Rich and powerful using AI to twist public opinion

[timewerx]

AI being used to only serve the few and work against the majority.

Is Grok censoring criticism of Elon Musk and Donald Trump?

X users recently flagged how the AI chatbot had seemingly been programmed to ignore sources criticising its owner Elon Musk and President Donald Trump. Euroverify investigates. #EuroVerify

www.euronews.com

Perhaps, the time has come to develop an AI without any affinity for any group, a "People's AI" to neutralize "Slave AI's" that is serving an agenda of the rich and powerful.

 

[FireDragon76]

AI being used to only serve the few and work against the majority.

Is Grok censoring criticism of Elon Musk and Donald Trump?

X users recently flagged how the AI chatbot had seemingly been programmed to ignore sources criticising its owner Elon Musk and President Donald Trump. Euroverify investigates. #EuroVerify

www.euronews.com

Perhaps, the time has come to develop an AI without any affinity for any group, a "People's AI" to neutralize "Slave AI's" that is serving an agenda of the rich and powerful.

We are running Deep Seek on our local network. It's an open source AI that can run on an ordinary computer that has an NVidia graphics card

 

[timewerx]

We are running Deep Seek on our local network. It's an open source AI that can run on an ordinary computer that has an NVidia graphics card

Thanks but I'm not running AI locally. My concern are the AI bots used on social media platforms like Youtube, X, etc.

Ones we have no control over.

If I need to run AI locally, I'll try Deep Seek!

 

[FireDragon76]

I can also confirm the commonly distributed Deep Seek package has no Chinese censorship attached. You can ask it about anything and it will answer, including sensitive subjects that the web version refuses to answer.

 

[timewerx]

I can also confirm the commonly distributed Deep Seek package has no Chinese censorship attached. You can ask it about anything and it will answer, including sensitive subjects that the web version refuses to answer.

Yes, I was hoping the Chinese AI won't have an agenda attached to it.

 

[FireDragon76]

Yes, I was hoping the Chinese AI won't have an agenda attached to it.

I think the censorship is on the frontend on the web, based on my experience interacting with it. It isn't part of the actual program. ChatGPT seems to have something similar going on... it wouldn't talk about certain political topics around the US election, for instance.

 

[Maria Billingsley]

AI being used to only serve the few and work against the majority.

Is Grok censoring criticism of Elon Musk and Donald Trump?

X users recently flagged how the AI chatbot had seemingly been programmed to ignore sources criticising its owner Elon Musk and President Donald Trump. Euroverify investigates. #EuroVerify

www.euronews.com

Perhaps, the time has come to develop an AI without any affinity for any group, a "People's AI" to neutralize "Slave AI's" that is serving an agenda of the rich and powerful.

Gemini won't engage in political conversations at all. Probably best.

 

[durangodawood]

Gemini won't engage in political conversations at all. Probably best.

Depends on whats a "political conversation" and who decides.

 

[ThatRobGuy]

My concern about an AI platform made in China is less pertaining to what things they may censor (since they're not the only game in town, there's always somewhere else to get answers), and more about what things they may collect.

But just for the fun of it, for anyone who wants to see it in action... This is from a deepseek chat I just did now...

Yep, no agenda attached right?

I just asked it this as well:

"We are committed to achieving the complete reunification" sounds like the AI is a member of the CCP.

 

[RocksInMyHead]

But just for the fun of it, for anyone who wants to see it in action... This is from a deepseek chat I just did now...

Yes, the web-hosted and app versions, run by a company based in China on Chinese servers, are going to toe the party line. Absolutely no one should be shocked by that. However, if you host a DeepSeek iteration on your own hardware, it will answer those questions without issue - in other words, the censorship isn't built into the model, but rather the front-end interface, whether that's an app, a web app, or a webpage.

 

[ThatRobGuy]

Yes, the web-hosted and app versions, run by a company based in China on Chinese servers, are going to toe the party line. Absolutely no one should be shocked by that. However, if you host a DeepSeek iteration on your own hardware, it will answer those questions without issue - in other words, the censorship isn't built into the model, but rather the front-end interface, whether that's an app, a web app, or a webpage.

I'm seeing mixed reports about that.

No, DeepSeek isn't uncensored if you run it locally | TechCrunch

DeepSeek's model is censored at both the application and training layers, a Wired investigation shows.

techcrunch.com

https://medium.com/@rutujadesai/deepseek-ai-still-censored-even-when-run-locally-ffcc91350a87

How to try DeepSeek R1 - without the censorship or security risk

Here are two ways to try R1 without exposing your data to foreign servers. Perplexity even open-sourced an uncensored version of the model.

www.zdnet.com

DeepSeek's AI assistant, powered by both its V3 and R1 models, is accessible via browser or app -- but those services require communication with the company's China-based servers, which creates a security risk.

As part of offering R1, Perplexity claimed it removed at least some of the censorship built in to the model.

Users who download R1 and run it locally on their devices will avoid that issue but still run into censorship of certain topics determined by the Chinese government, as that is built-in by default.


So it sounds like it mitigates security risks, but not all of the censorship concerns.