How To Be Series

How to be an AI & ML expert: A Webinar with Cloud Architect Subhendu Dey

Our Correspondent
Our Correspondent
Posted on 25 Jul 2022
19:09 PM

Live recording of the How to be an AI & ML Expert Webinar

Advertisement
Summary
The webinar outlined what AI and ML mean in today’s world and how students could get involved
Mr Subhendu Dey also laid out a comprehensive roadmap for those looking to start a career in AI and ML

Artificial Intelligence and Machine Learning as disciplines have taken the world by storm, particularly in the 21st century. While many youngsters have drawn inspiration from some of the best science fiction featuring AI and robots, the actual world of AI and ML has been growing by leaps and bounds. But what does the world of AI and ML have to offer? How can you transition from campus to career with AL & ML? And how can you be an expert in AI & ML? To answer these and many other questions, The Telegraph Online Edugraph organised a webinar with Subhendu Dey, a Cloud Architect and advisor on Data and AI.

The webinar saw participants from class 8 right up to those in advanced degrees, as well as teachers. Hence, the subject matter of the webinar contained takeaways that would be relevant at all stages. Mr Dey also highlighted that he would be focusing on showing how things that have always existed around us contribute to AI - giving students a more intuitive idea of AI and making it more interesting.

The webinar started by taking a look at a simple action like sending a text. People would find that their mobiles would keep suggesting words to them. Be it as soon as they have typed a few letters or after they have typed a few words, they would get suggestions that are surprisingly accurate. This is called Language Modelling and requires an intuitive understanding of language. A human may be able to do it from his or her extensive knowledge of words and language, but in this case, it is a fine demonstration of the intuitiveness of AI.

Advertisement

Let’s look at another aspect of AI - when we key in a question into the Google search bar, a decade or so ago, Google would have analysed the keywords and thrown up a list of links that feature the keywords. But fast-forward to this decade and Natural Language search is today capable of not just reading the keywords but also finding out the intent behind the query. This means that Google will, in addition to giving you the links, also give you the answer, as well as other questions that have the same or related intent. In fact, Google also has a system for taking feedback, which facilitates the Google AI to learn to be even more intuitive and better at giving suggestions.

The webinar was conducted by Subhendu Dey

The webinar was conducted by Subhendu Dey Edugraph

One need only look at the digital assistant - Siri, Google Assistant or Alexa - to understand the advancements in AI. From understanding spoken queries to giving intuitive, and often very witty, answers, these assistants communicate in a surprisingly human-like manner. Of course, there is a cycle of tasks that they must perform behind the scenes, which Mr Dey spoke about in detail.

While these changes that we can observe are new, AI has been around for a long time now. One of the earliest feats was in 1997, when the IBM Supercomputer Deep Blue beat world chess champion Gary Kasparov, in a six-match tournament.

Today artificial intelligence is a booming area of development and the Ministry of Electronics & Information Technology projects the addition of about 20 million jobs in the sector by 2025. In fact, this is also underscored by multiple studies and reports prepared by global auditing firms like Deloitte, NASSCOM and PwC.

However, one question that has always baffled scientists and engineers working in the domain of AI, is striking a balance between behaviour and reasoning on the one hand and human/irrational and rational on the other, when designing the various Artificial Intelligence agents. It has, however, been found that more intuitive AI agents with better user experience interfaces have a higher penetration in human society.

Next we take a look at Machine Learning. When an AI agent learns on its own from the interactions it has, this is known as Machine Learning. When humans learn something, it registers in some form in the mind. However, machines perceive data in the form of functions and variables. With Machine Learning, AI agents create models which exist as executable software components made up of a sequence of mathematical variables and functions. Hence, becoming an expert in AI and ML usually requires a person to have a sound understanding of mathematics and statistics.

Speaking of building a career in AI and ML, Mr Dey threw light on three avenues into the industry. These are:

  • As a scientist
  • As an engineers
  • As a contributor

Let’s take a look at each of these.

As a Scientist

As mentioned above, to communicate with AI, your query must be represented in a mathematical/logical format. Hence, when choosing your educational degrees or courses, go for courses that cover the following topics which contribute to the core of AI:

  • Vectors and Matrices
  • Probability
  • Relation and Function
  • Differential Calculus
  • Statistical Analysis

Choosing a major which covers these aspects should arm you with the knowledge and skills you need to become a scientist in AI.

As an Engineer

Being a scientist is not your only option, though. AI also depends heavily on engineers to grow and develop. From the engineering perspective, here is a list of functions that need to be carried out:

  • Visualisation/representation of data
  • Collection of data from multiple sources
  • Building pipelines to prepare data to scale
  • Using Machine Learning services/frameworks available on clouds to scale up
  • Test, audit and explain to various stakeholders the Machine Learning output

As a contributor

If you find you are not interested in being a scientist of an engineer, there are other significant ways you can contribute to AI. That could be in the following areas:

  • User experience design
  • Process modelling
  • Domain knowledge
  • Linguistic details
  • Social aspects

Mr Dey discusses all these avenues at length in the course of the webinar with examples. At the same time, he lays out the basic qualities that one must have - irrespective of which role one chooses to pursue. And these are creative vision, innate curiosity and perseverance.

Here are some courses that you should explore if you want to build a career in the core AI aspects:

  • A Bachelors or Masters degree in Computer Science or Engineering or Mathematics or Statistics.
  • A specialisation in any of the following areas:
    • Artificial Intelligence
    • Machine Learning
    • Data Science
    • Automation and Robotics
  • B Tech/ BE in other engineering fields, followed by work experience in the field of software or IT.
  • Artificial Intelligence
  • Machine Learning
  • Data Science
  • Automation and Robotics

The webinar ended with a detailed Q&A session which opened with some questions received from participants submitted at the time of registration and carried on to questions asked by participants in the course of the webinar. The Q&A covered a range of interesting topics like:

  • Neural networks/deep learning
  • Importance of Maths and Statistics in AI/ML
  • How valuable are practical projects for developing skills needed to work in AI/ML
  • Which programming language is the best to learn for a career working with AI/ML
  • Which are the best courses to consider as a student - traditional degrees or online certification courses
  • How does AI compare to the human brain
  • Will AI and automation endanger human jobs in the future
  • What are intelligent agents and how are they useful in AI

To learn the answer to these and many more questions, watch our video recording of the live webinar.

A career in AI and ML is an excellent choice now - and this small initiative of The Telegraph Edugraph was aimed at providing the right guidance for you to make the transition from Campus to Career. Best of luck!

Last updated on 26 Jul 2022
10:12 AM
Advertisement
Read Next