The Outcomes of Logical Analysis Are Unpredictable by Lorraine Ford

Dear Lorraine,

Thank you very much for your clarifications and answers. I invite you to look at my ideas and my understanding of Logic ("Logic of all logics"), on the basis of which Nature talks with us and my approach to the construction of the basic structure of Nature and cognition. I look forward to your critical comments and questions as this is very important to me.

You correctly wrote earlier in the FQXi Blog regarding the modern "the world being undecidable, uncomputable, and unpredictable". The current situation in the world introduces maximum Uncertainty into the existence of Humanity. Obviously, this also depends on the crisis of understanding in the philosophical basis of Fundamental Science. We all live in the hope of overcoming this crisis by joint and individual efforts to create a more sustainable joint future on planet Earth in an era of ever-increasing existential threats and risks.

Respectfully,

Vladimir

Dear Lorraine,

I look forward to your questions and critical comments on the ideas of my essay.

With best wishes, first of all health, in this difficult time for all Earthlings,

Vladimir

Hi Lorraine,

I hope your wombats, daffy ducks et al survived the fires ok. I spent much time with F&F around Aus recently. Great country in so many ways I was devastated watching the destruction.

I had to read your essay twice to really get what you were saying, lucky it was so short! But beautifully written as always. Fascinating and original way of looking at relationships, which of course I agree with. There's no direct link between the data analysis and the laws.

I found it particularly interesting as I addressed the very foundations of logical analysis and the Laws of Physics in my own essay, finding them flawed. I hope you find time to read and comment.

I don't find brevity an issue and think you talk more pertinent sense in 2 pages than many do in 9. Well done.

Very best wishes,

Peter

PS Whereabouts in Aus are you?

Dear Lorraine,

In a comment you say that "what underlies the world" has always been the same stuff, back then and right now. I agree completely!

As for 'logic', I found Schultz's discussion of algorithmic patterns vs non-algorithmic patterns very interesting. The limitations this contest focuses on are algorithmic. He suggests that the implied limitations on 'knowability' do not apply to non-algorithmic patterns.

Logical analysis occurs in the mind, and you say that "logical analysis is an aspect of the world that can't be explained by physics..."

I agree. My understanding of consciousness is as the self-interacting field, with dynamic changes induced by flows in neurons, which in turn induce changes in neurons. In such a situation every point in the local field affects every other point, and the whole process is non-algorithmic. [Of course non-linear interactions include linear approximations where appropriate, enabling 'logic', but this is a tiny part of the mind.]

Not arguing with you, just presenting my picture which I think supports your views.

I invite you to read my essay and comment.

Edwin Eugene Klingman

Dear Lorraine,

I understand that you may represent lights and sounds as numbers under certain system of representation. But when you say, "living things need to be able to respond differently to these different sets of numbers", these numbers become universal, not particular to the specific scheme of representation. How so?

You state, "This logical analysis and its associated outcome can only be represented (often only after the fact) as IF...THEN... statements." This statement does not leave a possibility for an associative processing. An associative processing results from direct activation of agents (neurons in the brain) as per the causal powers of incoming signals in hierarchy. And such processing may lead to actions, but without logically testing IF...THEN as we understand from computer codes. In fact, associative activation may also represent the relation of causal dependence of action on what the incoming signal represented, which will serve as semantics of functional or logical relation between the two and may further connect in processing where such a semantics of causal dependence statistically relates. One may ask, why do such connections form in the first place. The selection based evolution of neuronal organization can be the only basis to give rise to function of neurons that self-organize based on the statistical relevance as you correctly pointed out -- physical laws are blind to any purposeful action. As you said it yourself -- LOGICAL ANALYSIS AND ITS OUTCOMES CAN'T BE DERIVED FROM THE LAWS OF NATURE -- so selection based evolution is the only path left.

You say, "And clearly, life could not have evolved without the pre-existence of at least a small ability to logically analyse (what we would represent as) the numbers associated with the variables, and an ability to respond to this analysis." This statement goes against the understanding of selection based evolution of random modifications via mixing of chance variables. Selection and random modification are both natural physical process. Being a programmer, you may write a simple code -- think of any physical function of a set of elements with limited replenishment of resources that these elements consume. The function includes the abilities to join together in specific shapes with probabilities as atoms join to form molecules and molecules to much larger molecules and so on. Let the function also include differential need to consume resources and differential probability of replication of geometrical shapes or what ever you wish to base your probabilities on. Keep the probability differences as small as you may like. All of these functions that you coded are blind with no purpose, like laws of physics, yet, after passing of sufficient time, you will find only certain shapes are surviving, while others die out due to insufficiency of resources that different shapes needed to consume. If you observe this, then you will have to agree that any functional organization may arise, given enough time if it is better suited to survive under the blind laws of physics and enormous diversity of contexts. Even the neuronal network may emerge that represent the body conditions and their suitability limits, and neuronal function that self organize depending on the incoming signals from the environment and the the suitability of the body conditions (goals). And these neural organization may encode IF...THEN but at a huge cost of testing multitude of conditions at at each step, or simply reconnect the needful activations depending on the environmental conditions and body requirements based entirely on statistical occurrences of such relations. And such reconnections would gain stability with time requiring minimal modifications with changing context.

You say, "Computers/ AIs can't know that their high and low voltages are meant to represent zeroes and ones, and that these zeroes and ones are part of a binary digit system of representation." Indeed, this is absolutely correct. But all states of matter bear intrinsic causal correlation with the information of precursor states and what they in turn represent that make the states a reality. Physics has entirely missed out on this, that is why physics does not deal with the semantics of information, only with the bits. Though my essay is also brief, but you may find how meaningful information arise in nature.

Rajiv

Hi Lorraine,

I am always amazed that the beauty that I see around me is electromagnetic radiation generated by the motion of electrons. Our brains do all that amazing If....Then....Else.... processing and we comprehend a magnificent tiger or a songbird or...

I explore some ofthe issues that lead us to such abilities in my essay which I hope you will read and comment on , even though voting is probably finished.

I am going to do some voting now if the system lets me, as it is just past midnight in Melbourne.

Good Luck

Lockie Cresswell

Hi Lorraine,

In a previous essay comp I wrote:". With some luck some "rocky" planets formed in the Goldilocks zone, at distances and temperatures where liquid water could exist. On one or more of these planets, which harboured carbon, nitrogen, and liquid water, complex organic chemicals formed primitive amino acids in a process known as abiogenesis. In some particular environments, such as are found in alkaline deep sea vents, some arrangements of these organic molecules were able to replicate themselves by a sort of autocatalysis reaction. The important idea here is that the required complexity was forced by the environment, a form of top down causation. Different environments selected different organic molecule arrangements, and it turned out that some of these were able to replicate themselves. Thus was formed the first molecular memory, a very special arrow of time. Forced through endless changes in its environment, this memory enabled greater complexity to arise until, at some stage, what we regard as biological life was formed. Then the various processes of evolution continued to drive greater complexity, when finally at some point consciousness arose. Maybe it was just some lucky recursive wiring of neural networks, or maybe it was just a critical mass of them, but what consciousness allowed was the formation of intent. It probably started off in simple modifications to the environment coupled with further genetic evolution, but what ultimately occurred was the arrival of intelligence. Intelligence coupled with the ability to learn, to record and manipulate memories allowed us to gain insight into our environment, and thus began the natural sciences (physics, chemistry, biology, and geology) and mathematics. When all this knowledge was fine-tuned by scientific method over many centuries, humans gained the ability to build a radio telescope, which, in my biased opinion, is the definition of intelligence as a process."

Here I am setting the scene for how conscious 'programming occurs in sentient animals.

You state: "It is important and necessary to point out that computers/ AIs don't do logical analysis, and don't decide on outcomes. Does an earthworm or an amoeba do logical analysis? I suspect they (as sentient animals) do some primitive analysis according to their DNA program as they apply some form of If....Then... analysis making a muscle contract in response to a saline gradient or light or ...

In my current essay I noted: "In the computer science field of artificial intelligence, an intelligent agent is any device that perceives its environment and takes actions that maximise its chance of successfully achieving its goals. In order to do this an intelligent agent needs some form of map of the past. So therefore we need to include intelligent agents along with living organisms in the ability to sense the passage of time. It may well be that a class of skuld robot (a mechanical agent/entity spawned of artificial intelligence, and endowed with free will, self-awareness, metacognition, and problem solving [luckily for all us organics not yet invented, and hopefully never to be!]) will also have an innate ability to sense the passage of time and thus project their problem solving into the future. So we can see that living organisms and intelligent agents have the mechanisms to store selected maps of the "now" so that appropriate future actions can be taken, all occurring in the "now". Thus the passage of time, as we perceive it, does not exist as anything else other than an illusion that is necessary for living organisms to exist."

Regards

lockie

Hi Lorraine,

You stated:"The essential difference between living things and computers is that living things process information, but computers process symbolic representations of information."

I think the essential difference between skuld entities and computers is that computers don't think,saerehw skuld entities do. However I may be wrong on this as my computer typed 'wheras' backwards to inform me it was thinking. Truly it did this twice!

Regards

Lockie

I think that a lot of the to-ings and fro-ings about 'information' in this blog is because we do not define what sort of information we are talking about. There are many meanings depending on the context, a bit like the word 'space'. We dont generally define things too much in essays otherwise they can become less readable and may seem more like a journal paper. But by doing so we can risk readers missing our intent.

I'v just read your blog comments from the top and realise the topic is a very important one that needs discussion.

I have just flowsheeted a scenario we were discussing re physical information versus symbolic information as might be processed by a living being (skuld entity) or by an intelligent agent. Same physical input, same end outcome (after symbolic processing) So I asked myself what is the differece between the two? Well the IA will keep running until it is programmed to stop, and will do its job according to and within the accuracy of its design, wheras the skuld entity will stop when it decides the time is right (needs food; is tired; was meant to be relieved of duty hours ago;or we finished the job as required; or I may come back to this later; etc) ie. The skuld entity thinks about what it is doing wheras the IA doesn't.

A skuld IA of course doesn't yet exist (except for this computer which has a mind of its own)