Here’s an interesting perspective on intuition

[Salvijus]

1 hour ago, integral said:

If you ask it what's 100 + 101 it will guess.

It will write the answer that it was programmed to write. Wtf. For real. 

Edited Sunday at 06:08 PM by Salvijus

[integral]

3 minutes ago, Salvijus said:

It will write the answer that it was programmed to write. Wtf. For real. 

You're using language the way you were programmed to use language?

[Salvijus]

1 hour ago, integral said:

You're using language the way you were programmed to use language?

I use the structure of a particular language to express myself. That structure is always the same. 

Ai follows the structure of a certain computing code. That's its language. 

Edited Sunday at 06:21 PM by Salvijus

[Leo Gura]

1 hour ago, Salvijus said:

It will write the answer that it was programmed to write. Wtf. For real. 

No. It is intuitive.

Mankind has succeeded in reverse engineering intuition. That's the mindfuck of our time.

Edited Sunday at 07:54 PM by Leo Gura

[Joshe]

10 hours ago, Salvijus said:

It will write the answer that it was programmed to write. Wtf. For real. 

You misunderstand what current AI is. LLM stands for large language model. The whole point of it is to predict it’s next output based on a given input. They trained it on language and had it make millions of guesses and the programmers told it if the guesses were accurate. Over time, it’s predictions got better and better until eventually, it could learn unsupervised.

Im not the best to communicate it but AI is not programmed the way a calculator is. That would be simple. If you think an AI is the same as a calculator but just scaled up for every possible question, then you will be shocked when you learn what it actually is. They don’t even know how it does what it does. They can’t trace all its processing. They call that “the black box”. It literally is like a giant mind. It wasn’t trained to produce a specific output, it figures it out on the fly, guided by parameters of course. For example, they have it behave friendly. They did program that in. 

Look up neural nets, machine learning, and natural language processing, or better yet, just ask AI to explain the mechanics of itself. 

# Mechanics of AI Large Language Models (LLMs)

## Architecture
- Based on Transformer architecture introduced in 2017 (Vaswani et al.).
- Key components:
  - Encoder-Decoder structure (e.g., BERT).
  - Most LLMs like GPT use only the decoder.
- Utilizes self-attention mechanisms to process context.

## Tokenization
- Text is broken down into tokens (words, subwords, or characters).
- Each token is mapped to a numerical vector using embeddings.

## Training
- Trained on massive datasets containing text from books, websites, and other sources.
- Objective:
  - Predict the next token in a sequence (causal modeling).
  - Fill in blanks (masked modeling, e.g., BERT).
- Loss function: Measures the difference between predicted and actual tokens during training.

## Context Window
- LLMs have a fixed "context window" that limits how much text they can process at once.
- Larger context windows allow processing of longer texts but increase computational cost.

## Attention Mechanism
- Self-attention allows the model to focus on relevant parts of the input sequence.
- Calculates relationships between tokens to understand context and semantics.

## Layers and Parameters
- Consist of stacked transformer layers, each containing:
  - Attention heads.
  - Feed-forward networks.
- The number of parameters (weights) determines model size and capability (e.g., GPT-3 has 175 billion parameters).

## Pretraining
- Models are trained on general tasks using unsupervised learning.
- Enables them to learn patterns, grammar, facts, and reasoning.

## Fine-tuning
- Optionally adapted to specific tasks (e.g., sentiment analysis, code generation).
- Done using supervised learning on domain-specific data.

## Inference
- At runtime:
  - Given an input prompt, the model predicts and generates the next token iteratively.
  - Continues until it forms a complete response.

## Scaling
- Larger models tend to perform better but require significant computational resources.
- Scaling laws predict improvements with:
  - More data.
  - Larger models.
  - Longer training times.

## Limitations
- Prone to generating plausible but incorrect or nonsensical answers (hallucinations).
- Lack common sense and grounding in real-world truth unless externally validated.
- Memory is limited to the input/output within the context window; lacks long-term memory.

## Applications
- Can perform tasks like:
  - Text generation.
  - Summarization.
  - Translation.
  - Coding.
  - Question answering.
- Used in industries such as education, customer service, and research.

Edited yesterday at 05:08 AM by Joshe

[Salvijus]

Hmm, interesting points being made here.

Okay, let's say ai is a perfect guessing machine. But perfect guessing is not intuition either. Intuition does not guess, it KNOWS beyond guessing, it KNOWS. Like conciousness. Intuition and conciousness go together.

Being good at guessing and knowing something with certainty, with the certainty of God itself are not the same thing. That's why I'm skeptical of Ai having intuition or Love or conciousness etc. But I don't fully understand your position of what makes you say what you say so I can't give a proper counter argument. 

@Joshe I enjoyed reading your take btw.

Edited yesterday at 06:34 AM by Salvijus

[Leo Gura]

@Salvijus You have turned intuition into a dogma. So there is no reasoning with you. You believe it is infallible and so you are like a terrorist who believes Allah perfectly rewards their terrorism. All discussion ends there.

[Salvijus]

1 hour ago, Leo Gura said:

@Salvijus You have turned intuition into a dogma. So there is no reasoning with you. You believe it is infallible and so you are like a terrorist who believes Allah perfectly rewards their terrorism. All discussion ends there.

Yea but my arguments are more solid in the end. 

To guess something with high precision based on massive amounts of previous data and pattern recognition and to know something with perfect certainty and resonance of the heart while relying on no past data at all are two very different things. 

Edited yesterday at 07:02 AM by Salvijus

[integral]

25 minutes ago, Salvijus said:

Yea but my arguments are more solid in the end. 

To guess something with high precision based on massive amounts of previous data and pattern recognition and to know something with perfect certainty and resonance of the heart while relying on no past data at all are two very different things. 

Can you give me examples of when this happens?

[Leo Gura]

@Salvijus You have a dogma, not an argument.

[Salvijus]

1 hour ago, Leo Gura said:

@Salvijus You have a dogma, not an argument.

I thought that's a pretty good argument. 

"To guess something with high precision based on massive amounts of previous data and pattern recognition and to know something with perfect certainty and resonance of the heart while relying on no past data at all are two very different things." 

 

[Thought Art]

@Salvijus I don’t get it what are you arguing?

[Salvijus]

1 hour ago, integral said:

Can you give me examples of when this happens?

Intuition means to know something by resonance. 

The simplest example is the musical one that I gave before. Each note revebarates at a certain frequency. When one of those notes is off, it is jarring to the system and you know it is off not by intellectual analysis but by feeling the resonance of sounds. Same way Truth is a certain frequency. That frequency is love. And intuition means you can tell what is in harmony with love and what is not in harmony with love. That way your intuition shows you the way to every decision. For example, how do animals know what to eat? Did they study nutrition course? Or do they know it in their heart and need not use any mental faculties to know?  

Edited yesterday at 07:35 AM by Salvijus

[Salvijus]

1 hour ago, Thought Art said:

@Salvijus I don’t get it what are you arguing?

I have no idea myself. It's complicated lol

[Thought Art]

@Salvijus Just use your intuition . 
 

Edited yesterday at 07:35 AM by Thought Art

[gambler]

6 minutes ago, Salvijus said:

I thought that's a pretty good argument. 

"To guess something with high precision based on massive amounts of previous data and pattern recognition and to know something with perfect certainty and resonance of the heart while relying on no past data at all are two very different things." 

Why are you assuming intuition isn’t based on previous data? If you edited the genes of a human to be mentally like a fish, what would the fish mind be able to intuit? But if you concede it wouldn’t intuit like a human, then why do you have a problem with intuition stemming from pattern recognition from genetic programming or evolutionary information (as a baby you intuitively know how to learn a language and have innate fears), as well as experiences and learned knowledge? If data isn’t the reason why you can intuit, then what is the reason that if you were a fish, you wouldn’t intuit like a human? 

[Thought Art]

4 minutes ago, Salvijus said:

Intuition means to know something by resonance. 

The simplest example is the musical one that I gave before. Each note revebarates at a certain frequency. When one of those notes is off, it is jarring to the system and you know it is off not by intellectual analysis but by feeling the resonance of sounds. Same way Truth is a certain frequency. That frequency is love. And intuition means you can tell what is in harmony with love and what is not in harmony with love. That way your intuition shows you the way to every answer. For example, how do animals know what to eat? Did they study nutrition course? Or do they know it in their heart and need not use any mental faculties to know?  

1. What if what is true feels uncomfortable, or what if it requires critical thinking?

2. Animals have instincts that’s not really intuition 

[Salvijus]

1 hour ago, Thought Art said:

@Salvijus Just use your intuition . 
 

My intuition says it's just ego battles happening lol

[Thought Art]

2 minutes ago, gambler said:

Why are you assuming intuition isn’t based on previous data? 

Intuition is this essentially…

[Thought Art]

4 minutes ago, Salvijus said:

My intuition says it's just ego battles happening lol

I’ve learned never to argue with you about anything. Even if I am not arguing you will argue with me. 

Edited yesterday at 07:43 AM by Thought Art