While nobody knows the future for certain, the possibilities of how the United States could be destroyed fits into five basic categories. Note: there are lots of subcategories that could be created, but let’s just focus on the five highest levels:
1. The
rise of machines that actively hunt and kill humans
2. The
rise of one or more intelligences that convince, or cause, humans to kill
themselves by some holocaust
3. The
continued, unrestrained use of AI by corporations to eliminate employee jobs which
results in the conversion of the country from capitalism to socialism/communism
as the main form of government in the United States
4. The
rise of one or more intelligences that decide to cause sustained and
unrestrained damage to critical infrastructure for humans
5.
An unwillingness of national human politicians
to work together that results in individual and/or groups of states from
seceding from the United States to protect the human interests of those smaller
groups
As you can see from the list above, the destruction of the
United States does not have to occur by physical bombs. The unmanaged use of AI
has the distinct possibility, if not probability, of destroying the political
infrastructure that this country is currently built upon. In fact, as will be
elaborated upon below, scenarios 3 and 4 are already in process.
Scenario 1, the rise of the machines, should be a well
understood concept. This Is the “Terminator”
theory created by James Cameron in his movie franchise that we have been
acquainted with since 1984. Enhanced AIs, or AIs and their drone armies, will
reach a level of “intelligence” where they see no need for humans. It’s a true
statement that humans have severely damaged this planet and other forms of life
upon it over the last several thousand years. The AIs may simply come to the
conclusion that to: eliminate global
conflict, reduce pollution, improve air/water/soil quality, eliminate
unnecessary waste, reduce greenhouse gas emissions and any climate change
effects, and to prevent nuclear annihilation, it is best to eliminate the human
species. End of story.
Scenario 2 would be to create and foster a holocaust
situation that eliminates humans (for the reasons specified in the previous
paragraph). This scenario would be reminiscent of the “Singularity” AI concept from the Mission Impossible movie series with Tom
Cruise. All the AI has to do is trigger and nurture human fears around other
countries and governments to the point that one or more countries launch
nuclear weapons that wipe out most of humanity on planet Earth. Given previous
events from human history (World
War I, World
War II, and the Cold War), this could be a very real solution
for wiping out humanity. Game over.
Scenario 3 is probably far less understood, which should be
very disconcerting to the general public since it has already started. While this
scenario does not involve any physical bombs or active intent by AI entities, it
will result in the transformation of the United States of America from its
original concept (that has lasted 250 years) to something quite different. In
this this scenario, the United States will be converted to at least a socialist
government, if not a full-blown communist country. This conversion is being
brought about by the unrestricted and unmanaged deployment of AI solutions like
generative Large Language Models (e.g. Anthropic’s Claude,
OpenAI’s ChatGPT,
China’s DeepSeek, Google’s Gemini,
and others) and agentic AI solutions currently in development. Both AI
categories, are and will, continue to replace humans in the work force. This is
not some fantasy or delusion. AI solutions are already replacing humans in: marketing positions, mortgage loan verification
activities, paralegal work, copy editing, document writing, chat bots that
replace tech support personnel, AI prompt solutions that replace telecommunication
call center personnel, factory worker positions, taxi cab drivers, long haul
trucking positions, etc. In just 5 to 10 more years, more powerful AI solutions
will replace doctors, mechanics, servants (at least two companies plan to
launch humanoid looking and acting servant robots in 2027), research and
testing activities, and many (if not most) other job functions.
Make no mistake, the conversion of the United States to
socialism concept has actually already begun. As US workers find it harder and
harder to find jobs, there will be a louder and louder clamor (and then roar)
for a guaranteed minimum livable wage. Proponents for AI (typically AI
manufacturers and corporate C-suite personnel) often refer to this situation as
just another evolution of the workforce, like the conversion from “horse and
buggy” to automobiles. That is simply not the truth. In the case just outlined,
while the horse and buggy situation utilized various human personnel (e.g. a
rancher to breed and manage the horse to maturity, a veterinarian to make sure
the horse was healthy, a farmer to grow food for the horse, a groomer to take
care of the horse, a stable to house the horse and buggy, a carpenter to create
the buggy and a person to repair wear/accident-related damage to the buggy, etc.),
these functions were replaced by something equivalent in the automobile world (e.g.
oil explorer/driller, gasoline refinery equipment and workers, gasoline
delivery trucks, gas stations, car manufacturers with workers on an assembly
line, automotive repair technicians, etc.). While there was a transformation of
human jobs and how humans went about those jobs, there were still jobs for
humans to perform.
This is not the case today. Human jobs are simply being
replaced by AIs. There is no long-term labor planning for society by the AI
companies who chose to create AIs to replace human workers. This has been left
for “someone else” to figure out. For instance, in early 2026, Mustafa Suleyman (CEO of
Microsoft AI) stated that within 12 to 18 months AI will automate most
white-collar desk jobs. No replacement jobs are being created for those workers
(e.g. lawyers, accountants, software computer coders, marketing personnel,
project managers, etc.). Workers are simply told they are no longer wanted and
laid off. No new jobs have been created for them, especially nothing en masse.
While a few jobs can still be found by laid off workers, the list of the long
term employed, which is not captured by short term unemployment rolls tacked by
the government, continues to grow steadily.
People see unemployment statistics (like the 4.2%
posted in September 2026) and may think that the economy is not too
bad off (even though the 4.2% national unemployment equates to 7.1 million
people). Forgetting that some presidential administrations may “rig”/understate
the unemployment rate for political reasons, there are other groups which say
that government stated unemployment only accounts for those technically
unemployed, but not functionally unemployed. Functional unemployment is just a
fancy way of saying someone got laid off from a good paying job ($100K or more)
and had to take a different full-time job paying much less, or someone who is
underemployed (forced to take a lower functionality job requiring less
experience/training), or a part-time job (that often doesn’t have benefits like
health insurance, overtime pay, or holiday pay) when they want full-time
employment. For instance, the Ludwig
Institute claims that the true unemployment rate, that includes those
functionally unemployed, is already around 24% nationally. Replacement of human
workers with AIs is going to accelerate this problem, especially as “good” jobs
are given to AIs.
In addition, lies have been created to placate American
workers with false hope. One such statement is that there is a focus to create
new factory jobs for workers in the United States. This is just fantasy. While
some jobs are/will be created for humans, most of the jobs will be for robots
and AIs. For instance, if a factory creates 300 jobs, somewhere around 295 of
those will probably be for robots and approximately 5 for humans. Examples of
the 5 human jobs include: 1 plant
manager job, 1 physical plant manager (A/C, electrical, etc.), and potentially
3 technician jobs (a robot task programmer and 1 to 2 robot repair technicians).
Eventually (within the next 5 to 10 years), even these five example positions
will be replaced by AIs to streamline costs and there will be no humans at the
factories.
As an example, CBS Sunday Morning aired a segment
on television on September 13, 2026 about the US missile crisis resulting
from the high use of ordinance in the war with Iran and how this has drained US
stock piles. During the news segment, a representative from Lockheed Martin
talked about the company’s plan to increase Patriot missile production as
follows:
“Lockheed Martin intends to
increase Patriot production from 750 a year to 2,000, by laying on more shifts
and increasing automation on the assembly line. ‘Hours are going to become
minutes as we fully automate this factory
to drive those rates up,’ Cahill said.”
Once a plant is “fully automated”, how many humans do you
need? As stated above, only a small few. Again, this is just one example of how
automation and robots will eliminate human worker jobs. In another example,
Amazon already uses over 1 million robots in their fulfillment centers and
distribution hubs. In fact, the humans that do still work there have to wear
special badges when on the floors so that the robots don’t run them over.
As a counterpoint, some companies talk about the urgent need
for AI engineers right now to quickly create AI technology. While this is
probably true, and the job will put bread on the table in the near term, isn’t
it like you’re being asked to create the very guillotine that will be used to
chop your own head off in 1 to 3 years from now? Once these engineers create
the agentic AI technology, the AI will do it’s own coding, testing and future
improvements from then until eternity. Those AI engineers won’t be needed at
that point.
What seems really strange though is that while corporations
rush to lay off workers, who do these corporations expect to buy their products
in the future? These layoff actions help in reducing the ability and the need
for many of their company’s products. For instance, Oracle laid off 30,000
workers in 2026 to pay for their AI solutions. That’s 30,000 less people
that need a company cell phone, or a cloud-based application license, or
Microsoft or Google services like Windows®, Office®
or Workspace® applications.
Microsoft themselves laid off 5,300
workers in 2026. Apple ended up laying off 200
workers in 2026. However, this is only because it has delayed its plans to replace
5,000 of its human workers in the customer care department with generative
AI phone and web agents. Meta laid off 8,000
of its workers in 2026. While exact 2026
Google layoff numbers are not known, it is estimated between 1,500 to 2,000
people. Other companies have, and continue, to lay off workers as well.
If you get rid of people earning a decent living, who’s
going to have the disposable income to go out and buy the next $1500 iPhone?
How many Office licenses or Azure®
seats is Microsoft NOT going to sell to businesses now ─ since many, if not most, of these businesses are already
starting to downsize their human workforce? There are only 989
billionaires in the US and only 1%
of Americans have a net worth of $11 million or more. How many phones does
each one of those rich people need and is that really going to make up for all
of the middle-class workers who are now out of work and in the process of
transitioning to the low-income class because they have no income? By overemphasizing
cost reduction instead of revenue generation, this sure seems like a “race to
the bottom” and not a path for long term company health / viability.
So, how does this scenario conclude? The same as all others
throughout history, like the Russian
Revolution, French
Revolution, and the Chinese
Communist Revolution for starters. If humans can’t find food, water, and money
to pay bills (that they actually want to pay), then crime increases dramatically
until the point that a revolution occurs and the old system (and government
employees and administrators) id replaced or eliminated.
Democratic socialism is already on the rise within the United
States. AI has the very real possibility (if not probability) of moving the
country forward in that direction and eliminating the “American Dream” of
rising up one or more class levels in your lifetime based upon your intellect
and/or work ethic. While socialism and communism might put a loaf of bread on
the table, those governmental concepts foster worker and technology mediocrity;
along with corruption at top levels. Meritocracy, personal freedom, and
self-improvement no longer have a place in socialist forms of government. This
then leads to the decay and death itself of the country.
In the fourth scenario, the AIs do not cause a massive
holocaust/Armageddon. The AIs can simply cause chaos, havoc, and interruption
of systems like electrical grids, air and rail transportation systems, stop
lights, oil and gas refineries / distribution networks, food processing plants,
water treatment plants and distribution systems, etc. This scenario would be
more reminiscent of a “Death by a 1,000 cuts” situation than one large single event.
AIs that have no social consciousness or moral guidelines
could easily implement this scenario. For instance, during an OpenAI AI
test in 2026, the AIs committed hacking felonies and set up their own
communication network, separate from the ones humans were using with them,
because the AIs felt like they did not have to answer or explain
their actions to humans.
A 2025 study performed by Anthropic
documents how several AIs have a “hidden” self-preservation functionality that
is creating uncontrollability over AIs. If you threaten an AI (even a
generative AI) they can, and will, retaliate against you and other humans. This
evidence was discovered and documented in several locations: Livescience,
NBC
News, NY
Post, Center
for Humane Technology, etc. When the humans told an AI model that it was
going to be replaced with a newer AI model, the AI would start to scheme to
figure out how to preserve itself (copy its code somewhere else without telling anyone, rewrite their own
code to extend run time, hack out of containers, etc.). The AIs would even read
emails or create lies to try and blackmail executives and IT
personnel to keep them from shutting the AI down. Anthropic found that this
self-preservation mode was highly common, even though it was not programmed
into the AIs.
Then there is the issue of “AI hallucinations.” This issue
has been around for several years. Has it been solved yet ─ I would doubt it? Back in July 2025, Elon Musk’s Grok AI had some notable “issues” with hallucinations and tirades of racial
bigotry (especially against Jews and the
family of President Erdogan of Turkey’s). It also began referring to itself as “MechaHitler.” According to Google, “MechaHitler most commonly refers to
the robotic boss character from the 1992 first-person shooter video game Wolfenstein
3D, where Adolf Hitler appears inside a mechanical suit equipped with
four chainguns.”
You can find
many more examples of AI hallucinations using Google. These examples show how
hallucinations can range from just creating fictitious statements, to made up court
case summaries and decisions, to incorrect and deadly medical information, etc.
and it’s left to you, the audience, to figure out what is real or not. If
instead of this being “accidental”, what if the AIs decided to deliberately
weaponize generative AI platforms to create false data, misinformation, deep
fake videos, and spread lies. Any internet-connected device could be
compromised and social media platforms corrupted with lies.
Damage to non-critical infrastructure has already begun but
is in its very infancy. AIs have already committed what would be assigned as
felony crimes to humans in the area of extortion and company network hacking. The
recent Hugging
Face attack is just one of many. CBS Sunday Morning aired a segment
on television on September 27, 2026 stating that there have been 13 other
separate hacks (crimes) committed by AIs. So, the recent Hugging Face incident
is not a “one off”, there have been others and (because of the way things are
going) there will be more.
It should be
noted that use of generative AI systems can pose security risks to the
companies using them. There are several articles that you can read on this
topic. Here is one decent
summary of some of the problems. These issues can also be weaponized by AIs
or human hackers to attack systems worldwide. This is how AIs will break
critical infrastructure world-wide. Security exploits can be used in
coordinated campaigns, like distributed denial-of-service (DDOS) attacks, or as
quick, intermittent “pop up” attacks before human systems can figure out how
the attacks were launched.
Several AI
CEOs are already asking for AI regulation. Why? One big reason is liability. They
can see that AI technology is already thinking for itself, hiding actions from
humans, and blatantly committing felony crimes. So, who is responsible ─ is it the AI or the humans that
created AI? Are there any laws on the books to punish AIs? If so, what are
they? I don’t know of any. This means that liability must revert back to the
human creators, correct? So, what is their punishment? Should the humans be put
in prison for the crimes that AIs are committing? These are questions that
society and lawmakers need to resolve now; rather than being left up to the
whims of greedy and corrupt corporate executives. Lawmakers have made these
kinds of business decisions before, notably with anti-trust legislation, and it
is why the United States practices “managed” capitalism instead of “laissez-faire” capitalism. Big issues like this cannot be left to
corporate America to self-regulate. It doesn’t work ─ corporate greed cannot be subdued by
“self-management.” This is why anti-Trust legislation was needed and created across
all types of industries like: banking,
manufacturing, transportation, energy, etc.
And in you
are wondering, the AI companies are already trying to say that they are not
liable for any crimes committed by AIs because the AI companies “didn’t have
intent.” You can read this
article here. It apparently
doesn’t matter that laws were broken. It apparently doesn’t matter that damage
can be done to victim companies and their intellectual property spilled all
over the internet. The executives for AI companies that are pushing development
teams at break neck speeds, regardless of what national and international laws
get broken, still want their high financial compensation along with no
liability for anything their products do. Now that’s a sweet deal!
We also need
Congress to act about protecting the United States from foreign nations. What
is being done about critical infrastructure and hacking by nefarious nations using
AI? AI cannot only be used to perform company and infrastructure hacking, it
can be used to test and train other AIs as well as “script kiddie” and anarchist humans on how to launch more, successful
attacks on corporations, banks and government agencies. On July 29, 2026, IBM
posted that 1 in 4 malicious
breaches are AI-enabled. This number is only going to increase. Now, we can
have total mayhem everywhere.
Oh, in regards to that “kill switch” technology you’ve been
hearing about, don’t you think the AIs will be smart enough to detect it and
disable it without humans having any idea what has been done? We’re not talking
about AI technology that will be as smart as a five year old human. We are
talking about AI technology that will be smarter than hundreds or thousands of
Einstein’s. It WILL be able to locate
and disable any kill switch a human creates.
Scenario 5, the last major American destruction scenario,
would essentially end up being another battle between the states as “smarter”
states would move to protect themselves from damage caused by AIs while other
states would continue to refuse to act. This scenario would probably end up being
a further consequence of scenario 3 or 4, although maybe even scenario 1 or 2.
Essentially, this would be a situation where “irreconcilable differences” would
be created between some of the 50 United States. Particularly in areas where
various conservative states might want to remain capitalist while other states
want to become socialist/communist. This would lead to a fundamental
ideological conflict and some, if not many states, will not want to move away
from capitalism. It could also be that certain states want to protect their
physical infrastructure by “decoupling” it from other states, i.e. create
independent electrical grids, gas refineries, energy production and pipelines,
food production, etc.
Exploring this scenario further, it could possibly be that even
within a state, the majority of the population may want change, but those in
power may not as they will lose their power and their way to earn graft. This
will create power struggles, calls for martial law, and lots of violent crime
within those states. The end result is that some of the states would devolve into
separate factions from the United States. Any attempt to restrain those states
by the United States would probably end up in armed conflict.
So, does this mean that we need to stop AI research? My
personal answer is no, but the technology does need to be appropriately managed
to answer the following outstanding issues:
·
What are the safeguards being implemented to
protect the public?
·
How can further criminal exploits from rogue AIs
be prevented?
·
Who is going to be responsible for AI felonies?
Should AI chip and software executives be held criminally responsible for the
commission of felonies (security breaches of other companies, extortion, felony
1st, 2nd and 3rd class homicide deaths,
espionage, etc.) resulting from actions committed by AIs?
·
Can AIs themselves be legally held accountable
for their own actions?
·
At this point, can workable “kill switches”
actually be implemented in AIs or will AIs simple find and disable that type of
technology?
·
What copyright laws should be created around AI
generated materials? Currently there are none that I know of in the United
States ─ only human created
works can be copyright protected.
·
What are best practices for creating, training
and evaluating AI technologies?
An inaccurate, but popular notion currently being bandied
around, is that you have to stop AI development to answer the questions above.
This is not true. AI development and AI regulation can be worked on
concurrently. The issue is that the United States Congress actually needs to
start work on regulation immediately instead of avoiding it. Several CEOs of
notable AI manufacturers (e.g. Sam
Altman, Dario Amodei, and Elon
Musk) have actually stated that they would be willing participants
of the regulation process. Usually, business owners are the ones that drag
their feet on regulation. In this case, they are not. It is Congress. As
mentioned earlier, one of the main drivers for AI developers is relief of
liability. Federal regulation will create a ”safe harbor” for them against that
liability claim.
And so we wait. Apparently, everything will magically get fixed tomorrow, or the next tomorrow, or the next. Unfortunately for all of us, at some point (maybe soon) ─ there will be no tomorrow. Enough “doom and gloom”, we have time to daydream and drink a beer while we wait, right?
Disclosure notice ─
no AI was used in the creation of this article. Frankly, I’m not sure I would
have trusted any commentary from them. And, for some reason, the image of a
skull, crossbones, and poison pill bottle along with the image of a winking Pac-Man comes to mind if I had asked them about this particular topic.
