we all know about the famous A.I-based virtual assistant service of Marvel comic hero Tony Stark also known as Iron Man. One of the many things many of us idolise Iron Man as our favourite Marvel superhero is his artificial intelligence assistant system called J.A.R.V.I.S(Just A Rather Very Intelligent System). According to comics and MCU movies, JARVIS is capable of performing many powerful automated tasks such as controlling almost everything in the Stark household, running his business decisions, controlling his iron man suits etc.
That whole thing felt like a futuristic sci-fi movie to many of us who watched these movies for the first time. But is it actually just a Sci-fi movie?. before answering this, We must deeply research about the concept of JARVIS. Then we can test the real life application of it.
What is JARVIS?
First we must understand that JARVIS is more than just an AI assistant. JARVIS can almost think like a human and act as human in many aspect although it did not have a physical body until the suits. The main purpose of JARVIS is to make the human lives much more comfortable and easier with it assistant protocols. JARVIS is like a butler but not a living person.
Theoretically, JARVIS like AI systems can be achieved. But practically we are still far away from creating a near human general AI systems. There are so many factors we need to address to make these kind of systems.
So, let’s check how we can create these systems and how far we have came so far.
Main features of JARVIS.
JARVIS is designed to perform more advanced functions.
JARVIS is a very intelligent AI system that needs a very large amount of input data to process. This means Jarvis needs the most advanced and compatible hardware units to operate quickly and continuously.
JARVIS is heavily dependent on two subsets of the field of Artificial Intelligence; Machine Learning and natural language processing
1. Machine Learning
Machine learning is basically giving machines/ computers the ability to learn by inputting data and experiences, just like humans. This is a subset of the field of AI. There are several applications of machine learning such as data classification, regression, anomaly detection, clustering, reinforcement learning, etc. For example, voice and facial recognition is designed by using machine learning algorithms and this is a very big feature in an AI system like JARVIS.
2. Natural Language Processing
This is a subfield of Artificial intelligence. Basically in NLP, machine learning algorithms are being used to give computers the ability to learn and communicate in natural human languages rather than machine languages. NLP can enable computer devices to recognize human voices, understand the language as input data and communicate back to humans in the same language. It is just like two humans talking to each other. There are already voice assistant services that use NLP such as Apple Siri and Amazon Alexa. There is also a good demand for these applications and devices.
Besides these two main components, there are some other specifications too.
Computer Vision:
Just like in the name, this is about seeing the physical world through computers. Computer vision enables computers to detect and recognise the physical world through artificial cameras. It is equivalent to a human being seeing the world through eyes. So, computer vision is basically the process of using machine learning algorithms and deep neural networks to gather information by visual inputs(images and videos) and make decisions based on that input data just like a human eye and brain. Computer vision trains the computer devices to do this.
Internet of Things (IoT):
This is a very important part when creating a JARVIS-like AI system. Simply, IoT is controlling various devices, software and hardware and transferring data among them. Furthermore, IoT is actually a network of interrelated devices and applications that transfer data with other devices or applications or with the cloud. Examples of IoT devices are sensors and actuators. These devices can gather input data and act accordingly by automating some tasks. If we think about it, that is one of the major qualities of JARVIS.
Real-world problems of creating JARVIS:
Now we can understand that creating an intelligent AI system like JARVIS is not entirely a fantasy. There are already practical applications of AI. With that being said, we must realise that we are still limited by the technology of our time.
JARVIS is operated by a vast amount of Input data and information. As a generation lives with computers daily basis, we know it requires a large amount of processing power to handle that amount of datasets as these data will be gathered by various devices and applications. To do that we need more advanced hardware such as microprocessors. We already know that AI needs advanced hardware to operate. If you want to know more about the relationship between AI and hardware, check this article.
Safety and security are one major considerations. Today we can hear regularly about cyber attacks and hackings. So, if we create a system that is entirely operated by computers, we will become more vulnerable to these threats. Because thieves will be able to take control of your house or office without even a physical appearance.
Just like that, we have to consider about many more technical and social problems before creating these kind of AI system.
IS it even possible?
So, now we have discussed about the technical specifications and challenges. Finally we can decide whether it is possible or not to create system. Well it is not impossible but very complex process from designing to implementing.
Many researchers, engineers and scientists have made a significant progress in achieving this goal of creating advanced AI. But, we are still not there yet.
As I mentioned before and just like Howard Stark said, we are still limited by the technology of our time.