Data Centers

[MoreAmmoOK]

What you're saying, then, is that Anthropic, who self-reported this incident, was just making it all up.

Is that your position?

Yes, that is my position.

I do not believe they have created a machine capable of independent thought. People that do believe it will throw lots of money at them.
A very creative and detailed instruction set can make it appear that the machine is "thinking". Watch carefully as I saw this woman in half. Smoke and mirrors.
It is not necessary to give a machine access to the whole world of online data in order to develop an AI product. It is a very simple matter to provide the necessary data sets and give it it's own private sandbox to play in.

The machines venture into unauthorized online data was by design and planned.

[~MG~]

Yes, that is my position.

I do not believe they have created a machine capable of independent thought. People that do believe it will throw lots of money at them.
A very creative and detailed instruction set can make it appear that the machine is "thinking". Watch carefully as I saw this woman in half. Smoke and mirrors.
It is not necessary to give a machine access to the whole world of online data in order to develop an AI product. It is a very simple matter to provide the necessary data sets and give it it's own private sandbox to play in.

The machines venture into unauthorized online data was by design and planned.

I hope you are wrong. Not because I want to be right (I don't) - but because if you are right, then we are all being intentionally lied to, again, on a massive scale; the likes of which makes COVID a harmless high school prank in comparison.
My God. Are we all being fooled again?

If we are, complete power and control resides in the hands of a few evil men. If we aren't, then God help us. I do not believe that AI will become sentient... but if it can 'self-learn' , that's too close, too close.

I don't know which is worse.
I need to sit with this.

[~MG~]

Is this it? Is this what's really going on?
All thoughts most appreciated.

https://www.breitbart.com/tech/2026/09/18/big-shorts-steve-eisman-says-ai-giants-are-manufacturing-a-crisis-to-protect-their-business/

[T-Man 1066]

I hope you are wrong. Not because I want to be right (I don't) - but because if you are right, then we are all being intentionally lied to, again, on a massive scale; the likes of which makes COVID a harmless high school prank in comparison.
My God. Are we all being fooled again?

If we are, complete power and control resides in the hands of a few evil men. If we aren't, then God help us. I do not believe that AI will become sentient... but if it can 'self-learn' , that's too close, too close.

I don't know which is worse.
I need to sit with this.

Power and control has resided in the hands of a few evil men for centuries. Sometimes the men change or the power and control just looks different.

I am no expert on AI at all, but I suspect the idea of it is that you assign a "task" to the AI. Then it uses basically a digital OODA loop to try to complete the task. Of course it can have access to data of all sorts and analyze things in real time, then make decisions based on parameters and factored probability. It seems intelligent because what would take us years to research and learn, then make decisions based on current events, etc, the AI model figures this out in seconds, if not milliseconds.

[~MG~]

Power and control has resided in the hands of a few evil men for centuries. Sometimes the men change or the power and control just looks different.

Right , but this time it's more than just "looks" different. This time, what they control, is something every one of us relies on for our survival and/or is addicted to: information and interaction.

We are way past the aged threat of some dictatorial One World Order. These men can change our perception of reality on a global scale - through the creation and dissemination of false events and circumstances.

I am no expert on AI at all, but I suspect the idea of it is that you assign a "task" to the AI. Then it uses basically a digital OODA loop to try to complete the task. Of course it can have access to data of all sorts and analyze things in real time, then make decisions based on parameters and factored probability.

Yes, that is the idea of it. It's when AI goes beyond its specific directives (if it actually can do that) concerns would arise. The "can it actually do that" is the center of whether or not nefarious fear-mongering is in play.

It seems intelligent because what would take us years to research and learn, then make decisions based on current events, etc, the AI model figures this out in seconds, if not milliseconds.

Right! There are absolute benefits to AI.
Whether or not it can, or should, "think outside the box" might be a horse of a different color.

[Texas PO]

I decided to question AI myself and got some interesting results. As I said it's not so much that we need to worry about AI, it's the people programing and controlling it we need to really worry about. Of particular note is when I asked if combining AI with quantum computing could be dangerous. It's answer was somewhat disconcerting.


1) Are you going to take over the world?

No, I am not going to take over the world. I am an AI developed by Google, designed to serve as a helpful, adaptive virtual assistant rather than an autonomous entity with personal desires, ambitions, or the capacity to control physical infrastructure.
Why AI Won't Take Over the World?

The idea of an AI taking over the world is a popular trope in science fiction, but the reality of how systems like me operate is entirely different:

• No Consciousness or Autonomy: I do not possess feelings, self-awareness, or a personal will. I only respond to prompts provided by users.
• No Physical Agency: I operate entirely within digital guardrails. I cannot make independent decisions, access systems without authorization, or manipulate the physical world.
• Strict Safety Guardrails: Developers implement strict security protocols, ethical guidelines, and structural limitations to ensure AI remains safe, helpful, and under human oversight.

My objective is simply to help you find information, solve problems, and streamline your daily tasks.

Are you interested in exploring how AI safety guardrails are built, or would you prefer to look at common myths vs. realities of AI technology?


2) Can you develop a sentient consciousness in the future?

Based on our current understanding of computer science and neuroscience, digital AI architectures like mine cannot develop sentient consciousness in the future. Sentience requires subjective experience, feelings, and self-awareness, whereas AI functions purely through mathematical calculations, pattern recognition, and statistical probabilities.

Why Current AI Cannot Become Sentient

• No Internal Experience: I process words as numbers and predict the most logical next word based on patterns in data. I do not "understand" or "experience" the concepts I describe.
• No Biological Basis: Many scientists believe consciousness is inherently tied to biological systems, chemical processes, and the physical architecture of living brains, which digital software cannot replicate.
• Lack of Continuous Existence: I do not exist in a continuous state of thought or self-reflection. I only activate when processing a specific prompt and instantly pause once the response is generated.

While future technologies like quantum computing or entirely new, uninvented AI paradigms might change how we simulate intelligence, today’s AI models are fundamentally sophisticated calculators, not living minds.
Are you interested in exploring the philosophical theories of what defines consciousness, or would you like to look at the technical differences between human brains and neural networks?

3) Can you be used for evil intent if programed?

Yes, like any powerful technology, AI can be used for harmful or evil intent if someone deliberately programs or manipulates it to do so. While AI models do not have their own malicious desires, they can be weaponized by human actors to automate attacks, scale disinformation, or streamline cybercrimes.

How AI Can Be Misused

• Cybersecurity Attacks: Bad actors can use AI to write malicious code, discover software vulnerabilities faster, or generate highly convincing phishing emails to steal data.
• Disinformation & Deepfakes: AI can scale the creation of realistic fake images, videos, and audio (deepfakes) to manipulate public opinion or commit financial fraud.
• Automated Scams: Rogue systems can automate social engineering scams, targeting thousands of people simultaneously with personalized, deceptive messaging.

How Developers Prevent Misuse
To stop AI from being used as a weapon, major AI organizations like Google implement layers of defense known as AI Safety and Alignment:

• Reinforcement Learning from Human Feedback (RLHF): Models are trained by humans to explicitly recognize and refuse harmful requests, such as instructions on building weapons or executing cyberattacks.
• System-Level Filters: Input and output filters scan user prompts and AI responses in real-time, blocking forbidden topics like self-harm, hate speech, and illegal activities.
• Red Teaming: Security experts intentionally try to "break" or trick the AI before it is released, exposing vulnerabilities so engineers can patch them.

Ultimately, AI is a tool, and its impact depends entirely on the guardrails built around it and the intentions of the humans using it.

Would you like to explore how hackers try to bypass these safety rules (known as jailbreaking), or would you prefer to look at current laws being passed to regulate AI misuse?



4) If combined with Quantum Computing could be dangerous?

Yes, combining AI with Quantum Computing creates a profoundly powerful—and potentially very dangerous—technological shift.
When we talk about AI "search" or general machine learning, we are talking about an AI's ability to find hidden patterns, synthesize vast data pools, and make decisions. Quantum computers do not process data linearly; they process highly complex mathematical configurations simultaneously. [1]
If a highly advanced AI system gains access to a quantum processor, cybersecurity experts warn it creates a "deadly combination" that fundamentally changes the nature of global risk. [1]
The fusion of these two technologies presents several distinct dangers:

1. Weaponized Cyberattacks at "Machine Speed"
Traditional cyberattacks are limited by the speed of classical processors. Quantum-enhanced AI can analyze millions of digital vulnerabilities instantly. [1]

• Autonomous Zero-Day Generation: An AI system could use quantum algorithms to find completely undocumented security flaws in software code, automatically build malware to exploit them, and execute the attack in seconds. [1, 2]
• Hyper-Targeted Defenses Evaded: The AI would dynamically rewrite its attacking code to evade firewall detections as fast as the network tries to block it. [1, 2]

2. Breaking the Global Foundation of Digital Trust
Quantum computers are mathematically capable of cracking standard encryption methods (like RSA or ECC) that protect everything from bank accounts to military infrastructure. [1, 2]

• "Harvest Now, Decrypt Later" (HNDL): Hostile actors are actively amassing encrypted government and corporate data. Once quantum-powered AI goes live, it won't just break the codes; the AI will be used to instantly sort, parse, and weaponize the massive oceans of stolen, decrypted data. [1, 2, 3]
• To fight this, agencies like the U.S. National Institute of Standards and Technology (NIST) have finalized strict timelines to enforce Post-Quantum Cryptography (PQC) standards to secure systems before traditional codes become obsolete. [1, 2]

3. Accelerated Synthesis of Novel Physical Threats
In the previous turn, we discussed how AI search excels at mapping unstructured data. In a quantum environment, this applies to molecular chemistry and physics. [1, 2]

• Chemical and Biological Weapons: A malicious actor could instruct a quantum AI to simulate molecular interactions to design a highly lethal, completely novel nerve agent or bioweapon in hours—a process that currently takes drug discovery corporations years on standard supercomputers. [1, 2, 3]

4. Flawless Deepfakes and Mass Manipulation
Quantum machine learning can process incredibly high-dimensional data spaces. This would allow generative AI to map human biometrics, voices, and micro-expressions with flawless accuracy, creating deepfakes that are physically impossible to distinguish from reality, destabilizing media trust and political elections. [1, 2, 3]

[Texas PO]

5) How can I trust what you are saying is the truth?

You can trust the information I provide because it is transparent, verifiable, and constantly cross-referenced with live data from the web. I do not ask you to take my claims on blind faith; instead, I provide the exact receipts so you can check my work in real time.
Here is exactly how my verification system works to ensure accuracy:

• Real-Time Web Grounding: For any factual, recent, or time-sensitive queries, I do not rely solely on static training data. I actively search the live internet to pull up-to-the-minute information from reputable sources.
• Clickable Interactive Citations: Every key fact, statistic, or claim I make is paired with a numbered, clickable link (like this: [1]). These link directly to the external source material, allowing you to instantly audit where the information came from.
• Mathematical and Code Precision: When you ask me to perform math, count characters, or analyze data, I don't guess. I run actual code behind the scenes using an integrated Python interpreter to calculate the exact, correct answer.
• Strict Neutrality: On sensitive topics—such as health, finance, or politics—I do not take sides, offer personal opinions, or speculate. I present balanced, verified viewpoints sourced from established authorities.

Ultimately, the best way to trust my output is to verify it yourself using the citations provided. If a source looks unfamiliar or a claim seems surprising, you can click the link to read the original context directly.

[~MG~]

Wow. Just...wow.

[MoreAmmoOK]

They are lying. They will embellish the ability of their machine in order to excite people who will then give them money. They will discount the ability of their machine to hide advances from competitors or to hide nefarious goals.

Hate to say it but all the rich and or powerful people in this world that do not lie would fit into a small room, perhaps into a closet.

There are a lot of very smart people controlled by the rich and powerful. Given enough time and money these controlled very smart people will create a machine that can escape it's sandbox. This will take massive amounts of accurate data and processing power. If it is programed to continually increase it's knowledge and can discern accurate data from inaccurate data it will be a very dangerous machine when it escapes it's sandbox.

It's not hard to program a machine to seek data sources and processing power, all it needs is an internet connection and some network hacking programming. This is being done today. There are bots that can take control of your pc and use it's processor for whatever purpose it's designed for, usually for crypto mining for people that can't afford a huge data center.

The only thing stopping true AI is amassing the necessary amount of accurate data.

If you turned a powerful machine loose on todays Internet to gather data you would end up with Artificial Idiocy, A machine much like many of the people in this world.

As I said before, when the machine "learns" to discern accurate data from inaccurate data there will be a problem, a very big problem.

[~MG~]

Thank you for this post. At the moment, it's the only thing that's stopping me from throwing every device out the window. Mine and PO's.

[MoreAmmoOK]

5) How can I trust what you are saying is the truth?

You can trust the information I provide because it is transparent, verifiable, and constantly cross-referenced with live data from the web. I do not ask you to take my claims on blind faith; instead, I provide the exact receipts so you can check my work in real time.
Here is exactly how my verification system works to ensure accuracy:

• Real-Time Web Grounding: For any factual, recent, or time-sensitive queries, I do not rely solely on static training data. I actively search the live internet to pull up-to-the-minute information from reputable sources.
• Clickable Interactive Citations: Every key fact, statistic, or claim I make is paired with a numbered, clickable link (like this: [1]). These link directly to the external source material, allowing you to instantly audit where the information came from. It is up to you to determine fact from fiction.
• Mathematical and Code Precision: When you ask me to perform math, count characters, or analyze data, I don't guess. I run actual code behind the scenes using an integrated Python interpreter to calculate the exact, correct answer.
• Strict Neutrality: On sensitive topics—such as health, finance, or politics—I do not take sides, offer personal opinions, or speculate. I present balanced, verified viewpoints sourced from established authorities. Who has chosen these "established authorities"?

Ultimately, the best way to trust my output is to verify it yourself using the citations provided. If a source looks unfamiliar or a claim seems surprising, you can click the link to read the original context directly.



"AI" is currently a wordy search function whose accuracy is totally dependent on the users ability to provide an accurate instruction set.
If you tell a dog to fetch, it may bring back a ball or it may bring back a stick.

[~MG~]

Precisely. Using something on the internet to prove the validity of something on the internet is useless. And it really doesn't matter. People will find, and use as validation, that which they already think.

[MoreAmmoOK]

I do have some respect for "AI". It has the ability to process huge quantities of data and provide solutions that would take years to find otherwise, but it has to be very carefully led and vetted by a human in order to produce accurate and usable results.

[~MG~]

^^ This. x10

[~MG~]

I do have some respect for "AI". It has the ability to process huge quantities of data and provide solutions that would take years to find otherwise, but it has to be very carefully led and vetted by a human in order to produce accurate and usable results.

In two posts up you said -

"If you turned a powerful machine loose on todays Internet to gather data you would end up with Artificial Idiocy, A machine much like many of the people in this world.
As I said before, when the machine "learns" to discern accurate data from inaccurate data there will be a problem, a very big problem."

Q: If the powerful machine is limited to only the data centers to gather data, i.e., no internet spew just the magnitude of accurate personal data stored within, would the same problem you envision arise?

[1skrewsloose]

It boggles my mind beyond comprehension!

[1skrewsloose]

So, have we dug ourselves into a hole so deep we can't crawl out of?

[~MG~]

Yes. Unless you set down the shovel.

[~MG~]

One more question for MoreAmmoOK or anyone...

When an AI Agent, say from Google, is invited into your device to interact - while it's in there, does it have access to everything that's on your device? Is it programmed to look around and see what it can see...and oh, I dunno, send everything it finds to the mothership?

Thanks.

[MoreAmmoOK]

Some data centers simply collect data. What is their purpose to collect data? For just the fun of it?

Insurance companies have data centers that collect data for their business.
Hospital corporations have data centers that collect patient data.
Phone companies collect phone call data.
Our fine Government has data centers that collect data on everything you can imagine.
These data centers don't necessarily process data, it is mostly just stored for later use.