nsplayr Posted January 30 Posted January 30 Hey not for nothing, but make sure none of y'all download DeepSeek. It seems like it should go without saying, but do not willingly put CCP AI on your devices. I guess common sense is a superpower these days...
Banzai Posted January 30 Posted January 30 11 minutes ago, nsplayr said: Hey not for nothing, but make sure none of y'all download DeepSeek. It seems like it should go without saying, but do not willingly put CCP AI on your devices. I guess common sense is a superpower these days... Legitimately, why? It’s open source and can be run without an internet connection.
nsplayr Posted January 30 Posted January 30 (edited) 31 minutes ago, Banzai said: Legitimately, why? It’s open source and can be run without an internet connection. I for one will never install CCP-controlled/aligned apps on my personal devices. No need to give a hostile rival nation valuable data for free. If you wanna sandbox one in a totally offline environment for fun, I mean cool I guess, but that really limits the purpose of having an AI tool in the first place IMHO. I won't be doing that either. The Navy has already banned it. “We would like to bring to your attention a critical update regarding a new AI model called DeepSeek,” the email said. The memo said it’s “imperative” that team members do not use DeepSeek’s AI “for any work-related tasks or personal use.” It said recipients (all Navy personnel) were to “refrain from downloading, installing, or using the DeepSeek model in any capacity.” Edited January 30 by nsplayr
TreeA10 Posted January 30 Posted January 30 I did hear using Deepseek the user must agree to allow access to their computer data. Can't verify that, however.
Banzai Posted January 30 Posted January 30 (edited) 1 hour ago, nsplayr said: If you wanna sandbox one in a totally offline environment for fun, I mean cool I guess, but that really limits the purpose of having an AI tool in the first place IMHO. I won't be doing that either. I guess I just don’t buy the harm or your justification that being offline limits AI. It’s not like ChatGPT looks up its answers online. These are trained models that have weights built in to essentially a massive multidimensional matrix, for the most part. Which is why they can be run offline. And also why when I asked the most advanced ChatGPT o1 model, as of this afternoon, who the president was it told me: “As of my most recent confirmed information, Joe Biden is the 46th president of the United States.” This was with a good connection to the internet. The only thing being offline limits are calls to other APIs (like image generation), which you can verify within the open source code. Now, you do have to understand that they will be able to censor answers and information on things like Tiananmen Square or CCP perceptions using online answer scrubbing. But as long as you know that internet enabled answers come from a biased source, I still do not get the fear mongering. I have seen many folks that are able to get the offline DeepSeek models to talk about things the CCP would not want them to. edit: To be clear, I know if I used DeepSeek I should not ever type anything about myself or the government (I also know this about ChatGPT, Claude, etc.), as they will collect ANYTHING we put in the app. But I don’t understand those that say there is substantial risk in simply having the app. The navy reaction to me screams extremely overconservative easy button. Edited January 30 by Banzai
Lord Ratner Posted January 31 Posted January 31 9 hours ago, Banzai said: Legitimately, why? It’s open source and can be run without an internet connection. It is not open source, it's open weight.
Lord Ratner Posted January 31 Posted January 31 (edited) 7 hours ago, Banzai said: I guess I just don’t buy the harm or your justification that being offline limits AI. It’s not like ChatGPT looks up its answers online. These are trained models that have weights built in to essentially a massive multidimensional matrix, for the most part. Which is why they can be run offline. And also why when I asked the most advanced ChatGPT o1 model, as of this afternoon, who the president was it told me: “As of my most recent confirmed information, Joe Biden is the 46th president of the United States.” This was with a good connection to the internet. The only thing being offline limits are calls to other APIs (like image generation), which you can verify within the open source code. Now, you do have to understand that they will be able to censor answers and information on things like Tiananmen Square or CCP perceptions using online answer scrubbing. But as long as you know that internet enabled answers come from a biased source, I still do not get the fear mongering. I have seen many folks that are able to get the offline DeepSeek models to talk about things the CCP would not want them to. edit: To be clear, I know if I used DeepSeek I should not ever type anything about myself or the government (I also know this about ChatGPT, Claude, etc.), as they will collect ANYTHING we put in the app. But I don’t understand those that say there is substantial risk in simply having the app. The navy reaction to me screams extremely overconservative easy button. You are incorrect. Chat gpt absolutely has a web search capability. You are confusing the LLM with the many surrounding services that create a product you can use like ChatGPT or DeepSeek, including the phone or web app that acts as an interface between the LLM and the user. The app has all the same security risks any other app, such as TIK TOK, can have. Edited January 31 by Lord Ratner 1
SurelySerious Posted January 31 Posted January 31 OpenAI Investigating If DeepSeek Used Its Models: https://www.wsj.com/tech/ai/openai-china-deepseek-chatgpt-probe-ce6b864e?st=kvf16Z&reflink=article_copyURL_share
uhhello Posted January 31 Posted January 31 Spokespeople are saying those are 'billed to' not shipped to locations. Quibbling point.
busdriver Posted January 31 Posted January 31 Anyone who doesn't think China and Chinese companies did what was necessary to get their hands on the GPU horsepower required is naive to say the least. The top of the line GPUs in the fall of 2022 were fairly hard to get initially, as Nvidia prioritized getting as many chips to China as possible before the ban went into effect. I imagine many of those 4090s did not end up in gaming computers.
gearhog Posted Sunday at 08:05 PM Posted Sunday at 08:05 PM US State Department position on Taiwan Before & After 1
gearhog Posted Monday at 11:47 AM Posted Monday at 11:47 AM I think it's reasonable to conclude that the line "We do not support Taiwan independence" was purchased illegally by a foreign adversary from a fifth column here in the US. 1
Clark Griswold Posted Tuesday at 04:36 AM Posted Tuesday at 04:36 AM Getting crazy down under https://www.19fortyfive.com/2025/02/china-is-playing-a-dangerous-game-of-military-chicken-with-australia/
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now