Introduction to Artificial Intelligence with Talvinder Singh

In today’s dynamic product environment, AI takes the center stage when it comes to biting through the competition and building a rewarding product. 

We organized an exclusive session with our founder Talvinder Singh where he answered all your questions about the course as well as everything under the sky that you can relate to Artificial Intelligence for Product Managers.

We started the session with understanding the basics of AI. Talvinder says It’s very important to have a good network and hardware for a good production usage of AI and ML as we need to compute a lot of data at a very high speed. 

He says, AI is a bunch of algorithms that make the decision about data. These algorithms enable computers to perform the task automatically without human intervention. The main goal of AI is to create expert systems and create systems that behave and interact like humans. Applications of AI include gaming, natural language processing, speech and handwriting recognition etc and components of the AI system include agents and environments in which agents function.

Further, he moved towards agents. He explained the agent is a programme that can sense its surrounding through sensors. Sensors can be simple variable sent to the programmes example IOT, robots etc. Every agent has structure and structure has two parts 1) architecture- machinery on which agent works, 2)  agent programme- implementation of an agent function.

Getting deep into it, Talvinder explained in brief about several types of agents like:

Simple reflex agents – Takes action based on the current environment. It does not take into account historical actions.

Model-based reflex agents – It needs memory to store historical actions. It uses this historical memory to take into accounts unforeseeable aspects in the current environment.

Goal-based reflex agents – These agents have a goal and strategy to achieve that goal. All their actions are geared towards reaching the goal.

Utility-based reflex agents– The goal-based agents are concerned only of achieving the goal, without regarding if the way achieved was the best.

Learning agents – These agents learn from the experience. They have the ability to acquire information and integrate it with the system.

Now comes the environments! He says environments are the ecosystem in which agents reside. They provide information to the agents. Artificial environments are confined by keyboard inputs.  

When asked about the environment, he explained multiple types of AI environments:

Complete v/s Incomplete

> Complete AI environment- Environment that has enough information to solve one entire branch of the problem.

> Incomplete AI environment– Not enough information to anticipate solutions in advance.

Fully observable v/s Partially observable

.> Fully observable environment –  Has access to all necessary information to complete a particular task.

> Partially observable environment– This deals with partial information to solve problems.

Competition v/s Collaborative

> Competition environment – AI agents are challenged by each other in the same environment in order to optimize a specific outcome.

> Collaborative environment– These environments thrive on the cooperation between agents.

Static v/s Dynamic

> Static AI environment – Rely on data sources that are not volatile

> Dynamic AI environment – These environments deal with data whose sources keep on changing.

Discrete v/s Continuous

>  Discrete AI environment – In these environments, a finite set of possibilities lead to the final outcome of the problem.

> Continuous AI environment – In these environments, solutions rely on unknowns and rapidly changing data sources.

Deterministic v/s Stochastic

>Deterministic AI environment –  The outcome can be determined on specific data.

> Stochastic AI environment – The outcomes are not based on specific data.

When asked the difference between Machine Learning and AI, Talvinder concludes the session explaining Machine learning as a method of data analysis that automates analytical model building i.e. a branch of artificial intelligence based on the idea that systems can learn from data, identify patterns and make decisions with minimal human intervention.

Check out this exclusive AMA video to know more about our program and what Talvinder Singh has to say about the program:

October 10, 2018
© Pragmatic Leaders. All rights reserved.
Select your currency
USD United States (US) dollar