Which came first: sentience or programming?

Question

There is a high interest in creating an AI, despite the associated fears that such a creation could cause.

As I can figure man's approach is two-fold and its goal is being attempted by computer programming or by electronic physical neural nets.

As I can see it, intelligence is related to the sentience of humans and other animals. Which came first in the physical world we occupy: sentience or programming, genetic or an abstraction bound by physical rules?

Or as a direct result of the structure of the brains throughout the living world, which we may be possible to replicate. If sentience was first doesn't it mean that we will have to alter the definition of programming in order to achieve this goal by that method?

Answer 1

I do think physicists are most qualified to answer the part of this question regarding "which came first" regarding the universe or the rules of the universe. My understanding is that the physical reality we experience is based on a set of constants which have to fall into a certain range to produce a structure in which life-as-we-know-it can arise, and leads eventually to sentience and consciousness. I believe Stephen Hawking had something to say about this involving probability. Nevertheless there is a chicken & egg aspect to this part of the question that renders it probably not answerable (even though there were eggs long before chickens;)

Another problem is that terms such as sentience and intelligence can have a range of meanings depending on the framework. Because this is a purely philosophical question, there won't be a definite answer, merely viewpoints.

I tend to take a materialist approach and see the universe as combinatorial. In this model, intelligence and sentience arise from complexity of the system. (This is similar to the idea that Conway's Game of Life, on a gameboard of sufficient size, would produce intelligence. Sentience, defined as the capability to "perceive and experience subjectively" seems already to have arisen in human applications in Game of Life in that discrete structures can transmit and receive information.

It's unclear as to whether purely deterministic models can produce free will, a topic hotly debated for a very long time indeed, but deterministic models and reasoning can produce intelligence, albeit in a still limited form. (i.e. even a deterministic Tic-tac-toe algorithm has "intelligence", limited as it may be.) In AlphaGo, a stochastic model is employed in the form of Monte Carlo Tree Search, with very good results. Thus:

• It seems to me that the approach which has recently yielded what is sometimes categorized as "strong, narrow AI" is a product of evolving approaches to programming AI.

Quantum computing may be the next big evolution in approaching the problem as the capability of quantum computers advances.

Answer 2

The mind comes first. The brain is just a representation of the mind by itself [1]. For the mathematically-inclined reader of this answer, an analogy (mine, not taken from the cited book) could be the representation of arithmetic by itself through Gödel encoding.

[1] Herbert F. J. Müller, Brain in Mind: Ontology Becomes Pragmatic Design in the Unstructured, 2010

Answer 3

Sentience is a horribly imprecise word, even qualitatively. It is like, in physics to say something is imposing. Imposing why. Is it heavy, large, moving fast, jagged edged?

Many of the dictionaries say that sentience is the capacity to feel, perceive, or experience subjectively. That illustrates my point. Which one? Feel, perceive, or experience, and what to those even mean quantitatively. Furthermore, how can one feel, perceive, or experience objectively? Even the dictionary doesn't know what to say about the word.

Perhaps it is best to let the word fade away from disuse.