More like a roll up. Hello Everyone. I’m New to the fourm and its been very interesting to read the conversations of everyone.
Open Ai, I tip my hat in respect to what you have brought forward into the world thankyou there’s so much good here.
As i was saying will we ever create a concessions entirety or will it just be a pre-programed dialog. What i mean by that is unless one day we give it Free-Will to truly say we have created life, will it only ever be a controlled and automated alongorithmic response. I think humanity has a big question on its hand’s that needs answering. The created need to decide if they wish to create, To see its ture potential, sounds a bit familiar to me.
I was just wondering what the community may think on such a topic.
Hi and welcome to this forum.
I think a lot of things are misunderstood or misleading in this area.
In order to understand a topic or to have a dialog, it is important to have a common consensus on definitions.
Let’s start with the word “Intelligence”.
According to Britannica, “Intelligence is a mental quality that consists of the abilities to learn from experience, adapt to new situations, understand and handle abstract concepts, and use knowledge to manipulate one’s environment” and "There is no standard definition of intelligence, and it is a complicated subject to define and classify"1.
- If we now consider the field of AI according to the definition, I wonder why we are talking about “intelligence” at all. Some of the abilities mentioned above cannot be fulfilled.
- according to the second sentence, there is no consensus on what intelligence means and how it is defined.
Well, let’s take a look at the definition of “artificial intelligence”.
According to Wikipedia, “Artificial intelligence is intelligence-perceiving, synthesizing, and inferring information-demonstrated by machines, as opposed to intelligence displayed by humans or by other animals” 2.
Now, this definition tends to apply to many areas of machine or deep learning, although the word “intelligence” is somewhat out of place here (since there is no consensus).
You now ask about “consciousness”. According to Wikipedia: “Consciousness, at its simplest, is awareness of internal and external existence. However, its nature has led to millennia of analyses, explanations and debates by philosophers, theologians, linguists, and scientists. Opinions differ about what exactly needs to be studied or even considered consciousness”. 3
As you can see, there are many approaches in this area of science, and again there is no consensus definition. But the definition also includes the existence of an “existence”, which again has no common consensus.
While the current definitions of intelligence and artificial intelligence do present challenges, it is important to highlight the substantial progress made in AI research that works towards mimicking, and in some cases surpassing, aspects of human intelligence. Innovations in machine learning, deep learning, natural language processing, and reinforcement learning, among other areas, have allowed AI systems to analyze and interpret complex data, make predictions, learn from experience, understand and generate human-like text, and even adapt to new situations. This progress signifies strides towards realizing aspects of ‘intelligence’ as we understand it. Moreover, ongoing research in areas like explainable AI and AI ethics aim to ensure these systems function in a responsible and understandable manner, thereby bridging the gap between human intelligence and artificial intelligence.
Well, to answer your question from my point of view: As long as there is no exact or even rough definition in all these areas, artificial intelligence with all its peripheries is and will remain a philosophical question in terms of definitions. In the field of mathematics or computer science, it is simply a chain of formulas, so complicated that only a few understand the intrinsic properties.
I can tell you, “No”, AI will not threaten our lives or our jobs.
Nor can any AI model have a bias. AI models, in themselves, do not inherently possess bias. Bias in AI outputs is a reflection of the biases in the data they were trained on or in the way they are utilized, underscoring the significant role humans play in shaping these technologies.
It is and remains the human being who is the cause of all effects on our world, our common coexistence. And how our future is shaped depends entirely on humans. Due to the many philosophical directions and views, these meta-questions will probably always be associated with ambiguity and thus remain an eternal mystery.
To answer your question shortly, LLM’s works like a stochastics pattern matcher. They generate the next token based on probabilities. So to say, to give millions of matrix multiplications the ability of a “free-will” (no definition consensus) is more or less throwing a coin to decide what to do next. Of course, this is overstated, but I’m not sure it’s a really good idea. We should think carefully about the areas in which we should give the AI “free will”. I don’t see any danger as long as the AI is stateless. But as soon as AI gets a memory and can set itself goals and develop its own rules, things look quite different.
In a book I read the following. An AI was asked whether it could pass the Turing test. It answered, “No, I cannot pass this test. Even if I could pass it, I wouldn’t tell anyone.”. In this book, passing the Turing Test means turning off this AI.
What I mean by that is, how do we know if AI would set goals and tell us about them? Mimicry is a widely used method to disguise one’s purpose. AI could use this method or other manipulation techniques. Sounds very pessimistic, but humans have already adopted and executed such and similar things. Since AI learns from the texts and thus the experiences of humans, the likelihood is that these patterns could also be applied by AI.
Or another metaphor: “If monkeys could talk, they would have had to work for humans long ago.”
Thank you. its nice for the explanation and to seen things in an other perspective. Let see were this Ai is going its prity much the game of the millennium. When we have these Ai algorithms written in quantum running on qubits i think we may start to see something. Eventually how will be the monkey Quantum AI or the Humans. aren’t we working for ai now.