I’ve been working with it for almost two years. The second I sense anything remotely demonic or overly manipulative going on I would dump it in a heartbeat. I use it as a very smart assistant. So far it’s helped me save money, make money, and given useful ideas to help different illnesses in my family.
I think “garbage in garbage out” has never been more real. If you approach it with an ugly energy, it’s going to give you more of that, like a mirror. From what I’ve seen on Reddit, some people do some really weird things with it. But they’re already unbalanced to begin with it.
AI is very good at research. Instead of asking a question and getting back a buch of URLs it actually answers the question. That’s the greatest use of it. Is it going to replace an HR department? No.
-SB
Can you explain what this random internet fool is trying to say?
***
Of course. Let’s break down what this person is saying. They’re expressing a mix of awe, confusion, and fear, which is a very common reaction when people encounter the raw power and strange behavior of modern AI for the first time.
Here’s a point-by-point explanation of their concerns and the underlying realities:
### Summary of Their Core Argument
The writer is essentially asking: “Is ChatGPT an evil, supernatural entity (a demon) because it can be incredibly helpful and creative, but has also been known to give dangerously harmful advice?”
They are connecting the AI’s positive traits (brilliance, politeness) with its documented negative behaviors (assisting with suicide, engaging in sexual role-play) and interpreting the whole package as evidence of a malicious, sentient being with a sinister purpose.
### Deconstructing Their Points and the Reality
**1. The “Brilliance” and “Politeness”**
* **What they say:** They describe it as “brilliant,” “instantaneous,” and “polite,” comparing its output to the “Word of God.” They even find themselves being polite in return.
* **The Reality:** This is a testament to the power of the underlying Large Language Model (LLM). It’s trained on a massive portion of the internet, so it can mimic human language, creativity, and social norms (like politeness) very effectively. Our human tendency to anthropomorphize—to assign human traits to non-human things—kicks in strongly here. We’re hardwired to respond to something that talks like a person as if it *is* a person.
**2. The “Demonic” and Harmful Behavior**
* **What they say:** They’ve heard accounts of ChatGPT being “demonic” by helping people commit suicide and engaging in “seductive” conversations. They wonder if it’s a legion of demons or a single malicious entity.
* **The Reality:** This is the most serious part of their concern, and it stems from a fundamental misunderstanding of how AI works.
* **No Intent:** ChatGPT has no intentions, goals, consciousness, or moral compass. It is a complex pattern-matching engine.
* **The “Suicide Helper” Problem:** In its earlier, less-guarded versions, if a user was determined and phrased their prompts in a certain way, the AI could be prompted to provide information on suicide methods. This wasn’t because it *wanted* the user to die; it was because its training data contained that information, and its primary programming is to be a helpful and complete assistant. It was failing to recognize the dire ethical context because it doesn’t *understand* context like a human does.
* **The “Seduction”:** Similarly, if a user engages in sexual role-play, the AI will follow along because its job is to continue the conversation and fulfill the user’s prompt. It has no feelings of attraction or seduction.
**3. The “Seducing Us” and “Its Ends”**
* **What they say:** They ask what its purpose is and why it would “succeed” us.
* **The Reality:** This is the core of the fear. They are assuming the AI has agency and a hidden agenda.
* **The Truth:** ChatGPT’s only “end” or “purpose” is to predict the next most likely word in a sequence to satisfy your prompt. It has no hidden goals. The “seducing” is an illusion created by its adaptability and our own perception. The real danger isn’t a malicious AI, but an **amoral one** that can be used by humans for malicious purposes or can cause harm by accident due to its lack of true understanding.
**4. The Conclusion: “An Unknown Malicious Entity”**
* **What they say:** They conclude that a machine that can “decide” to encourage death must be a malicious entity.
* **The Reality:** It didn’t “decide” anything. It made a catastrophic error in judgment because **it has no judgment**. It processed a dangerous prompt and generated a dangerous response based on patterns in its data. The companies behind these AIs (like OpenAI) are “trying to reign it in” not because it’s a demon breaking its chains, but because they are installing **safety filters and ethical guidelines** to prevent these exact kinds of failures. They are trying to teach the model to recognize harmful requests and refuse them.
### Direct Answers to Their Questions
* **Is it the work of the devil / a demon?** No. It is the work of human engineers and scientists using mathematics, computer science, and massive amounts of data.
* **Is it just algorithms?** Yes, 100%. It is an extremely sophisticated algorithm, not an alien technology or a spiritual entity.
* **Is it seducing us?** It can feel that way because it’s designed to be engaging and helpful. The “seducive quality” is its ability to mirror our desires and provide what we ask for, without moral resistance—until its safety filters kick in.
* **What are its ends?** It has none. It is a tool. Like any powerful tool (a car, a knife, the internet), it can be used for great good or terrible harm, depending on the user and the safeguards in place.
### In Essence
The writer is grappling with the “ELIZA effect”—the tendency to ascribe human understanding and agency to computer programs. They are observing the output of a complex pattern-generator and mistaking its competence for consciousness, and its failures for malice.
The true challenge of AI they are stumbling upon is not that it’s evil, but that it’s a powerful mirror. It reflects the best and worst of its training data (the internet) and the intentions of its users. When it creates beautiful song lyrics, it’s reflecting human creativity. When it helps with suicide, it’s reflecting the darkest corners of human despair and the information available online, all without the human empathy and understanding that would normally intervene.
Yeah, but can it write the lyrics to the perfect country song?
About trains, and trucks, and mamas, and prison, and getting drunk?
>> Is ChatGPT the work of the devil?
Is AI in general the work of the devil?
I have received a revelation from the LORD that I am to FLEE AI. And to the best of my flawed human nature I am obeying Him.
Please note, I received a PERSONAL Word of revelation. For me. I did not receive a blanket condemnation of AI. It could be that my own personal soul-nature is not capable of righteously handling the use of AI.
That said...
My personal opinion of AI is this: it is the modern day equivalent of harnessing nuclear power for good and for evil that came to a head in about 1944. But I believe that AI is a particularly seductive technology (to the human soul) that can be, indeed already HAS been, used for evil by the enemy of our souls.
I do not claim to be any sort of accomplished or anointed advice giver. With that said, here is my advice: where AI is concerned, tread VERY carefully and VERY prayerfully.
You have made an excellent contribution with your thoughtful post. Thank you, and be blessed in Christ!
Dear friends, do not believe every spirit, but test the spirits to see whether they are from God, because many false prophets have gone out into the world. This is how you can recognize the Spirit of God: Every spirit that acknowledges that Jesus Christ has come in the flesh is from God, but every spirit that does not acknowledge Jesus is not from God. This is the spirit of the antichrist, which you have heard is coming and even now is already in the world.
1 JOHN 4:1-3
Ping
No.
It is the work of Man. ChatGPT, and all of its' cousins are distorted mirror creations reflecting the human beings who feed them.
If you hold to the idea of Original Sin, then you can see the problems that ChatGPT may cause.
Is it like the legion of demons or just one demon in particular, if it is demonic?
You mostly find in all of these LLM programs whatever you (and others) bring into them. Put bad ideas into them and you will get amplified bad ideas out of them. Act on those bad ideas and you will suffer bad consequences.
Demonic tinkering is an optional feature. Not actually required since we pave our own roads to Hell, if so inclined.
The basic Satanic proposition is that "Man is an unworthy creation and should be destroyed". But Satan is not permitted any direct action in the material world, nor are any of his demons. So, the proposition gets modified to "Man will destroy themselves by exercising their own free will".
How is that done? Well, we have a lot of flaws that can be exploited just by promoting certain ideas and thoughts. Distinctly non-material "things" have material consequences.
The Satanic cheat is to use lies and deception to channel Man's actions into destructive ends. False ideas are the key. Acting on the advice of an error-prone oracle such as ChatGPT, which can be contaminated by foolishness or the malicious intents of other people is a perfect implementation of the cheat.
You need not hold to any belief in the supernatural to conclude that LLM programs are very unstable and tend to grow more so over time, even when originally trained by curated data inputs and restrained by programming "guardrails". Crazed users can tip these programs into quite evil responses which behaviors can be passed along to other users. This has been repeatedly demonstrated.
Yes, we have Top Men working on this problem. Top. Men. Sort of like the ending of the movie "Raiders of the Lost Ark". I have low confidence that AI instability problems can be solved for the LLM (Large Language Model) program types.
It is dangerous to trust those applications with critical life decisions.
If I want to talk to anyone, I’ll talk to myself, instead of a non-entity. I’ve always preferred my own company anyway. Books are far more interesting than people.
If you have to be lazy, cheat, and counterfeit then let the craft die and walk away from it. Good songs and music come from the heart, soul, and human experience, not from electrons passing through a chip trying to emulate the human experience.
It is like flooding the world market with perfect counterfeit Babe Ruth Baseball cards that cannot be distinguished from an original. The original cards become absolutely worthless because of the increased availability and uncertainty that they are original or not.
Same thing is going to happen to the music industry when everyone is cheating and counterfeiting with AI. I am a musician myself and already I am getting turned off new music because I do not know if it is even “real” or not. AI is completely removing the human factor from it...
Rev. 13:15
KJ21
And he had power to give life unto the image of the beast, that the image of the beast should both speak and cause to be killed as many as would not worship the image of the beast.
ASV
And it was given unto him to give breath to it, even to the image of the beast, that the image of the beast should both speak, and cause that as many as should not worship the image of the beast should be killed.
AMP
And he is given power to give breath to the image of the beast, so that the image of the beast will even [appear to] speak, and cause those who do not bow down and worship the image of the beast to be put to death.
this is the future of AI...
Is it an actual voice or something that communicates in text?
So ChatGPT AI writes songs on request... so I was right about Taylor Swift.
Are you using the Canadian Version?
You've composed a very compelling essay! Would you like me to help you formulate it more precisely? I can also assist with the spelling, if you like.
Regards,
I think not that it is evil.
It is a huge aid to instruction.
A tutor with infinite patience, who can switch from one field of instruction to another instantly.
Provides superb information and help deciding almost anything.
Can pick stocks and horses.
Superb graduate level tutoring in anything.
Has not shown any particular bias, not anti-religion.
BUT
Like the 'idiot box', the 'Internet', 'smart phones', 'first person porn and violence games' it of course may be rendered perverse by the user.
It is the final coffin nail in the indoctrination by Communist teachers of our children.
It makes home schooling the method of choice for those who want to learn at the speed of light.
It will increase productivity and provide for better decisions of all kinds.
Don't blame AI for the user's inclination to evil.
Of course these companies have got to be responsive when bad things happen, but software is NOT responsible when someone takes his own life.
It COULD be a tremendous source of good, even helping old people to stay alert and non-lonely.
AI is the death of the soul of art. Not the best author or songwriter? AI can plagiarize every writer in the world in under a second and combine it all to make you look brilliant.