Tech space is constantly evolving and growing at a speed which no one had anticipated. So, when in late 20th Century the study of AI began, no one expected that within a decade it would become such an important part of the entire tech industry.
Even though AI can never fully replace humans, the AI integrated tools today are capable of gathering useful customer patterns from data, address to a customer’s needs and even come up optimal solutions as and when required.
Everyone has a different stand on AI, but at the end of the day we all agree that AI is helping us in ways which were manually not possible. Let us take a deeper look into this field and try to understand what it is all about.
What is AI?
AI stands for Artificial Intelligence, and it’s technical definition is something like this
The theory and development of the computer systems to be able to perform tasks which normally require human intelligence, like speech recognition, decision-making, visual perception and even translation between languages.
All in all, the purpose of AI is to make machines “smart” and perform tasks which would otherwise cannot be performed without human intelligence.
AI is generally made out of a process called Machine Learning, a discipline that explores the complexities and algorithms of learning from data.
Artificial Intelligence or any Intelligence has five basic elements:
- Learning : There are many forms of learning but AI mainly emphasises on the trial and error mode of learning.
- Reasoning : Reasoning involves drawing inferences that are relevant to the task or situation in hand.
- Problem Solving : Problem-solving is a huge part of AI and is incorporated in a variety of methods such as: finding the winning move in a board game, or identifying a person from the given photograph.
- Perception : Perception requires scanning of the environment by the means of various sense organs, real or artificial. The data gathered is then used to identify the object or environment by analysing its features.
- Language understanding : A language is a system of signs having meaning by convention. Every AI requires a language that can be interpreted in a productive manner.
Types of AI
Now, let us take a look at the four types of AI
1. Reactive Machine : This is the most basic type of AI system, in which the system is purely reactive and doesn’t have the ability to have memories or use past experiences to form current decisions. The greatest example would be “Deep Blue,” IBM’s chess-playing supercomputer that beat the grandmaster Garry Kasparov in late 1990s.
Deep Blue can identify the different pieces on the chess board and also knows how each of those pieces are supposed to move. It can easily make predictions as well about the next move and also select the most optimal next move. While it can do all of the Deep Blue doesn’t have any memory of what has happened before, and it ignores everything before the present moment. It only looks at the pieces on the chess board as it stands right now and choose from next moves.
This intelligence works on perceiving the world directly and only acting on what it sees. These systems don’t participate in the world quire interactively, instead they behave exactly the same way every time they encounter the same situation.
2. Limited Memory : These AI systems can use past experiences. Self-driving cars already have this AI integrated in them. For example, they observe the speed and direction of other cars and this can’t be done in a single moment. They identify various objects and monitor them for a period of time.
These observations are added to the pre-programmed representations of the world which are already present in the cars, such as the lane markings, traffic lights and other important elements. These elements are include, so that when a car decides to change lanes, it can avoid cutting off another driver or getting hit by a nearby car.
While it can use past experiences, these pieces of information about the past are transient. They aren’t a part of the car’s library of experience from which it could learn, in the way humans compile their experience over years behind the wheel.
3. Theory of Mind : This is the point where we differentiate between the machines that we are using at this moment and the machines that we will have in the future. The different types of AI systems, when ready in the future, should be able to recognise people’s emotions, beliefs, thoughts and should be able to have a proper social interaction.
In psychology, ‘Theory of Mind’ means understanding that people, creatures and objects in the world can have emotions and thoughts which can affect their own behaviour. This theory is quite important to how human formed societies as they allowed us to have different types of social interactions.
If we fail to understand each other’s motives and intentions, and don’t take into consideration what someone might know about something or the environment it would become impossible to co-exist.
If ever AI systems are going to walk amongst us, then they will have to understand that each individual has thoughts, feelings and also expectations of how we should be treated. And that they will have to adjust their behaviour accordingly.
4. Self – Awareness : The final step of AI development would be to build systems which can form representations about themselves. Ultimately, AI researchers would not only have to understand what makes consciousness but also work on building machines which have consciousness.
Alternatively, you can even consider this type of AI as an extension of “Theory of Mind”. Conscious beings are very well aware of themselves, they know their internal states and are able to predict the feelings of others. If you take an everyday example of someone honking behind us in traffic, we automatically assume that the person is impatient or even angry. We are able to arrive at these conclusions because that is how we feel when we honk at someone else in the traffic. Without theory of mind, we won’t be able to make such inferences.
We are quite far from creating machines which are self aware and can completely replicate human intelligence.
Pursuing a career in AI
All the major companies are looking to adapt AI technologies and the demand for AI professionals is increasing everyday. The median salary of an AI professional is approximately INR 14.3 Lakhs per annum. The entry level salary in the field is around INR 6 lakhs per annum, while senior-level professionals can expect to earn INR 50 lakhs per annum.
AI is a field that has been in demand for quite some time now and with many major companies using AI technologies, like SIRI by Apple, the demand for AI specialists is only bound to increase.
Let us take a look at how you can enter this field
Step 1: Get a Bachelor’s Degree
You need to have a bachelor’s degree in mathematics and computer science. This would help you land an entry level job, but having this degree will help you go a long way in your career.
Step 2: Understand the different roles and select your specialisation
The second step towards your AI career would be to understand the different job roles that fall under the field of AI. The most popular AI job roles are:
- Machine Learning Engineers : They are responsible for building and managing various platforms for machine learning projects. They should be able to make use of predictive models and even be able to use NLPs (natural language processing) when they work with massive datasets.
- Data Scientist : They are responsible to collecting, analysing and interpreting large datasets by using both machine learning as well as predictive analysis. They also work on developing algorithms which enable the collection and cleaning of data for proper analysis.
- Business Intelligence Developer : The primary role of a business intelligence developer is to analyse complex data sets in order to identify various market and business trends. They are also responsible for designing, modelling and maintaining complex data in highly accessible cloud-based data platforms.
- Research Scientist : They are experts in multiple disciplines, which include mathematics, deep learning, machine learning and even computational statistics.
- Big Data Engineer : They work on developing the ecosystem which enables various business systems to communicate with each other ans also collate data.
Now that you know about these fields, you need to pick one and specialise in that field. In order to get into the field of AI, you need to have a Master’s Degree or a Doctoral Degree. Some of the institutes that offer Master courses in AI in India are:
- University of Hyderabad
- IIT Bombay
- IIT Madras
- IISc Bangalore
- ISI Kolkata
Tips for Beginners to learn AI
1. Understand Numbers : You need to have an understanding of various statistical concepts, so you know how and when to use your data effectively. Some of the ideal statistical learning topics you can start with are
- Mean & Distribution
- Statistical Decision Theory
- Mean Square Error, Least Squares
2. Learn a Programming Language : Learning a programming language can seem like a long process, but it isn’t. You need to find a program which is popular, easy to learn and also commonly used for machine learning and data analytics, such as Python or R.
3. Understand the basics of Machine Learning : Machine learning deals with processing a lot of data and it involves specific steps. You will need to understand basic concepts such as data science, programming, algorithms etc. You will also need to invest some time in understanding the basics of machine learning and data science.
4. Perform Exploratory Data Analysis : You need to study a dataset and understand the shape of data, feature correlations and also the various signals within the data which can be used to build predictive models. This analysis can help you determine how to improve the products, understand user behaviour etc.
5. Look into Deep Learning Models : The deep learning algorithm is built with connected layers, in such a manner that the neural network can learn complex data features as it goes through each layer. With deep learning you can turn predictions into actionable results as it can help with pattern discovery and knowledge-based predictions
Biggest AI tech in the world
Now that you know how and why AI is becoming such a hot topic these days, its time to take a look at some of the hottest AI technologies in India (get excited):
- Manthan : Founded in 2003, Manthan’s AI-powered retail analytics platform provides descriptive analysis for users, helps to grow the customer engagement and recommend necessary actions.
- SigTuple : With the help of AI and Machine Learning technology, they have developed medical diagnostic solutions. It can perform screening and advanced diagnosis of urine and blood samples along with X-rays and retinal scans
- Haptik : They specialise in developing AI based chatbots for enterprises, service companies and consumers
- Uncanny Vision : An AI-based Computer Vision startup focused of making surveillance cameras smarter with real-time, edge-based intelligence for a safer and secure world.
- Arya.ai : They help other AI based startups in building and solving complex problems at a faster pace. They serve various industries such as Banking, Insurance, Health Care, Retail and Oil & Gas.
Want to learn explore the world of AI yourself, then sign up for the MyCaptain Artificial Intelligence Workshop.