This is where I think that DeepSeek won’t be as much of an economic threat as it seems. What company outside of China will put their sensitive data into a Chinese GAI platform? It’s one thing for a bunch of dumb kids to use TikTok. It’s quite another thing for an American firm to open up their data to a Chinese bot.
The U.S. Navy has instructed its members to avoid using artificial intelligence technology from China’s DeepSeek, CNBC has learned.
In a warning issued by email to “shipmates” on Friday, the Navy said DeepSeek’s AI was not to be used “in any capacity” due to “potential security and ethical concerns associated with the model’s origin and usage.”
A spokesperson for the U.S. Navy confirmed the authenticity of the email and said it was in reference to the Department of the Navy’s Chief Information Officer’s generative AI policy.
Nothing wrong with Deepseek if you use it locally. It is Open Source. But if you are putting your data into THEIR site… yeah, good luck to you.
What does that mean exactly? Can China not see anything entered into their AI? Can we be sure of that? Just asking as I am a bit ignorant on the subject and probably should not be. What if part of its ‘learning curve would always be to assess threats to China and it would choose to report back? I know that could be possible, but I don’t know if that could be figured out through analysis of the program and I don’t know if the report back could be made untraceable.
I wonder just how disappointed China is that Democrats did not win the election. I don’t think it is a coincidence this came out in the first weeks of a new Presidency and I wonder if China was hoping it would start being used heavily by the West with, shall we say, less Government scrutiny of its potential threat to us.
The model is open source. So you can see how it is run. And you can host it locally in your own systems so that it is “offline” to the public internet and can’t talk to China or anyone for that matter.
OR you can use their hosted version that most definitely sends data back to China. But if you put any proprietary data into someone else’s AI… you don’t have a clue as to what will happen to it.
This has little to nothing to do with the current or prior administration. This is about AI market share and showing off what you can do. Plus there were multiple other models released as well recently out of China. They got a lot of smart people there doing amazing things. All under the CCCPs thumb.
Deepseek most certainly vacuumed up data from OpenAI to train their model which is against OpenAI terms of service. But OpenAI did the exact same thing to millions of other websites on the web such as YouTube, Reddit and other copyrighted sources. They stole from a thief.
Thanks. I was not trying to politicize it specifically, but it took Trump to ban Tik-Tok which is openly collecting data for the Chinese and Dems do seem to be blase’ about it. I think Tik-Tok is not banned under KH and Deepseek would not have had a negative or competitive reaction to it, that is all. Maybe it would be better if it were that way, no one knows for sure, but i am pro-America as opposed to pro-Government so I am pro-MAGA in the economic sense.
AIs are learning processes and I do know an AI ‘expert’ who thinks it can be initially programmed to have certain learning paths and certain rules, kind of like Asimov’s 3 laws of robotics. I don’t know if what he says has any truth to it and am wary of his talking points so I asked here.
‘AIs are learning processors’ is what I meant.
I don’t think they should ban TikTok personally unless they show us PROOF of it being a danger. Data collection happens everywhere and is sold to everyone around the world. No stopping it. And while I was not a user of Tiktoc, no one showed me any reason to believe it wasn’t just another social media platform.
Yes all these things learn based on what is input in them. So they could have a left/right bend to them certainly. Even the Deepseek model won’t really talk about Chinese controversial subjects.
Even the Deepseek model won’t really talk about Chinese controversial subjects.
Which proves the lie of “artificial intelligence”.
As for Tik Tok, when you sign up you must agree to have any and all your info, posts, requests, credit card info, geolocations, communications, phone contacts, etc, etc., sent to its CCP-run servers in China.
Btw, if you want to stop Google’s new annoying “artificial intelligence” from showing up at the top of your search results, just add a swear word to your inquiry.
>Even the Deepseek model won’t really talk about Chinese controversial subjects.
Is that in the code and is it visible in the code? How would it not talk about it if you used the outside source product? You can enter any information in it and it should work based the info provided, shouldn’t it? If you buy the open source Deepseek product and fill it with negative info on China and have it working on ways to bring China down, it will do so with no hesitation, right?
The conspiracy enthusiast in me believes that gathering all data on everyone is a danger no matter who runs it, Orwellian society is very close and better means of information gathering continues to bring us closer. If you believe that information gathering by any big brother agency is okay, we disagree on it and that is fine. With that belief, or the fatalist belief that we can do nothing about it, I am sure you will never believe that any kind of “proof” that is presented to you will convince you that it is dangerous.
For instance, if an avalanche of false charges came in from cards that were used on TikTok in order to bring down or hurt one of our major credit companies, but the addresses of the charges could not be definitively linked to China other than that they all happened to be used on TikTok and Google and/or Facebook, that would not be proof to you because it could also have been a different info gathering giant. To me it would be proof that some info gathering giant was perpetrating the crime (or hackers thereof) and so even if it wasn’t China, they COULD have been responsible and so it is stupid to give that kind of info to a nation that wants to harm the US.
We have a military because there are many countries world-wide that want to harm us. We don’t give those countries codes to our weaponry. Similarly, we should not give them anything that they could then use against us.
Is that in the code and is it visible in the code?
Yes, and no, respectively.
The deception of “artificial intelligence” is the of the same make in “open source”. OS is nothing more than free-to-use tools used to access and use core op code to create programs/apps. In reality, any programming language is open source. The thing is, once you write your code (or these days, cut and paste modules) you need something to interpret that code. That is the core code, which only the entity offering you “open source” tools knows the ins/outs of.
Apple offers “open source”. So does Google. But ain’t no way you gettin a peek at their core.
FYI, this is from the DeepSeek user agreement: We store the information we collect in secure servers located in the People’s Republic of China.
Same with Tik Tok. Same with RedNote.
Thank you!
I guess JV was right after all. The media will (or won’t) print anything based on money. Without the money coming from USAID, politico can’t make payroll this week. The New York rag may go bust too with the spigots turning off from Government money. Apparently lib media has not been able to pay their bills from subscriptions so Dems bought their own media sources with taxpayer money.
Good call JV!