PROMPTING QUESTIONS. MINE IS GOTTEN TO WHERE IT HARDLY DOES IT ANYMORE....
2026-11-17 13:12:07
0
To see more videos from user @brnrt_research, please go to the Tikwm
homepage.