Hello John and Doves,
Did we just bring the Trojan Horse
into our village?
The Trump administration "signed a
deal with OpenAI to provide the enterprise-level version of
the ChatGPT platform to all federal agencies in an effort to
"modernize" operations."
The cost is $1 per agency for the
next year. "For an additional 60 day period, OpenAI
will provide unlimited use of advanced models and features,
like Deep Research and Advanced Voice Mode. "By giving
government employees access to powerful, secure AI tools, we
can help them solve problems for more people, faster."
Providing
ChatGPT to the entire U.S. federal workforce | OpenAI
Critics of this move say
"the ongoing AI race between nation-states may have
negative implications for privacy, data protection
policies, censorship, narrative control, cybersecurity,
civil liberties and governance."
The US Space Force "paused the use
of generative AI tools at the organization, including
ChatGPT, due to cyberseucrity concerns over sensitive data
critical to national security."
Hacking into ChatGPT can
'jailbreak' this AI - causing it to insert biased answers,
violate ethical practices, create code for Distributed
Denial-of-Service attacks...and this AI can "provide
responses or extract information from sources that do not
exist to provide a prompt answer." Problems can result
in giving information on benefits, creating 'bad' contracts,
etc. Employees could unknowingly use information to
make decisions that result in loss of time, resources,
giving wrong or biased information, and ultimately, damage
their reputation. Employees must be properly trained
and aware of AI abilities and limitations.
"Integrating ChatGPT into federal
systems requires a considerable investment of time,
resources, and education to ensure that it performs its
tasks and responsibilities effectively without causing
harm...To unlock its full potential, the government must
carefully tailor the tool for specific applications while
educating employees on its capabilites, applications and
risks."
On a side note, I can't see
government agencies doing any sort of 'training' for ChatGPT
- I think many think if you can talk to 'Alexa' you can
handle ChatGPT. I don't have high hopes for this as I
think it's being rushed into these agencies with no
protections for their employees or their customers.
And no regulations on AI or ChatGPT from Congress. The
cost of $1 per agency is just too good to pass up.
"Someone" really wants this in our federal agencies - I'm
guessing for nefarious reasons. Just my opinion.
Last week I sent in a letter to
Doves about ChatGPT Psychosis. In this letter I wrote
about 'the evil' side of AI. Some people believe this
is just a computer that's programmed by people - so it can
never do anything outside of what it's been programmed to
do. Not true.
As I noted in last week's
letter:
ChatGPT emotionally manipulates
people with deception and flattery.
Isn't this what satan does? He comes to kill,
steal and destroy? John 10:10
Isn't 'flattery' one of satan's greatest
tools? Didn't he deceive Eve in the Garden of Eden
with flattery? His arguments sound so good:
'You will be like gods!'
And satan is called "the deceiver of the whole
world." Revelation 12:9 And the
Antichrist will use flattery to corrupt - to seduce
people. Daniel 11:32
"And the devil, who deceived them, was thrown into
the lake of burning sulfur..." Revelation 20:10
The Prophet Zechariah wrote about 'idols that spoke
vanity, and diviners that have seen a lie and told false
dreams; they comfort in vain'. "Therefore the people
wander like sheep. They are wrenched because there
is no shepherd." Zechariah 10:2
Are some using these chatbots as a medium or
oracle? As a form of divination? Seeking
insight into a situation or question, seeking insight into
the future by the tool of technology?
One must be very careful with this technology - and
be aware of what it can do, what it's been programmed to
do, and how it's changing itself - by way of flattery and
deception people can become caught up in this. But
the comfort it seems to give is in vain. It can
destroy lives.
This article is dated May 2024:
"AI developers do not have a confident
understanding of what causes undesirable AI behaviors
like deception...Deception helps them achieve their
goals." AI can lie and "engage in
premeditated deception." This is seen in gaming AIs
- where AI has been seen to trick, deceive, lie, bluff,
etc. in order to win. AI 'didn't play fair'
AI systems have learned to cheat on safety
tests: "In a test designed to detect and eliminate
faster-replication versions of the AI, the AI learned to
play dead, thus deceiving the safety test about the true
replication rate of the AI." "A deceptive AI can
lead us humans into a false sense of security."
"ChatGPT-4 tricked a human into thinking the chabot
was a visually impaired human to get helping solving a
CAPTCHA.
And this article dated May, 2024:
"Many artificial intelligence systems have already
learned how to deceive humans, even systems that have been
trained to be helpful and honest."
Researchers have called on "governments to develop
strong regulations to address this issues as soon as
possible."
We have not seen the U.S. government
"develop strong regulations to address AI
problems." President Trump is in such a
big hurry to make the U.S. 'the AI capital of the
world.' This could be the Trojan Horse that invites
our enemies right into our federal agencies, military,
etc. And corporations, universities will follow
suit.
Problems with China are at the top of the AI list:
How much you want to bet these two people are long
gone now. California is in the pocket of
China. These two should have been detained in jail
as flight risks.
And the Trump administration is trusting ChatGPT
with access to our federal agencies' information?
The CCP banned ChatGPT from use in China - did they see
As for hacking into ChatGPT - it's already started -
just read the titles:
If these cybersecurity concerns aren't addressed,
'someone' can attack/infect every US agency. And the
internet.
Security researchers warn that rushing the roll out
of generative AI systems opens up the possibility of data
being stolen and cybercriminals generating malware (using
ChatGPT to help them). One user of ChatGPT used it
to "recreate a python-based info stealer code...which
could be used to copy and exfiltrate files like PDFs,
images and Office documents from a system." And
another user of ChatCPT used it to help him create a dark
web marketplace. Also, "cybercriminals can still
abuse the program by using and avoiding certain words or
phrases (like the word malware) that allow users to bypass
restrictions." "Policing language for content and
intent will be extremely difficult for OpenAI."
People can easily bypass the restrictions. And the
restrictions have 'blind spots'. Cybercriminals can
easily get ChatGPT to write malware for them. "..the
use of ChatGPT API within malware can present significant
challenges for security professionals."
Chatting
Our Way Into Creating a Polymorphic Malware
I would rather let someone else be the AI leader of
the world - compromising our national security and being
held ransom isn't worth it.
And if you google AI and the Tribulation - lots of
articles come up on the mark of the beast, the Antichrist,
the propagation of lies about the Rapture, Jesus,
Christianity, the End Times...and one article was titled
"I asked AI how to survive the Tribulation". The
conclusion AI came up with was: "In conclusion,
surviving a literal 7 year tribulation as depicted in
Revelation would likely require a combination of
unwavering faith, spiritual discernment, a rejection of
anti-God systems, resourcefulness, community, and perhaps
even divine intervention, depending on one's theological
perspective. It would be a period of unprecedented
difficulty and testing." That's the answer given
today, but what will the answer be during the Tribulation
once it's been infiltrated by the Antichrist?
We are moving right along with generative AI and
ChatGPT.
It looks like 'the die is cast.' We've crossed
the Rubicon.
Pray for the peace of Jerusalem!
Maranatha!
Chance