Sep 29, 2026 路 4:24 PM UTC

117
1,220
30,197
612,347
Sort replies: Relevant Recent Liked
Replying to @_devJNS
7
33
1,431
20,927
this is crazy 馃槶鉁岋笍
28
6,718
Replying to @_devJNS
Deepseek exists bro
2
115
15,633
tf would deepseek do?
2
56
12,033
more replies
Replying to @_devJNS
many open weight models will definitely build you most illegal apps, but you need to self-host.
5
1
359
13,889
Replying to @_devJNS
Qwen 27B Uncensored exists for a reason.
1
4
891
Replying to @_devJNS
but Qwen 27B Uncensored can :)
188
Replying to @_devJNS
Chinese know the pain so they are busy launching open source uncensored models. No one should feel left out
81
5,660
Replying to @_devJNS
You and I have a very different idea of localhost
49
4,462
Replying to @_devJNS
Jailbroken opensource models will actually so um your wrong
47
4,760
Replying to @_devJNS
We already have pretty powerful local LLM models, which are getting uncensored by community to have 0 refusals. You need special rigs with a lot of VRAM to run best models, but if you're taking it seriously, it's worth it. Having no rate or tokens limit is huge.
3
39
6,195
Replying to @_devJNS
If you can't jailbreak ai, just say that
1
29
3,706
Replying to @_devJNS
Since Claude and Codex banned you, and Grok still has a long way to go...
1
12
6,058
Replying to @_devJNS
GLM 5.3 abliterated wya
1
8
3,422
Replying to @_devJNS
Uncensored Kimi and GLM will.
6
1,484
Replying to @_devJNS
Can't we do with the help of prompt injection ?
5
4,669
Replying to @_devJNS
Oh it will, Claude just need some elite gaslighting
4
852
Replying to @_devJNS
The trick is to make a similar but legal software with AI, then make it illegal manually. Not life advice. I would not do this.
1
5
489
Replying to @_devJNS
Abliterated model: are you sure about that?
3
2,382
Replying to @_devJNS
Open weight models exist bro
3
1,811
Replying to @_devJNS
Abliterated Qwen will, but don鈥檛 break the law kids
3
1,551
Replying to @_devJNS
AI won't create it's own backdoor馃拃
5
2,493
Replying to @_devJNS
Learn to get Ai to build them for you.
3
1,922
Replying to @_devJNS
Or we just find a way to compromise the AI鈥檚 values
1
859
Replying to @_devJNS
Factually wrong and its actually living in the twilight zone. Bro i had ai cracked for 2 years now, i havent done anything because ive been afraid of misusing it, and being branded like some robin hood vigilante. This is a retarded take. Come back to reality.
2
3,148
Replying to @_devJNS
i have zero respect for your courts or laws
1
868
Replying to @_devJNS
I guess my bro hasn't heard of uncensored models 馃ゲ
1
319
Replying to @_devJNS
Sorry, but open source uncensored AIs will build illegal apps, as they do now.
1
603
Replying to @_devJNS
abliterated models: sup
1,617
Replying to @_devJNS
Try uncensored models
1
2
2,360
Replying to @_devJNS
I regret learning how to code, yes it helps me know about things when it comes to prompting, but 90% of the time i just get another ai to prompt for me anyway.
1
622
Replying to @_devJNS
The clip is from where?
1
1,691
Replying to @_devJNS
Finally! A logical defense! Its refutable but at least it makes sense and something you can have a conversation around.
2,360
Replying to @_devJNS
Meanwhile yemen people were using claude to create rockets 馃槀
1
1,484
Load more