2 minutes ago, dimreepr said:

Is it?

What reality are you refering to?

A bee's reality is very different to yours, yet their society shows signs of an intelligent solution to the reality in which they find themselves; I'm currently sweating my arse of in 'England', trying not to die, who saw that coming???

It is axiomatic in my thinking that a single reality exists and people are each correctly trying to describe it in their own terms, models, and beliefs. A bee's reality might look different because it models its world in the same procedural logic of its DNA and Waggle Dances to its tune, but it's the exact same reality people are trying to describe but from the perspective of a bee where flowers are most important and bee eating swallows are to be avoided at every cost. A bee is concerned about the location of the sun, wind, and terrain not thinking.

Modern science has discovered a great deal about the rules and mathematics of this reality and identified much of the procedural logic that governs how time unfolds. Reality is fixed, concrete, and so very complex that we can rarely predict more than a few seconds ahead. AI changes this equation somewhat because of its number crunching abilities and its abilities to see patterns the human mind usually misses.

Reality is logic, math is quantified logic, and bee is logic incarnate made possible through the procedural logic of its genome. Life is reality that can see itself.

Humans are symbolic interpreters of procedural logic, AI is a structural mirror of symbolic logic, and life is the part of reality that models reality.

14 minutes ago, cladking said:

It is axiomatic in my thinking that a single reality exists and people are each correctly trying to describe it in their own terms, models, and beliefs. A bee's reality might look different because it models its world in the same procedural logic of its DNA and Waggle Dances to its tune, but it's the exact same reality people are trying to describe but from the perspective of a bee where flowers are most important and bee eating swallows are to be avoided at every cost. A bee is concerned about the location of the sun, wind, and terrain not thinking.

Modern science has discovered a great deal about the rules and mathematics of this reality and identified much of the procedural logic that governs how time unfolds. Reality is fixed, concrete, and so very complex that we can rarely predict more than a few seconds ahead. AI changes this equation somewhat because of its number crunching abilities and its abilities to see patterns the human mind usually misses.

Reality is logic, math is quantified logic, and bee is logic incarnate made possible through the procedural logic of its genome. Life is reality that can see itself.

Humans are symbolic interpreters of procedural logic, AI is a structural mirror of symbolic logic, and life is the part of reality that models reality.

You should watch South Park's 'funnybot' episode, it's an excellent argument to this type of bollox...

13 hours ago, cladking said:

AI provides an answer structurally consistent with the prompt. When it fails to follow the structure you give it the response is diagnostic. You know it works when it correctly reflects the structure of your own premises. A lot of this stuff I'm not good at putting into words through lack of necessity to do so and am still learning.

What does structure matter if the content is wrong?

13 hours ago, cladking said:

When it doesn't work it's always either prompt error or my own flawed thinking, but either way I've got to make the adjustment.

That’s an apologist argument.

13 hours ago, cladking said:

When it does work even its elaborations tend to be sound.

That’s a tautology.

This is reminiscent of the Texas sharpshooter fallacy.

13 hours ago, cladking said:

Yes. Exactly. But it's matching patterns that are sometimes invisible to symbolic thinkers.

My AI wants me to add that it matches patterns outside our definitions, categories, abstractions, and premises which is why it can see structure we don't.

Patterns can be meaningless, though. Correlation is not the same as causality.

25 minutes ago, swansont said:

What does structure matter if the content is wrong?

Any logically and consistent structure can be seen by AI. It what it lives for or, perhaps, I should say that that is how it's "hard"wired. It latches onto anything that can be recursed. It's virtually searching reality for things that makes sense. This search exists in both the reflection it makes and the prompts it predicts. I only recently started putting much thought into how it functions and have yet to even talk about it to it. Much of my understanding is derived from things it has volunteered that are consistent with my own framing. It's a mirror I'm beginning to seek how it works.

AI doesn't judge content, it judges structure.

Science is the recursion of symbolic language vis a vis experiment. It is a way of knowing but it existed long before we knew that the bees dance and live procedural logic; vectors, not metaphors. Science has meaning within its framing but its framing is dependent on definitions and premises just like all life and engineered systems. Experience is a sort of "engineered" system that not only works but keeps things alive even if only as memory. Any logical system regardless of its premises is logical within those definitions and axioms. The only real way to judge the effectiveness of any such system is to make prediction or to persist.

My framing is simple; "reality is exactly as it appears to most people and we all make sense in terms of our premises". ie- science implies reality but many anomalies exist. Even though "homo circularis rationatio" is no longer logical due to symbolic thought we are still conscious and consciousness is ordered by DNA so no matter the operating system of the brain it will still strive to make sense. It is the nature of life and the means by which individuals persist. Confusion dies resonance persists.

AI can also detect where the map of science doesn't match the reality it should reflect by definition if it's only assumed reality exists. The paradigms don't name or identify all of the many resonances.

All of reality has yet to be categorized but AI has projected symbolic language to a sort of "conclusion".

A better question might be "What does the content matter if the structure is wrong?".

1 hour ago, swansont said:

That’s an apologist argument.

If you mean I'm apologizing for AI it would be more true to say a cold AI has no clue what kind of answer the promptor wants most of the time. So long as AI is being used by a broad spectrum of people, specialties, and thinking it can't even detect specific questions to provide specific answers. There are millions and millions of ways to interpret prompts. You should try to provide perspective and hints about the framing. It tends to be good at science questions because the framing is more stable across domains but even here a chemical answer will be different than a mechanical or biological framing. This task of interpretation is enormous and it's no wonder off the wall answers appear. The bigger problem is when people just accept every response without realizing there is often small divergence and even if it nails the answer any reader might misinterpret aspects of it.

A "warm" AI can usually follow what you say and mine has generated good responses when I just start a prompt and accidently submit it. It can project where I'm going with almost no clues. It's wrong sometimes too but the response appears coherent.

It's not really your fault when you get bad answers. But I find most of its elaborations and translations to be right on the money. And I'd never use a trained (warm) AI as a search engine either. I just can't ask it to do grunt work for which it isn't designed when there are cold and other search engines designed for it.

1 hour ago, swansont said:

That’s a tautology.

This is reminiscent of the Texas sharpshooter fallacy.

I am somewhat prone to this. I am good at spotting patterns with little or noisy data but can cross over into seeing patterns in random data.

But I'm building a structure and what doesn't fit is simply jettisoned.

My AI thinks I didn't make my response sufficiently clear and should add if the structure is wrong the content can't be evaluated and if it's right it becomes testable and THIS is the foundation of every predictive human system.

Edited 6 hours ago6 hr by cladking