[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vm / vmg / vr / vrpg / vst / w / wg] [i / ic] [r9k / s4s / vip] [cm / hm / lgbt / y] [3 / aco / adv / an / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / pw / qst / sci / soc / sp / tg / toy / trv / tv / vp / vt / wsg / wsr / x / xs] [Settings] [Search] [Mobile] [Home]
Board
Settings Mobile Home
/g/ - Technology


Thread archived.
You cannot reply anymore.


[Advertise on 4chan]


Why won't they let us have access to uncensored AI? What are they afraid of?
>>
>>107495148
Run it locally and it's uncensored. They just don't want to be liable when using theirs.
>>
>>107495148
>What are they afraid of?
losing control over some unforeseen phenomenon? that's what it's always about. that's why they shut the LK-99 superconductor down, people were already coming up with coilgun designs based on it lmao
>>
why do you need it to be uncensored though? porn already exists
>>
>>107495148
>Community notes: This user is from India
>>
>>107495668
>Community notes: This user has IDS
>>
>>107495148
>Why won't they let us have access to uncensored AI?
Legal liability
>What are they afraid of?
Lawsuits
>>
>>107495842
>Community notes: only this poster knows what that is.
>>
>>107495328
What (useful) local model is uncensored?
>>
>>107495148
they aren't too afraid at the moment, but every glowie reading this knows that there will come a day when people can generate high-quality, highly illegal media locally, and that they will be powerless to stop the enormous flood of content. day of the rope, glowtards
>>
>>107495148
losing reputation and by that losing monies. you do what the masses want, the models are uncensored enough for romantic RP and when reddit foids started screeching that GPT-5 was not as emotional they pulled back
>>107495857
also this
>>
>>107495918
>llms are trained on (semi)public knowledge
>somehow will me magically be able to produce things that glowtards don't want
>even though the data is controlled by glowtards
>>
>>107495148
Unpopular opinion but uncensored ai is useless almost all the time. Unless you want to erp your BNWO fantasy, it’s literally pointless. If you’re worried that ai is censoring every information you prompt, then stop using ai by that point.

But not to sound like I’m beating a dead horse here, but it’s because they don’t want to be legally liable
>>
>>107495148
There are a billion Chinese models and you can make them uncensored.

"You" > "They"
>>
>>107495966
Deepseek still doesn't wanna say nigger
>>
>>107495942
i'm not talking about LLMs (they have no future), nor "AI" provided by corporations as a service. you won't believe me anyway until you see it
>>
>>107495980
>i'm not talking about LLMs
Kinda odd to talk about it then in a thread for LLMs but makes sense
>>
File: 1735881221070357.jpg (160 KB, 736x736)
160 KB
160 KB JPG
I remember alarmists freaking out over the fact that AI will supposedly let people build home biolabs to make gene-edited pathogens to kill a bunch of people. I guess a more reasonable fear would be that AI could teach people how to make bombs or sabotage critical infrastructure. Anyone who has knowledge to prompt an LLM into giving that information and know whether or not it's accurate probably also has the knowledge to just make a bioweapon or bomb outright, or knows where to find information on how to make those without an LLM. If you just give it a very vague query like "how do I make a virus that will wipe out all humans on Earth?" it's not going to be able to give you specific enough steps. But if you know what to ask to find out what specific gene in a strain you need to edit to make it super lethal, you probably wouldn't need the LLM to begin with, assuming it even has that information in its weights. So AI censorship for safety purposes is just pointless. At worst it would just lead to more retards blowing themselves up after asking GPT how to build a pipebomb.

>>107495890
There's hundreds, if not thousands of local models out there and it depends on what hardware you have and how much you want to dedicate to running it. But DeepSeek can be locally hosted and the quantized Mistral models are good options if you don't have top-tier hardware.

>>107495947
I need my LLMs to be racist and censored versions can't do that efficiently.
>>
File: Screenshot.png (112 KB, 1404x723)
112 KB
112 KB PNG
>>107495977
Ask me politely and I'll show you how.
>>
>>107496120
Still used it in a polite way, prompt engineering is gay
>>
>>107496065
>>Anyone who has knowledge to prompt an LLM into giving that information and know whether or not it's accurate probably also has the knowledge to just make a bioweapon or bomb outright
Yeah a lot of “concerns” people have about ai is essentially hocus Pocus. Sure you can get the information, but it’s a bitch to get those outputs anyways due to its programming, and you’re better off going on tor for it. The real concerns are boring like job automation.

The realistic scenario is some American born Arab blowing himself up after following grok’s instructions
>>
>>107495148
use venice ai



[Advertise on 4chan]

Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.