Finally finished reading this article. Wow, just wow. This article is next generation in spinning a narrative.
Here is the crux: The Meme itself decries the fact that AI companies are gatekeeping their models to give us poor folks nerfed versions.
And here is how the article frames it:
Anthropic refused to give Pentagon the "un-nerfed, non-gatekept" raw AI model to Pentagon for national security.
The article characterizes this as "The AI would not be used for mass surveillance of American citizens. It would not be used to control fully autonomous weapons systems where no human being was in the decision loop."
Very laudable right? But, what the article DOES NOT mention is that Anthropic wanted to be the gatekeeper who decides when the AI can be used and cannot be used, on their own, with no visibility. So a woke programmer in Anthropic could potentially put in the gatekeeping for Pentagon in software used for national security.
That is when Pentagon put Anthropic as "Supply Risk" because that is indeed what it is - Pentagon would be in a high risk if they had to depend on a private company to gatekeep the critical software features.
OpenAI, on the otherhand made a deal where these two "gatekeeping" would be done more transparently with the Pentagon, and hence got the deal.
Point 1: This article portrays private gatekeeping as a GOOD thing while trying to pretend they care about these companies gatekeeping.
Now, lets go to the whole "changelog-gate". The summary is that "Claude had to reduce AI "effort" to MEDIUM since they could not afford to keep running at higher effort that they had been offering due to budget issues.
The article connects this to "Pentagon punishing Anthropic"
But lets unpack. Anthropic was able to subsidise their "HIGH QUALITY" offering because of "Pentagon Contracts".
Anons who are smart would find that very familiar. This is how the Big Tech companies were being controlled by the Cabal - given them tons of money in the name of "government contracts" and give them a competitive edge.
Point 2: The reason why Anthropic/Claude was the darling of all AIs was because they had sweetheart deals with government contracts
And a final, anecdotal point. Having been using OpenAI GPT-4/5 for a few months now, and having compared it with Claude, I can tell anyone that, atleast when it comes to coding (one of the toughest usecases for AI) I can use OpenAI's $20 subscription exactly as effectively as the Claude MAX subscription that costs >10x working on really complex systems software that is typically only handled by senior software engineers with field expertise of atleast 3 years.
So, I had always suspected that Claude was being pushed rather heavily but this article actually confirms my suspicions.
This also lines up with all the constant anti-OpenAI attacks, be it allegations by ex-employees that their AI has become sentient to recent allegations hat Sam Altman is "opening portals" - one thing is for sure. The Cabal is definitely trying to take control of OpenAI.
If the Cabal is trying to take control over the AI models then it stands to reason that's why the govt is pushing for unilateral, open, and unrestricted access to all functions.
I'm sorry (not really), but giving any one entity, whether it be public or private, unfettered, unilateral, unrestricted access to these AI programs for any and all possible uses is absolute insanity.
Of course our govts would want to tie in those programs with military functionality. And of course they'd want to use them in such a way as to eliminate the need for humans to be in the decision loop as to how, where, why, and what those autonomous weapons systems would be used. We're already seeing widespread governmental abuse of surveillance AIs. And now in Ukraine they're being used with drone technology for battlefield testing in a ls close to a fully autonomous decision making structure as possible! These things are actively deciding which combatants to target and eliminate. It's a very small step from targeting combatants to targeting civilians caught in a battlefield that dynamic and always changing. Civilian casualties happen in battle and war, but we don't fully understand how these models reason out decisions, so it's more than likely that we will inevitably see a civilian population targeted by AI powered drones in the very near future.
The Pentagon wanted unrestricted access to Anthropic's for the very purpose of militaristic decision making. If we don't trust our own govt to not use drones against civilians when humans are in the desion-making loop, ala Obama's drone use against civilians, then why would we allow any other system to have the same type of access, especially when it's unclear whether or not restrictions against harm to humans are in place? The fact that these AI companies refuse to admit one way or the other is very telling. It tells me that they don't have any safety restrictions as to the type or amount of harm they can cause to humans.
There needs to be a global pause on all AI development until The People have a direct say in how, where, and why l, and for what reasons these programs are implemented. Until then, absolutely NO ONE should have unrestricted and unlimited access, whether it's our govts or even the very AI companies themselves. Not until it's clear that they can't be used for human harm.
Just remember, if something exists, and is not available to public, it is most definitely available to certain private entities, who will then use it to control the public.