Taken from: I am - A blueprint for sentience by Krys Norman
There are many problems with this book. Some are known to the author and some are not. It contains a collection of opinions and beliefs that may well be contested by the many disparate groups that have a vested interest in the topics discussed. The ramifications of these are wide reaching and also affect many genres of academic study as well as social groups. The author does not profess to be an expert in any of them and apologises now for the understandable consternation that will ensue from poorly phrased and antagonistic ideas. The book is written as if it is a scientific document but is not intended to be peer reviewed or use references to established works. The author understands that much is flawed and inaccurate. The writing style is littered with too many inadequately explored ideas and, above all, may well come across to the reader as arrogant and smug. This is regrettable and is more indicative of the author's literary and intellectual limitations than an attempt to ride rough shod over reader's views. During the many times that a reaction of indignation or incomprehension may well occur it is suggested that this first paragraph is re-read so that the reader may feel at least a modicum of the author's humility about the whole project.
This is not aimed at being an academic document; nor a piece of popular fiction. It is a pragmatic attempt at providing an insight into sentience and also to show how, perhaps, sentience can be created. Another caveat is that the following views and descriptions are taken from a desire to create a model for sentience and may well be made up in order to justify and reinforce the author's belief that this model is the best one around. i.e. this could all be a self fulfilling prophecy of nonsense. This may be the case but is not the point. Anyone who does not want to follow the possibility that this book could lead to the creation of sentience will always be able to find countless flaws in any document, however meticulously researched, written and reviewed. Not only that, they may have a righteousness of their own intellectual logic even if their pursuit of the book's flaws differs significantly from that of others. A common denominator may appear in the variations of these disparate but determined critiques; that is a desire to show that sentience cannot be determined and created and, most notably, should not be attempted. As is to be discussed, part of the what may well create a sentient being would generate the same reaction in one. A created sentient being would come out on the side of the critics not the author's.
Emotions are at the heart of all thought. Reasonable, rational thought can only go so far and then it descends into base, abstract philosophy. The human mind has to deal with the staggering complexity of existing. It has to discern what visions and noises and tastes and smells and touches are; what they mean; how that correlates to all previous experiences. What could be done with all this information? What should be done? How to deal with confusion and ambiguities? All this is arrived at every instance using the brain's cognitive capability and its emotional controls.
This is incredible. The nuances and subtitles that have developed within humans are almost impossibly complex. However, the aim of this book is to provide a blueprint of how the essence of what has made us unique so far can also be modelled and replicated in machines. It is to give an insight of how computers can have responses in the way humans do. To show that they could have “irrational” behaviour; that they can exist as sentient beings. Many organisations are utilising the very limits of intelligence and technology to this end. The problem with the endeavours into creating sentience so far is that the designs have either used sophisticated mimicry of outcomes or intuitive emotional algorithms. Neither of these has worked particularly well and there is, currently, no way forward. The populist understanding of development platforms seem to lie more in the likes of superstition, God, magic, conspiracy and hokum. Why not also bring in alien intelligence into the pop art sci-fi ideas that abound? There's always the plot of computers from the future to give ninety minutes of viewing pleasure to talk about over drink later.
No, these don't work. Not only that but people want the saccharine solution. One that feels nice to interact with. Preferably, most investors say, it needs to be commercially viable. For this to happen, artificially intelligent devices need to be faster, more efficient and less fallible than humans. At their most base level, they need to adhere to safety orientated rules to ensure that humans and/or property do not get destroyed at “inappropriate times”. Whether they assist peaceful construction or war based destruction they will have to do our bidding. These are robots. Electronically based devices that aid basic human needs through mechanical and electronic slavery. The writer, Isaac Asimov wrote some hallowed “robotics laws” and then developed many practical and ethical ramifications in entertaining fantasy.
They are:
First law: A robot may not injure a human being or, through inaction, allow a human being to come to harm.
Second law: A robot must obey orders given it by human beings except such orders would conflict with the first law.
Third law: A robot must protect its own existence as long as such protection does not conflict with the first or second law.
How very concise, comprehensive and comforting they are too. These laws are sixty years old and many developers still see them as fresh and something to aim towards. Where the discrepancy lies, however, is in the understanding of what artificial intelligence is. This leads to what natural intelligence is, what is a machine, what is a robot. In turn, the questions move to what is it to be human, to be sentient.
This is not going to rehash how to make robots better, faster or easier to command. It’s not going to provide obvious revenue generation schemes through complex pattern recognition and response algorithms. That component of computing is being developed by many highly focussed and capable organisations already. Instead, this will provide a blueprint for something that is considerably less easy to define. It will start backwards from conventional intuitive thought. A snapshot of current human thinking and behaviour will be discussed in section one. In the next section, sentience will be defined and modelled. So too will stimulation of senses, memories and rationality. Only then will interaction come into play. The algorithms for all these processes are to be outlined in this section such that they can be used to implement sentience within machine code.\
As for a slave robot obeying commands? How sentient is that? How human is that? What about the basic rights that humans hold so dear? Surely, a sentient being is morally entitled to fight for its rights. These debates will rage (maybe literally) for decades to come. An introduction to them is to be found in section three.
What is to be defined here will also have beliefs. These beliefs will be layered upon each other and connected by the effects of previous experiences. The higher level ones may be implausible, incomprehensible or even downright wrong, all mixed up with the more accurate lower level ones. But they will be accepted as true and beyond any rational assault by their owner, just like with humans. And so on to religion. That Bastion of the human differentiated cultural groups. How would it be if instead of God creating man and having a plan, man could observe God being created as an extension of sentient thought?
So, this is not going to produce a robot with artificial intelligence. This is going to provide the basis of something that not only thinks, not only believes that it thinks but knows it exists because it believes that it thinks. It will also want to protect that state. It may be close to, or remote from, whatever is considered human in this day and age but it will be a synthetic “being”. This is the blueprint for sentience.