How Chat-GPT Use Artificial Intelligence and Machine Learning?

How Chat-GPT Use AI and ML:

Let’s Begin by understanding this new technology in depth. Artificial Intelligence what is It and how does CHAT-GPT work exactly? It’s important to understand this so that later when we use this technology for our benefit and transform our lives by using it, we can understand its limitations. What are its boundaries? by understanding the process behind it, you will make yourself more future-proof because in future more advanced AI software will come. But if you understand the foundations, you will also be able to understand the basis on which future software would work.

This Word here, “Artificial Intelligence.” What does it mean? Does it mean that computer scientists and researcher have used computer to create intelligence? Just like humans have intelligence in their brains, has it been recreated on computer? Have they understood the human brain and replicated it on a computer? Not at All.

Roles of Scientists in AI and Machine Learning:

According to our current technology standards, scientists have not yet understood how the human brain functions exactly. It is constantly being researched. But one thing is for sure, scientists have tried to imitate some specific functions of the human brain. And they have been successful in this effort many times. 

For example, in playing chess. For long time, people thought that playing chess was something that can be done only humans. But in 1997 Artificial intelligence was used to invent a robot named deep blue which defeated the then chess champion, Garry Kasparov.

Also Read:  The Next Big thing in IT automation.

Evolution of AI and Machine Learning:

This was a big turning point because people start believing that it could be done. Some specific aspects of human cognitive functions can be carried out by computers. And be better that than humans at that.

So, if you look at the definition of artificial intelligence, it means that it "is the capability of a computer system to mimic human cognitive functions. For example, learning, problem-solving, and playing chess. The latest example, chatgpt, which human cognitive function is it mimicking? Talking like humans, being able to chat. Like human’s chat or write messages on WhatsApp. This function is being mimicked by this function is being mimicked by this artificial intelligence software.

What is Artificial Intelligence?

Now, the topic of artificial intelligence in itself is not a new topic. It has been talked about for a long time, since the 1950s and 60s after the invention of computers. However, it is only in recent times that its popularity among the public has increased. It was only after 2012-15 that our processing power increased so much that we were able to process such large amounts of data.

As a concept, scientists knew about for a long time but they were not able to test it practically because processing the data of thousands of GBs and TBs require a lot of computing power. There are many terms in the field that are mentioned repeatedly. One is ML or machine learning.

What is Machine Learning?

Machine Learning is a subset of Artificial intelligence you can call it a type of AI or Artificial intelligence. In which, as the name suggests, you are teaching something to the machine. ML is already being used I many places, even where you might not be able to guess. When you watch a tv show or film on Netflix, the suggestion that comes from Netflix, as “For You” is achieved through machine learning. When you scroll through social media feeds, the automatically suggested posts are also due to machine learning. It is also used in self-driving cars. Another such subtopic in artificial intelligence is NLP or the natural language processing. Natural language processing it means understanding human languages through a computer. How a computer can process natural language. Another ai type is artificial neural networks.

Here, the connection between the neurons in the human brain are imitated in the machines. When i use the word 'imitation' here, it doesn't mean that the computers should be made in the likeness of the neurons in our brain. It doesn't mean that. Here, imitation means that the way the behaviors and functions, the way the behaviors and functions, can be performed by humans to be able to generate same output through computers. If you understand this technology at a basic level, you will understand that this is not magical. This is quite simple, really. Usually, in traditional computer programming, we program the computer and define the function for it. We provide the input and the computer has to merely use the function to derive the output from the input.

Also Read: GPT-4 vs ChatGPT Which one is more powerful?

 For example, when we define the function of addition for the computer, and then provide the inputs to the computer of 2 and 2. The computer takes the input of 2 and 2 and uses the function of addition. It gives us the output,2 + 2 = 4.

How Machine Learning Work?

In machine learning we, provide both the input and output without defining the function that was used. We ask the computer to guess what the function should have been. We provide it with a large set of inputs and outputs. To enable it to recognize the pattern between these inputs and outputs and to derive the function used on its own.

Suppose we want the computer to recognize through machine learning how to write the number 4. So, in the form of input, we will provide it with samples of handwriting. On the various ways of writing 4. Different ways humans write 4. But the output of all the ways is the number 4. So, the pattern among the samples, through machine learning, the computer can recognize it by itself, by looking through different pixels.

Also Read: How AI and Automation Will Shape the Future of Work?

For example, the styles used in writing the ways it bends based on the handwriting but all of it represents 4. So, what will happen next is that the computer has created its own function. Next time, if there is a handwritten 4, then using this function it will be able to recognize that handwritten 4. Fundamentally, the computer is recognizing patterns. This allows us to use many existing algorithms with machine learning.

HOW CHAT GPT USE MACHINE LEARING?

If we look at this in a very simplified way, chat gpt is also doing the same. It is recognizing patterns. We have fed it thousands of GBS of training data. We have provided it with inputs and outputs. We have provided it with samples of how people talk to each other. How English sentence are formed it has recognized patterns it has discovered the same functions and now it is able to talk to us.