What is AI?
I struggled for years trying to answer the question "Who am I?"
The profound sense of peace that I got from defining myself is already challenging to sustain; now, pundits are claiming that Artificial Intelligence (AI) may know me better than I know myself.
If I do not keep pace with the workings of these emerging technologies, I may be declared non compos mentis. I may just totally lose myself.
Computer science is not my discipline, so my technical knowledge on AI is limited. However, I have been exploring the topic, because AI always aims to leverage existing communication technologies. The interrelatedness of information science and communication science is axiomatic.
Defining AI is like defining media – the more it is delineated, the more it is debated. The semantics is comparable to love, time, mind, god, and soul; these concepts are so woven in everyday life that it cannot be separated from our being. It is challenging to come up with a universal definition for anything close to "I."
AI is becoming more intimate to us. Its inventions can now be viewed as extensions of our bodies. Like what Mark Weiser claimed on powerful technologies, AI has successfully entwined itself into the fabric of everyday life that it is becoming indistinguishable from it. Without our conscious knowing, it is stealthily becoming a part of ourselves. The logic is similar to having a phone: when it was first invented, it was just considered a luxury; in the current era, it is already a necessity. A phone is an extension of our hand; we cannot socialize and work without it. There are now more phones than humans, and AI is going towards the same destiny.
Most sci-fi movies portray AI as a robot or as a cyborg. Google Images also predominantly show it as an advanced being that belongs to the white racial classification. It cannot be black or brown, so the masses cannot be blamed for having a distorted limited idea of AI as just "a white robot" that performs human functions.
In fact, AI can morph into various forms according to its desired purpose. It does not have one definite shape. Nowadays, it merges itself into the body of existing technologies.
John McCarthy, one of the fathers of Artificial Intelligence, defined AI as the science and engineering of making intelligent machines, especially intelligent computer programs.
There is no question on understanding AI as "artificial," because it is easily understandable as an invention. However, the "intelligent" part can appear vague. Like media as "communication technologies," the definition of AI as "intelligent machines" is also broad. It is challenging to grasp whether a technology qualifies as AI if there is no universal definition of intelligence.
In Computing Machinery and Intelligence (1950), the mathematician Alan Turing explored the idea of thinking machines in computer programming. For him, an intelligent machine can be determined through The Imitation Game. If a person cannot distinguish whether it is communicating with a computer or with another person, then a machine can pass as truly intelligent. This test has been a motivation for AI developers: the basis of a strong AI is dependent on its capability to trick its users that it is a real human being.
Turing also believed that machines can only imitate the adult human mind if it is programmed to learn. It shall not exist to just follow the instructions set by its creators; it must be programmed to simulate the mind of a child that is subjected to the experience of education. This idea resonates with the current disciplines of Machine Learning and Deep Learning where an AI technology is dependent on patterns and examples to improve itself.
Only does what is instructed | Instructed to learn |
Has a predictable response: it strictly behaves according to what it is set to do | Predicts and decides what to do according to the rules provided |
Example 1: If it is asked who is more handsome between Piolo and Coco, it is already programmed with an answer: Coco is more handsome than Piolo. | Example 1: If it is asked who is more handsome between Piolo and Coco, it will be instructed to process many pictures of handsome and unattractive men to come up with a reliable answer. |
Example 2: If it is asked to identify who is Piolo and Coco, the specific facial features of both men will be readily provided for it to come up with an answer: Piolo is the one with the reading glasses; Coco has a good eyesight. | Example 2: If it is asked to identify who is Piolo and Coco, it will be instructed to process many pictures of both men to see who's who. |
In this sense, it is more practical to define AI as intelligent machines guided by algorithms to learn, relearn, and unlearn/self-correct through collecting, processing, and analyzing data.
If Julia is asked who is more handsome between Piolo and Coco, she will rely on the experiences that she had with the two men. She thinks both have good facial features, but Coco spends more time with her.
As Julia is fond of Coco's presence and sense of humor, her emotions directed her to conclude that Coco is more handsome than Piolo. |
If an AI image classification technology is asked who is more handsome between Piolo and Coco, it will rely on this guiding algorithm: it must look into the height, eyes, skin color, and body build of men to determine the answer.
The AI decides based on the pictures of men mostly voted as "handsome." As the data it has collected shows that tall muscular brown-eyed Caucasian men are deemed more attractive by women, it concludes otherwise: Piolo is more handsome than Coco. |
In the intelligent machine case above, the more pictures that the AI can collect of handsome men, the more reliable its answer will be. Though this example involves a petty situation, the logic is clear: AI improves with better algorithms and more data.
Strong AI is yet to be achieved: there is still no machine that can perform with multiple levels of intelligence like humans. The current AI technology is still weak, as each invention only specializes in a single level of intelligence; nonetheless, this is not an excuse for AI not to have drastic effects in the world now.
We live in a world of specializations. As an AI technology gains more expertise in a single level of intelligence, professions are at risk. Having a niche in the job market may not even work anymore. Self-driving cars endanger human drivers; AI-powered diagnostics are more accurate than doctors; AI private tutors are more patient than teachers; and AI surveillance cameras replace security officers. In fact, artists are not also excused: AI can create films, songs, and other works of art. AI bots in social media also challenge the journalists in defending truths. In farming and food industries, AI may be more ingenious and creative in developing recipes according to the the current crop supply.
We are used to leading the machines that support us; now, we are programming our machines to lead us with our minimal supervision.
The positive contributions of AI are indubitably beneficial for maximum efficiency, but our current economic, political, ethical, and even religious ideologies will have to adjust.
As Yuval Noah Harari explored the future of humanity in Homo Deus (2017), he philosophized on the new data religion. He explained the concept in this manner:
Religion : Dataism
God: Data
Main belief: Freedom of information is the greatest good of all. This is different from freedom of expression; freedom is given to information, not humans.
Commandments: (1) A Dataist ought to maximize data flow by connecting to more and more media, and producing and consuming more and more information. (2) A Dataist ought to link everything and everyone to the system.
Greatest sin: Block data flow (the right of information is prioritized over the right of humans to own data and restrict its movement)
Meaning of life: Human experiences are valueless if they are not shared, and that we need not – indeed cannot – find meaning within ourselves. As the global data-processing system becomes all-knowing and all-powerful, so connecting to the system becomes the source of all meaning.
Motto: If you experience something, record it. If you record something, upload it. If you upload something, share it.
If we will still fully own our private lives, AI cannot also reach its full potential. As argued, it needs large sets of data to become stronger. This technology becomes more intelligent as it combines information from all individuals to serve the collective.
Time will surely come that we will listen to AI algorithms more than we listen to our individual selves. In fact, this is now manifesting: we trust the health insights of the AI in our smartwatches more than how we feel with our bodies, and we believe the AI in GPS navigation applications more than our personal knowledge on street directions and maps. The purpose of AI is really to know us more than we know ourselves, and to know what we do not.
As everyone (and even everything, including cars) becomes one through AI and the Internet, are we headed to lose our selves?
My questions are related to my previous findings on defining myself:
I am not my thoughts and emotions. I am who watches it. | Can intelligent AI become conscious too? If not, will it still make better decisions than humans? |
I am nothing, but this also makes me everything. | As AI demands data from all individuals, will privacy be obliterated? What will happen to free will in making choices? |
I am not existing to pursue happiness. I always have it. | What will AI prioritize, individual happiness or collective happiness? What will make us happy? |
I am not meant to find myself. It has always been with me. | As AI finds itself in all human tasks, what will happen to the workforce? What will happen to the government? |
Ironically, I am not meant to define myself to be me. | Will the "individual" be redefined as a "collective?" |
I always view the audience reach of each technology as a metaphor for how it will impact humanity. As AI unites us all, are we about to create a flag for the entire planet Earth? This is positive, but we still have a long way to go, and a lot to do.
Click my pic below to return at the Digital Galaxy main page. *lol, judged*