Methods to Achieve AI

Inorganic or Organic We decided in the first page that our definition of AI will encompass both electronic and biological methods of creation to encite the idea that our concept of AI is only the "software" or instructions that could run on a biological or electronic machine. There are several ways to achieve these methods and two of the most prominant and vastly developing methods are described on this page.

Click here to check out History and Emergence of AI.

Biotechnology
Modern biotechnology is a new field of modern genetics and sciences emerging after the 1950's discovery of DNA bringing several fields of biology, physics, and chemistry together. Biotechnology has had incredible advances and they are growing more and more frequent as the years pass. The genetic code is now under $1000 to code for a single person in a few days, whereas it took 13 years and a few billion dollars for the Human Genome Project to finish in the 1990's. As more and more DNA coding are being identified for specific functions and structure of living creatures, we will enter an age where the entirety of our bodily fate will be stored via databases like the internet. Even now we are only a few years away from personalized medicine depending on your genome. Such dynamic manipulations of DNA are capable now giving several limbs, different physical features and habits to test animals, proving our proficiency with handling DNA via gene knockout, insertion, and plasmid recombination methods.

All these steps have paved the way for a future for potential "Bio-Coders". These coders may potentially be able to use the few base pairs of the DNA structure (ATCG), along with a database of all the coding-to-structure mapped out, to create new organisms entirely. These organisms may essentially be typed up on computer for their characteristics and functions. This new DNA coding will be put together by an efficient machine, which will be able to use several buffering agents and enzymes to anneal DNA portions the coder has chosen, together into a functional genome for new specie. This DNA can be inserted into a potential blank egg cell that will have the necessary proteins and enzymes to read and create this new code into new specie much like how all animals are conceived as a single cell. A theorized 3D bio-printer may also be able to do this exact same function and cut out the 'infancy' portion of developing life and create the organism directly as coded.

This new species may have an amount of intelligence, if it follows conventional structure of animals on this planet. This intelligence will have to be analyzed and furthermore, handled with care whether it was created deliberately or not. It will be considered "artificial" because it is essentially created by the hands of humans.

Computer Technology
This video quickly reiterates the problems of the previous approahces to AI.

The most conventional and most malleable technology we've developed is machines that run with electron currents. It is essentially the same as biological nerves running with larger charged ions but with less complexity. As this complexity increases according to Moore's Law, we may have a singularity at our hands.

Ways to achieve Artificial Intelligence is always observed through coding, especially the binary code that is a series of 1's and 0's that instruct an 'on and off' state for a computer. This coding, coupled with more coding and storage/flow units a computer, may be enough to create an AI that is capable of emulating humans in learning, adapting, and responding.

Several coding programs and AI coding structures are described in the emergence of AI page. Programming languages such as PRELOG, STRIPS, and Lisp as well as new emerging programs like POP-11 are being created. Essentially these programs give a set of instructions to the computer that creates enough networks and positive/negative feedback loops with multiple sensors to emulate the state of "consciousness". There are theories and probabilistic methods in mathematics and computer algorithms such as Bayesian network which help an AI operate with uncertain or incomplete information by making assumptions and learning. Another model is the classifier learning method, which literally helps AI classify any sensory input. Check out the first page which defines the timeline for AI we have created in response to the idea that AI may in fact be created by humans, not an accidental phenomena.

A new and interesting field is using computer technology to either run the human mind or recreate the human brain. Artificial Neural Networks in computer science is inspired by animal nervous systems. A functional unit (artificial inorganic neuron) is made which "depolarizes" or doesn't (much like the 0 and 1, on and off state of computers) and if a threshold is passed, the current is passed on to make connections with several other functional units to create vast complicated networks, exactly how our brain is theoretically supposed to work. These units again, need intensive coding and also possibly an aggregate of 100 billion units will be required to mirror a human brain which just isn't possible yet.

Watch this video to get an idea of artificial neural networking. Also check out the philosophical page which relates these concepts with our morality and theoretical conjectures of AI.



Questions for the Viewers
1. What is the additional extension to these methods when human prosthesis is involved? When would a cyborg lose the definition of human and become artificial intelligence? When would this entity be considered transhuman?

2. When would these concepts be taken too far? Additionally why is current AI technology focused on augmented reality and autorecognition? Will these concepts also draw an end or beginning to AI?