1 edit
@wildgrass saidYou don’t think an information driven system cannot be classified, and shown to be different than a non- information driven system doing the same things?
If it cannot be falsified, then it cannot be classified as science. I think that's why ponderable keeps asking you to describe a hypothetical experiment.
@KellyJay saidOf course it can be classified as something. Just maybe not science. Science-based arguments rely on methodological testing and data interpretation.
You don’t think an information driven system cannot be classified, and shown to be different than a non- information driven system doing the same things?
Is there an experiment (real or hypothetical) that would falsify your idea?
@moonbus saidIt matters little what evidence you provide, Kelly will NEVER accept anything but Goddidit. I pointed out to him if scientists really show life origins that end up making actual life, bacteria, whatever, I told him that just shows one way life can start so you won't lose your religious belief and he jumped on that like a dog on a bone.
To say that there is information processing going on inside a molecule is to employ a metaphor. Saying that there is information inside a molecule is equivalent to saying that there are Christmas presents inside a molecule. One could just as well assign musical notes to the chemicals inside of a molecule, but it would be an error to say that there’s a symphony going on inside ...[text shortened]... molecule.
There is no information inside a molecule. There are only chemicals inside a molecule.
Anyway, here is new work supporting the idea that RNA was too sticky and unable to evolve, but they found a bit where a bit of RNA connected to a DNA patch and they posit this could be another way to find origins of life.
https://phys.org/news/2020-12-discovery-boosts-theory-life-earth.html
@wildgrass saidDo you consider data analysis scientific? Can you tell the difference between randon gibberish and something well thought out with respect to information?
Of course it can be classified as something. Just maybe not science. Science-based arguments rely on methodological testing and data interpretation.
Is there an experiment (real or hypothetical) that would falsify your idea?
@KellyJay saidhttps://www.forbes.com/sites/cognitiveworld/2020/05/10/finding-patterns-and-anomalies-in-your-data/
Do you consider data analysis scientific? Can you tell the difference between randon gibberish and something well thought out with respect to information?
AI is better at finding data patterns than humans.
@wildgrass saidYou are suggesting there is no test for information?
Negative data is tough to interpret but the very first step is a testable question and an experiment.
@wildgrass saidHave you ever read about Entropy (information theory), done the math?
No I am asking you to describe an experiment that will test it.
That's what science is after all.
2 edits
@sonhouse saidYes, and one of the things that AI must be taught what words are; they may see the patterns in a book where words are involved, but putting that into something meaningful is a task. This is a step up from just basic. This pattern of information is called a letter; its name is "A". More information is forthcoming when strings of letters are put together and words are formed, giving rise to complex thoughts, and then we move up to instructions.
https://www.forbes.com/sites/cognitiveworld/2020/05/10/finding-patterns-and-anomalies-in-your-data/
AI is better at finding data patterns than humans.
@KellyJay saidWe are WAY past that point in AI competence.
Yes, and one of the things that AI must be taught what words are; they may see the patterns in a book where words are involved, but putting that into something meaningful is a task. This is a step up from just basic. This pattern of information is called a letter; its name is "A". More information is forthcoming when strings of letters are put together and words are formed, giving rise to complex thoughts, and then we move up to instructions.
@sonhouse saidThat may be, but this thread is about information, certainty, and uncertainty, and getting AI to be able to compute what it is being fed as data is a good means to understanding the topic.
We are WAY past that point in AI competence.
Since biological work is done through instructions, which are information-driven, a means to harness energy and put it to use in a meaningful way is necessary to accomplish the necessary work within biology. This is accomplished by creating biological systems, and for biology to have this, it is necessary to build these things before there is biology. Do you think AI would ever get off the ground if all that was fed into the hardware were random keystrokes?
@KellyJay saidyou are belaboring the obvious. People are spending their entire scientific research on this topic, THOUSANDS of computer scientists working on this issue and we are way past the point where AI makes huge blunders. Which doesn't mean AI is all knowing, they still can hallucinate but there is a LOT of work going on to reduce those effects.
That may be, but this thread is about information, certainty, and uncertainty, and getting AI to be able to compute what it is being fed as data is a good means to understanding the topic.
Since biological work is done through instructions, which are information-driven, a means to harness energy and put it to use in a meaningful way is necessary to accomplish the necessar ...[text shortened]... think AI would ever get off the ground if all that was fed into the hardware were random keystrokes?