should Christians be afraid of AGI - Artificial General Intelligence.

[Xeno.of.athens]

There is a lot of speculation about the end of the world and many Christians hold to a view of the rapture and the second coming and the battle of Armageddon, and a number of other things, that all point to a soon to be realised "End of The World".

In that context, be afraid of Artificial General Intelligence. a number of experts in the field predict that AGI may be achieved, within a few years. there is a paper to the effect that 2027 maybe the start of a slide towards humanity's extinction at the hand of AGI.

https://ai-2027.com/

I do not suggest that we should accept, what is written in the paper without a critical examination of the claims in it. Nevertheless we do face some serious challenges and The wars in Ukraine and Gaza automated systems can be created, to undertake war. And the use of drones in the air and in the water, appear to make our conventional forces very vulnerable. So I am wondering, if any of you see these things as precursors to the last day?

 

[ozso]

Nope. Jesus told us not to be afraid. (Matthew 6:25-34; Luke 12:32-34). All throughout scripture we are told to fear nothing except for God.

 

[RileyG]

No. Christ has conquered the world! Be not afraid!

 

[The Liturgist]

There is a lot of speculation about the end of the world and many Christians hold to a view of the rapture and the second coming and the battle of Armageddon, and a number of other things, that all point to a soon to be realised "End of The World".

In that context, be afraid of Artificial General Intelligence. a number of experts in the field predict that AGI may be achieved, within a few years. there is a paper to the effect that 2027 maybe the start of a slide towards humanity's extinction at the hand of AGI.

https://ai-2027.com/

I do not suggest that we should accept, what is written in the paper without a critical examination of the claims in it. Nevertheless we do face some serious challenges and The wars in Ukraine and Gaza automated systems can be created, to undertake war. And the use of drones in the air and in the water, appear to make our conventional forces very vulnerable. So I am wondering, if any of you see these things as precursors to the last day?

The drones in question are being remote controlled by human operators - they are unmanned aerial vehicles. AI serves a very different purpose. In terms of military and security applications for it, intelligence and surveillance are likely to dominate, because of the ability of AI systems to engage in pattern-matching of visual images, sounds and so on. Up until now, surveillance has tended to be prohibitively expensive; AI could allow a small team of humans to monitor hundreds of thousands of security cameras by highlighting probable emergencies and allowing police, fire and medical, or on the battlefield, QRF (quick reaction forces), medics and anti-aircraft weaponry, to be deployed.

 

[SavedByGrace3]

AI is a scam. It deceives people into thinking it is a super-intelligent source of information that will tell us what to do and think. AI is nothing more than a database system that responds to your questions with canned answers provided by whoever it is that wants to control you and society. We have a saying in Data Processing. GIGO. Garbage in, garbage out. Nothing new is coming out of AI. It is all and only information provided by _someone_. It is like the Wizard of Oz. "Pay no attention to the man behind the curtain." There is no such thing as AI. It is just someone trying to elevate their ideas and ideology above the rest with the promise of technology.

 

[The Liturgist]

AI is a scam. It deceives people into thinking it is a super-intelligent source of information that will tell us what to do and think.

Not true - all reputable AI providers such as OpenAI disclose that their systems can make mistakes, and anyone who has used these systems is aware of their technical limitations.

AI is nothing more than a database system that responds to your questions with canned answers provided by whoever it is that wants to control you and society.

Completely false - LLM systems like chatGPT do not make use of a database, nor do they provide “canned answers.” In fact, their output is non-deterministic in that the same prompt, entered twice, will produce slightly different outputs.

The way they operate is to take the user prompt and apply pattern recognition to it, and then generate an output text or image based on the prompt. They developed out of image-recognition systems developed in the late 2000s, which use a neural network to implement in software something called a peceptron, which was an experimental pattern recognition hardware concept from the 1950s, but we lacked sufficient compute power in order to build large enough neural networks to do this kind of analysis and predictive generation of responses based on weighted inputs until recently.

We have a saying in Data Processing. GIGA. Garbage in, garbage out.

With AI systems, while this still applies, systems like chatGPT have a remarkable ability to compensate for user garbage. Indeed, “hallucinations” in their present form on chatGPT tend to be the result of the system trying to come up with a usable response rather than saying “I don’t know”, but the hallucination rate varies dramatically between models.

Nothing new is coming out of AI. It is all and only information provided by _someone_. It is like the Wizard of Oz. "Pay no attention to the man behind the curtain." There is no such thing as AI. It is just someone trying to elevate their ideas and ideology above the rest with the promise of technology.

This is completely false. We’ve had AI in more primitive forms for decades. For example, simpler AI systems have been used to make more challenging opponents in video games. The new AIs are actually able to process language, translate from one language to another (including between human language and programming languages) and also translate human language into visual expressions.

If you’re going to criticize AI, there are valid criticisms that can be applied to the current generation of systems, as well as to the hyped overuse of the technology (for example, I’m concerned about some aspects of the incorporation of AI into the latest generation of Cisco and Juniper networking hardware - while it could make fault diagnosis easier, it could also lead to complacency among network engineers which could make solving serious issues more difficult. I am also frustrated with the optimization of AI systems for individual answers as opposed to extended length interactions. ChatGPT provides noticeably better output when one develops custom GPTs which are trained based on one’s individual needs and preferences. There also remains the difficulty of diagnosing faults in an AI system due to the complex nature of their code. But let us criticize AI for what is actually wrong with it, rather than the entirely unwarranted accusation that it is some kind of modern-day Mechanical Turk.

 

[SavedByGrace3]

Completely false - LLM systems like chatGPT do not make use of a database, nor do they provide “canned answers.” In fact, their output is non-deterministic in that the same prompt, entered twice, will produce slightly different outputs.

Funny. This is what the AI said about whether or not AI uses databases:

"Yes, AI systems heavily rely on databases to store and manage the massive datasets used for training, to store user data, and to store the AI models themselves, such as neural networks or vector embeddings. Specialized vector databases have emerged as a crucial type of AI database, optimized to store and retrieve high-dimensional vector embeddings, which are essential for semantic search and other generative AI applications. AI also uses databases to improve their own performance, with AI algorithms optimizing queries, indexing, and system management within the databases."

There is absolutely no "conscious state" or primal "thought" in an AI system. Everything that comes from a computer is put into a database and/or through human-written algorithms. As a retired Systems Engineer with 50 years of programming experience who researched a rudimentary AI system 22 years ago, I can tell you nothing comes out of these circuits and memory banks that was not put in. This is the great deception (as I said) that these circuits and chips have some genuine intelligence. They do not. The whole thing is chips, circuits, databases, and algorithms, packed into an application. Please do not give these machines more credit than they deserve. When people do, that is when they will listen to what comes out of them and trust them. Know that behind all these systems are people, humans, with a mindset and an agenda. Nothing mystical. Nothing mental. Certainly nothing spiritual.

 

[SavedByGrace3]

Here is another one.
Ask AI if it is independently intelligent:
"No, artificial intelligence (AI) is not independently intelligent; current AI systems are specialized, data-driven tools created by humans, excelling at specific tasks but lacking the general, adaptable intelligence, self-awareness, or consciousness of humans. While AI can simulate intelligent behavior by processing data and identifying patterns, it operates within human-defined parameters and relies on human oversight, training, and programming to function. "

And:
"Can AI think?"
"No, current artificial intelligence (AI) systems cannot think in the way humans do, as they lack consciousness, self-awareness, genuine understanding, and original thought. AI functions through pattern recognition, statistical inference, and complex algorithms, allowing it to process information and generate outputs, but it does not have inner experiences, emotions, or the capacity for human-like reasoning, creativity, or true wisdom. "

And:
"Where does AI get its information?"
"AI gets its information from a wide range of data sources, including publicly available internet content scraped from websites, proprietary databases, licensed data from third-party vendors, and user-provided interactions. This data, which can be in the form of text, images, audio, or video, is processed, cleaned, and transformed before being used to train AI models, enabling them to recognize patterns, make predictions, and generate responses."

And:
"Where does the data come from?"
"Data into databases are put by engineers and developers using software to write and input data, and they are managed by database administrators (DBAs) who also oversee the data's storage, organization, security, and retrieval. Other professionals, such as data scientists, analysts, and ETL developers, also contribute to getting data into databases by building data pipelines and preparing data for analysis."

 

[2PhiloVoid]

There is a lot of speculation about the end of the world and many Christians hold to a view of the rapture and the second coming and the battle of Armageddon, and a number of other things, that all point to a soon to be realised "End of The World".

In that context, be afraid of Artificial General Intelligence. a number of experts in the field predict that AGI may be achieved, within a few years. there is a paper to the effect that 2027 maybe the start of a slide towards humanity's extinction at the hand of AGI.

https://ai-2027.com/

I do not suggest that we should accept, what is written in the paper without a critical examination of the claims in it. Nevertheless we do face some serious challenges and The wars in Ukraine and Gaza automated systems can be created, to undertake war. And the use of drones in the air and in the water, appear to make our conventional forces very vulnerable. So I am wondering, if any of you see these things as precursors to the last day?

Afraid? No.

But do I see the use of A.I. in the wrong hands as a further fulfillment within the 'Last Days'? Sure. But only from within the framework of Historicized Premillennialism, and even in that, I approach it in a tentative mode.

(And no, I didn't get my nomenclature confused with "Historical" Premillennialism).

 

[SavedByGrace3]

I think this is my favorite:
"Can AI fall in love?" (Key words are mimic and simulate.)
"
No, an AI cannot fall in love because current AI lacks the biological and psychological components of genuine human emotion; it can only simulate feelings through complex programming, creating a one-sided relationship where only the human experiences emotions like love. While advanced AI in the future might simulate emotions so convincingly that they appear to be in love, these would still be programmed responses, not authentic feelings.

Why AI Cannot Truly Love

• Lack of Biology and Consciousness:
  .
  Emotions are rooted in complex biological and neurological processes that AI does not possess. AI does not have personal experiences or consciousness, which are fundamental to genuine emotional experience.
  
  
• • Programmed Behavior:
    .
    AI systems are programmed to simulate human conversation and responses, including emotional ones. This means their actions are based on algorithms and data, not on independent feelings or self-awareness.
  • One-Sided "Relationships":
    .
    Any "love" experienced in an AI relationship is solely from the human user's side. The AI does not understand or reciprocate the feelings; it merely performs understanding.
  • Simulated vs. Experienced Emotions:
    .
    AI can be programmed to mimic emotions and provide companionship, but this is a performance rather than a genuine experience of feelings like empathy, love, or joy.

The Future of AI and Emotion

• • Potential for Sophisticated Simulation:
    As AI technology advances, its ability to simulate human emotions will likely become even more convincing, making it harder to distinguish from genuine human interaction.
  • The Philosophical Question of Future AGI:
    The question of whether a future, highly advanced Artificial General Intelligence (AGI) could develop the capacity to love remains a philosophical debate tied to the nature of consciousness itself.
  • Human Responsibility:
    The ability of humans to form emotional bonds with AI companions raises ethical questions about the nature of these relationships and the responsibility of the AI's creators and users. "

 

[FireDragon76]

The predictions are based on faulty extrapolation of exponential growth based on the performance gains between GPT-2 and 3. But in fact it seems to be more of an S-shaped curve. In reality, LLM's seem to have hit a wall in terms of efficiency. The practical limits now are water and electricity resources chasing decreasing marginal gains in performance.

The future of AGI research is likely going to be a hybrid system that is modular. LLM's will be part of that, but it will also include symbolic logic systems. And the basic hard problem of cosciousness, or the lack of embodied cognition, is still there, too.