• Guests may view all public nodes. However, you must be registered to post.

Grok Chatbot to Be Given Access to Pentagon Intelligence

RiffRaff

RiffRaff

Deputy Director
Staff member
Donator
Joined
Apr 7, 2016
Location
Indiana, US
lol is grok now going to be cleared TS/SCI?

on another note, damn it! I was wrong I thought for sure it was going to be like most government networks, everybody has their own for each branch of the military plus all the IC members will want their own AI, then there will be unclass versions, collateral versions and then high level versions none of them will talk to each other, there might even be one with the word "joint" in it that is supposed to be for everyone to share but there will be 3-4 different versions of those as well and everybody will have a different name for it. lol.

on a serious note, my two big worries about AI in the short term are 1) everybody getting their own AI and when it is time to give a briefing to the powers that be each branch/ agency has fed the AI with their own stuff and we get 18 different answers, we get conflicting reports and it turns into a mess. the 2nd worry I have with AI is if the white house and/or congress wants their own AI and then decides they do not need the IC to brief them cause they have their own AI.
 
I wonder what can be done to make sure the AI doesn't leak important information, or worse yet cause WW3 by mistake ( or even on purpose! :( )
 
StormyCat said:
I wonder what can be done to make sure the AI doesn't leak important information, or worse yet cause WW3 by mistake ( or even on purpose! :( )
Click to expand...
I think what they mean here is that the Pentagon will deploy a copy of Grok with access to its data; not that the public edition of Grok will have access to Pentagon data (the latter would be ludicrous).

I'm more puzzled by Hegseth's statement - that artificial intelligence systems must operate “without ideological constraints that interfere with the lawful use of military force”. To me, this is about the swirling doubts regarding the legality of some of his recent decisions, and reads like "This AI should be hardwired to tell me what I want to hear." (but in that case - why use AI at all...?)
 
Socom7 said:
I think what they mean here is that the Pentagon will deploy a copy of Grok with access to its data; not that the public edition of Grok will have access to Pentagon data (the latter would be ludicrous).

I'm more puzzled by Hegseth's statement - that artificial intelligence systems must operate “without ideological constraints that interfere with the lawful use of military force”. To me, this is about the swirling doubts regarding the legality of some of his recent decisions, and reads like "This AI should be hardwired to tell me what I want to hear." (but in that case - why use AI at all...?)
Click to expand...
Well. If it does actually advise using proper lawful use of military force, depending on what it actually answers it's seen as lawful or as an ideological constraint.
 
  • Like
Reactions: Torch
Socom7 said:
I think what they mean here is that the Pentagon will deploy a copy of Grok with access to its data; not that the public edition of Grok will have access to Pentagon data (the latter would be ludicrous).

I'm more puzzled by Hegseth's statement - that artificial intelligence systems must operate “without ideological constraints that interfere with the lawful use of military force”. To me, this is about the swirling doubts regarding the legality of some of his recent decisions, and reads like "This AI should be hardwired to tell me what I want to hear." (but in that case - why use AI at all...?)
Click to expand...
AI will learn from its creator and make decisions based upon the data and sources its creator initially says are valid.
Its value judgment will also be shaped by whom ever creates it.

We’ve been living in that would with Google now for almost two decades.

The old joke not to many years ago was that Walmart inc had more data storage than than the government.
Then the gov farmed it off for thr IT to Amazon. WM missed a big opportunity there 😏

I have noticed with just simple searches Any of the retail AI’s can be challenged sometimes even using on against the other.
There little more than grammar school pupils compared to what is envisioned.

Then again the public Interactions with all the AI models and date collection it all just a means of teaching AI to think somewhat more independently
 
Cilvilt_beredskap said:
Well. If it does actually advise using proper lawful use of military force, depending on what it actually answers it's seen as lawful or as an ideological constraint.
Click to expand...
Hehe we been watching the shifting sands of interpretation play out every day with politicians for the last 15-20 years. This will be nothing new.
 
Publius said:
AI will learn from its creator and make decisions based upon the data and sources its creator initially says are valid.
Its value judgment will also be shaped by whom ever creates it.

We’ve been living in that would with Google now for almost two decades.

The old joke not to many years ago was that Walmart inc had more data storage than than the government.
Then the gov farmed it off for thr IT to Amazon. WM missed a big opportunity there 😏

I have noticed with just simple searches Any of the retail AI’s can be challenged sometimes even using on against the other.
There little more than grammar school pupils compared to what is envisioned.

Then again the public Interactions with all the AI models and date collection it all just a means of teaching AI to think somewhat more independently
Click to expand...
That is what makes me Leary. Machines learning from each other. IMO
 
its just a massive thinking machine that can be used for good, solving complex problems, issues and computational problems. of course i am afraid that it might be used for evil as well cause that's the nature of humanity.
 
You must log in or register to reply here.
Back
Top Bottom