Microsoft's OpenAI Skynet For The US Military

Microsoft performed an OpenAI coup so they could militarise AI

Dec 13, 2023

Back in August 2023, Bloomberg announced Microsoft and OpenAI were competing to win Pentagon contracts to supply the US military with AI:

It is based on OpenAI technology trained on US military data, weaponising the AI into Skynet:

Back in May 2023, the Intercept asked OpenAI if the Pentagon could use their AI technology and they refused to answer. It wasn’t too long ago that OpenAI had in their terms of service that “We kindly ask that you refrain from sharing outputs that may offend others…” in regards to GPT-3.

Having tested the service, we can also vouch it originally had Terms that prohibited the use of the AI in political and military capacities as well. So what changed?

The Coup At OpenAI

To understand the coup, you must understand the context of what OpenAI was to know what the coup has done.

Sam Altman, Elon Musk, Peter Thiel and others, had committed $1 billion to the start up of OpenAI back in 2015, calling it a “research lab” and a “nonprofit”, Business Insider repeated the same evaluation of $1 billion.

In 2015, website good.is also painted OpenAI as having a “nonprofit approach” with “its open-source, public-domain nature”.

In 2016, Wired wrote an article on OpenAI stating it was a “chance to explore research aimed solely at the future instead of products and quarterly earnings”. In the same article, OpenAI CEO Sam Altman declared AI technology would be “given away for free”.

Also in 2016, for-profit Microsoft had joined the OpenAI partnership, claiming it was to ‘protect humanity’s best interests’.

In 2017, the Open Philanthropy granted OpenAI $30 million ($10 million per year), on the basis OpenAI were effectively an altruistic, charity organisation. It was ironically focused on the ‘catastrophic risks’ from AI.

Clearly then, OpenAI were painting themselves as a pro-humanity, money-adverse non-profit intending to be transparent, open-source and public domain, even to the point of being free, and even going so far as acquiring grants under this guise.

Then OpenAI Changed

March 2019, Greg Brockman and Ilya Sukskever, employees of OpenAI, declared OpenAI had changed from being a “nonprofit” previously known as “OpenAI Nonprofit” into “OpenAI LP”, declaring itself a “capped profit” model.

Their objectives no longer aligned with their publicly stated goals.

Barely months after switching from their non-profit facade to a profit model, in July 2019, OpenAI CEO Sam Altman had negotiated a billion dollars from Microsoft, and it included a partnership to use proprietary Microsoft Azure services.

A year later, in September 2020, Microsoft just so happened to get an exclusive contract to use GPT-3, in violation of OpenAI’s publicly stated mission goals used to secure philanthropic funding for technology ‘given away for free’.

Microsoft had essentially infiltrated OpenAI, and as recently as September 2023, Microsoft committed another $10 billion to OpenAI.

A month after, in November, OpenAI CEO Sam Altman was fired by the OpenAI board for supposedly being dishonest. Conveniently, during this short window frame, it enabled Sam to be hired by Microsoft no less than 3 days later, along with ‘for-profit’ announcer Greg Brockman, before Sam conveniently rejoined OpenAI days after that.

It was a perfectly engineered, well-timed coup.

It was most likely Sam Altman took the firing to get out of an OpenAI contract. Likely one obliging him to be honest and not work for other companies.

Microsoft Takes Control

After this, Microsoft joined OpenAI’s board with a 49% ownership stake. Altman rebutted, claiming they were taking a ‘non-voting, observer position’.

They wouldn’t need to vote, Microsoft already controlled several key members and multiple OpenAI employees that could vote, including Greg Brockman and the CEO Sam Altman himself. It was a thin veneer of separation.

Antitrust Provoked

Microsoft have most recently misled investors by claiming it is moving away from OpenAI, when evidence shows the opposite is true. This is likely due to the fact the UK government are considering opening an anti-trust probe into OpenAI and Microsoft, seeing their ‘collaboration’ as far too close.

Business Insider bragged how Microsoft CEO Satya Nadella has ‘outsmarted them all’ with his coup of OpenAI. He’s not looking so smart now.

Privacy Flaws

What’s in it for Microsoft? Besides the Pentagon contracts, they gain a leading edge over their competitors whilst bypassing the moral objections the original developers had to the development of AI.

They also have access to reams and reams of whatever data happens to get passed through any of the AI products hosted by Microsoft OpenAI Azure services.

In July 2023, the FTC opened up an investigation as to whether or not OpenAI were deceptive in relation to their privacy and potential harms.

August 2023, OpenAI were hit with a GDPR complaint they had violated EU privacy laws.

September 2023, OpenAI were hit with a class action lawsuit for violating privacy rights in California, and another complaint over privacy was fired off in Poland.

This isn’t considering Microsoft’s own extensive privacy violations. For example, they datamined children and got fined a meagre $20 million, by the FTC.

The Pentagon involvement is even more alarming.

This Isn’t The Pentagon’s First Foray Into Skynet

In 2015 Motherboard published an article covering former DoD (Department of Defense) employee Chuck Hagel’s plans to mass automate the US military’s warfighting capability under the Defense Innovation Initiative.

Highlands Forum Andrew Marshall had been pushing for the DoD to adopt Skynet since their Skynet study (part 2 covering his handiwork can be found here), which envisions a future involving robot armies.

This same army would be fueled data from social media determining who to kill, and makes the fact that X/Twitter’s AI called ‘Grok’ datamines Twitter’s postings all the more suspicious given the timing of the rollout coincides with the Pentagon competition.

Wall Street Journal even tried to censor the ‘datamines X’ part from their article:

The DoD aren’t the only ones with an obsession with Skynet. The NSA do too, and it used to determine who to kill:

Documents pertaining to the NSA’s SKYNET program were originally covered by the Intercept back in 2015:

And this included evidence it was based on social media behaviours:

Page 3 from SKYNET: Courier Detection via Machine Learning

In our article titled “Elon's Sinister Plot At Twitter” (bearing in mind Elon Musk has an investment in OpenAI and is connected), under the Elon’s True Purpose subsection, we commented:

The US government has two options now, they can either torch the place and close down Twitter for good with no replacement, or they can datamine and track people’s habits to find all the domestic terrorist parents they hate. It looks like they’re aiming for the latter in an attempt to suppress the growth of alternatives like Gab.

We even predicted the name he would use:

Worst case he makes the datamining X app that tracks every single post you make allowing the government to track your history and suppress your speech in the real world lest you become a parent (AKA domestic terrorist). I wonder if it comes complete with financial CDBC and in-car Tesla tracking following your movements in real-time? Mmm, mass surveillance and clean-up crew scrubbing, yum.

Do you now understand why Twitter wants your facial recognition data?

Returning to point, the reason Microsoft essentially ‘stole’ OpenAI is the developers who had signed up to only do ethical AI research were tricked into supplying their best work, only for it to be gobbled up by the US military as part of an automated warfare campaign to target the populace.

Goes from bad to worse, doesn’t it, dear reader?

The Daily Beagle

Discussion about this post