Artificial Intelligence

Trying to create artificial intelligence is trying to create a being that is different than us, yet somehow similar. It’s like some alien beings come to earth, except, we are the creator.

The simplest way to create a being with intelligence is to make a human baby. If the education’s done right, intelligence will follow up. We biologically complex brain tissues with bipedal mechanism creatures seem to be the highest form of intelligence. So, if we want to achieve a manmade intelligence, what would it be, than simply just a human baby?

Artificial intelligence means to accomplish greater feats us humans simply can not: It functions 24/7 non-stop as long as we feed it energy. It can collate and analyze massive data in short time. It is intelligent.

In short, artificial intelligence might be the most powerful tool we mankind have ever created.

Hypothetically, let’s try to make a human baby brain, a machine that possesses the brain complexity of a baby.

Thus machine must fulfill three basic traits to be a human baby brain:

  1. Basic language
  2. Basic emotions
  3. Basic wants

Language is the start of everything, the start of human civilization. A machine currently does not necessarily have to have the comprehension of abstract concepts, it needs only to understand imagery words like “food”, “water” or “data”. Wolf children raised by wolves don’t understand human languages, but they know what they need. Even the most basic language is useful in engaging with human activities and interacting with humans. We don’t need machines that don’t understand us.

Emotions are what define us humans, emotionless seems to be the case for many AI we imagine it to be, in films, novels and media generally. Without emotions, one simply does not appreciate the normal human behaviors and common activities we engage in.

Wants, or needs, are the primary reason we can learn. The wants are proactive, spontaneous. Because we lack something and we have the need to fulfill it. It is simply distinguished from automation. Although right now we can make a machine automatically collect data, process data and conclude outcomes, it is not a spontaneous activity, it is a mere human manual labor enforced on the machine. The machine does not collect data out of its own need of absorbing information, it has no understanding of its lack of resources, it does not proactively engage in the activities.

Thus, we need to solve these 3 basic problems for a true artificial intelligence to work on its own.

Philosophically, through out the human civilization, the human nature has not changed. We suffer from our very own desires–greed, envy, sloth, lust, pride, wrath, gluttony. Yet, we want to create something beyond ourselves.

It is a paradox to think about it.

The current machine learning means to create a powerful problem solving tool. It is different from the artificial intelligence we’re talking about here.

But it’s not wrong. Humans make tools, and each generation comes more advanced tools to be used.

What current machine learning tries to solve, are simpler problems, compared to the aforementioned problems. Machine learning aims to create an environment, a platform, for programs to filter and select, the best algorithm to be implemented in the reality. The more data, the higher the accuracy it can provide.

It is a process to slowly replace manual coding, to let machines generate codes for itself. The issues programmers need to solve are just how to set parameters, and this environment for the programs to collect, process and generate.

Leave a Reply

Your email address will not be published. Required fields are marked *