Programming a Human: A different perspective at Artificial Intelligence
“It’s just a computer”, “Computers cannot think”, “They are far from being intelligent” are some of the statements we get to hear in our day to day lives when discussing the matter which is ‘Can computers think?’ As proposed by Alan Turing in 1950, a test was conducted which involved 2 humans and a computer in three separate rooms; a judge and two test subjects in which the human test subject and the machine engaged in an anonymous conversation with the judge. The goal of the test was for the machine to trick the judge in to thinking that the judge was talking to the human test subject instead of the machine. The test was to determine whether machines could imitate human behavior to trick a judge in to considering that the judge could possibly be having a conversation with a machine and still be convinced that a human is carrying out this conversation. This was later known as the ‘Turing Test’ which is still an accepted way of testing a machine’s intelligence.
Years have passed and machines have become more ‘advanced’. I choose not to use the word ‘intelligent’ since intelligence is based on several factors which need to be satisfied in order to qualify as ‘intelligent’. Intelligence in definition is the cognitive capabilities of a living organism to understand basic concepts of ‘life’ and one’s capacity to solve problems or to be self-aware. Humans and other similar living organisms are born with the ability to ‘think’; the ability to reason and understand. This ability is purely based on ‘concepts’ which are derived through experience. A child born to this world not knows that he/she exists, that life empowers his/her body or that one’s life is controlled by their own actions. A child however grows in to understanding these concepts and starts to make decisions through ‘instinct’. This ability is known to man as ‘consciousness’ and has become one of the most controversial topics of discussion between scientists, given that it has many definitions depending on how one defines it.
The gist of this is, humans have evolved with time and it has become a universally accepted truth that machines are far from being human and that the ‘machinery core’ or in other words, the algorithms that empower a machine are far from being intelligent when compared to humans. I choose to disagree. In my opinion, intelligence is even the slightest presence of an ‘ability’ to carry out functions demanded by decisions that were self-proposed. Machines qualify to this theory and let me prove this right.
Being a programmer myself, one of the most entertaining activities I found myself doing was to develop computer programs that were able to carry out tasks. ‘Making things happen’ as advised by the engineering plan one carried out is one of the simplest joys in an engineer’s life. I started programming when I was an Ordinary Level student at Trinity College Kandy and the programs I wrote were simple as applications that determine ‘How awesome you are’ depending on the number of times you logged on to Wikipedia (for research of course) or simple calculators that carried out my math homework. All of these simple applications were ‘programmed’ to carry out tasks and therefore critics might disagree with me when I call them ‘intelligent’. Here’s my perspective. In a way, humans are programmed too; children are taught the good and the bad by their parents, children are taught by their teachers, children are guided throughout their childhood and to rephrase this: Children were programmed.
A simple calculator if programmed right, should give out accurate results of different mathematical operations performed on numbers within milliseconds and we have been using digital calculators since the 1960’s. One cannot interfere the algorithms and one cannot prove a calculator wrong unless there’s a technical fault within. A minute form of ‘intelligence’ is displayed by a calculator when performing these tasks and when you come to think of it, we have been using an artificial mind to solve our mathematical problems since the 1960’s. This form of intelligence present in machines was coined as ‘The ghost in the machine’ in ‘The Concept of Mind’ by Gilbert Ryle in 1949 which proposed the statement that machines have an intangible form of generating a certain consciousness that was similar to that of a human. This was derived from the concept of ‘dualism’ put forward by the French physicist René Descartes who suggested that the mind and the body is two separate entities that were independent from each other. This was a theory that was subjected to many controversial discussions and experiments and still unproven as to whether it is a valid or invalid theory.
In Buddhism and Hinduism, a similar theory was put forward as “Nāmarūpa”; Nama describing the spiritual properties of a being and Roopa describing its physical presence. Nāmarūpa is also known as the ‘Five Skandhas’ which in Buddhist Phenomenology refers to the five functions that form a sentient being. According to the Five Skandhas, ‘Roopa’ was considered to be the physical body, ‘vedanā’ being sensations, ‘saññā’ was considered to be the ‘perception’ or cognition, ‘saṅkhāra’ was considered to be the mental formations and mental prejudices and ‘Vijñāna’ was known as ‘the human consciousness or the ‘life force’ that empowered life.
Moving away from the consciousness, let’s direct our attention toward advancements in computer programming and technology. As of now, we have the right tools and research to develop an artificially intelligent ‘being’ or a brain that empowers a robot. Based on this I conducted a research known as “Eddie”, an intelligent research assistant that reads and analyses online material and answers questions asked by its users. The program is in its initial development phase but is advancing rapidly over time (For more information: Youtube.com/EarthlingBots). Some of these tools are known to programmers as “Machine Learning algorithms” that enable a machine to learn through experience and decide how to solve a problem. Another one of the research I conducted during my free hours, was developing an “Artificially intelligent Rat” that was programmed to mimic a real-life rat. The rat was programmed with the most vital survival instincts of a real-life living organism; seek shelter, scavenge for food and avoid danger. This project was implemented using a sub-field in Artificial Intelligence known as “Genetic Algorithms” which in definition means, ‘a process that mimics the natural selection of this world based on Darwin’s Theory of Evolution’. In simple terms, this algorithm creates a population of a defined species and develops their genetic structure by emulating ‘evolution’ and creating generations of the defined species. With each new generation, the species could have evolved with nature-inspired techniques such as inheritance, mutation, selection and crossover. Inheritance being the process of passing traits from the parents to their offspring, mutation is when an offspring naturally ‘differs’ from one generation to another and their DNA being slightly altered positively or negatively, selection being the process of naturally choosing the most fittest organisms of one generation to encourage breeding and crossover being the process of breeding using the selected ‘fit’ chromosomes. The rat initially has no knowledge of survival basics such as movement, food scavenging or self-protection but develops these traits gradually with each new generation all thanks to the techniques mentioned before. Genetic Algorithms could allow machines to technically evolve by mimicking the natural evolution and natural selection.
What I believe is that ‘Genetic Algorithms’ could be implemented in different computer programs such that they evolve over time and develop their problem solving skills. The future seems tough and mysterious with the mindset humans possess yet it’s possible to solve important problems that require immediate attention such as Cancer or HIV/AIDS. However this could take decades but by using machines that possess far greater and powerful thinking power than humans, the remedy to these problems are only a few years away. Humans are powerful as it is but we have achieved more due to technological advancements and history is a constant reminder about how easy our lives have got since the invention of the telephone, refrigerators, vehicles and the list is endless. Machines can be trained to determine answers depending on a situation and the remedies it could suggest is possibly the greatest invention we can hope for. Artificial Intelligence is the key to transcendence and the pathway to a safer, comfortable and reliable future where diseases or environmental hazards can’t affect any valuable human lives.
However there’s one more outcome of AI that I should direct your attention to and if I do not address it, my arguments or research would be void. World renowned physicists such as Stephen Hawking and entrepreneurs such as Elon Musk have publicly warned the world about an oncoming ‘technological apocalypse’ as commonly portrayed by blockbuster movies such as “Terminator” or “I-Robot”. This is a prediction known to the computer literate as “Technological Singularity” which predicts that technology will surpass human intelligence and act on its free will, demanding human slavery thus creating a newer generation of machines that ‘control the world’. I neither agree nor disagree since the development of systems that learn and develop on its own could finally lead to a self-conscious generation of computers that gradually become a species. However, this event is void if machines do not possess an algorithm to mimic ‘human consciousness’; the feeling of being alive, self-consciousness or the knowledge that life exists and the world is a playground for living and non-living beings. These concepts are currently mysterious even for humans yet alone for a machine to theoretically derive through experience so therefore I believe that it’s exaggerated in movies and of course everything has its downfall if misused. The good in AI outweighs the negative ‘assumptions’ and that’s what these statements are right now: Assumptions.
Artificial Intelligence could lead humans to a far more advanced generation where natural disasters would still happen but will take no lives due to advanced prediction systems, diseases that take the lives of innocent people would still spread but the cures are at hand and daily tasks that take up time and energy could be simplified. This is the transcendence we should hope for and the only way we can find answers to mother earth’s most mysterious questions. Last but not least, if anyone is interested to further expand their knowledge in Artificial Intelligence and synthetic life, feel free to follow our posts on facebook at fb.com/ArtificialEarthling.
Dinuka Jayasuriya
