Unit 1: Introduction to Artificial Intelligence

• What is AI?
• Daily uses of AI?
• AI- good or bad?
• Various Concepts in AI
• AI  ML DL and Data Science
• AI Ethics and Impact to society
Unit 1: Introduction to Artificial Intelligence

• What is AI?
Intelligence: Intelligence is the ability to acquire knowledge, understand concepts, think
critically, solve problems, and adapt to new situations. It encompasses various cognitive
processes, including reasoning, memory, learning, and perception. Intelligence enables humans
and certain animals to interact with their environment, make decisions, and exhibit complex
behaviours. It involves the capacity to process information, apply logical reasoning, demonstrate
creativity, and learn from experience. Intelligence is a fundamental characteristic that
distinguishes humans and plays a vital role in our ability to navigate and thrive in the world.

Artificial Intelligence: Artificial Intelligence (AI) refers to the development and implementation
of computer systems that can perform tasks that typically require human intelligence. It aims to
create machines capable of reasoning, learning, problem-solving, and decision-making in ways
similar to humans. AI encompasses various subfields, such as machine learning, natural language
processing, computer vision, and robotics. By utilizing algorithms and vast amounts of data, AI
systems can analyze, interpret, and generate information to automate processes, recognize
patterns, provide recommendations, and even simulate human-like behaviour. The goal of AI is to
develop intelligent machines that can assist humans, enhance productivity, and tackle complex
challenges across multiple domains.
Unit 1: Introduction to Artificial Intelligence

• AI- good or bad?

Unit 1: Introduction to Artificial Intelligence

• AI- good or bad?

Unit 1: Introduction to Artificial Intelligence

• AI  ML DL and Data Science

Unit 1: Introduction to Artificial Intelligence

• Various Concepts in AI

Narrow AI vs. General AI:

Machine Learning:

Deep Learning:
Unit 1: Introduction to Artificial Intelligence

• Narrow AI (Weak AI)

• Designed and programmed to perform a specific task or set of tasks with a high level of
proficiency. These systems excel at a particular domain but lack general cognitive abilities.

• Scope: Narrow AI is focused and specialized. It can perform specific

functions like language translation, image recognition, or playing chess.

• Learning: Narrow AI systems typically use machine learning techniques, such as supervised
or unsupervised learning, to improve their performance within their designated task.
Unit 1: Introduction to Artificial Intelligence

• General AI (Strong AI)

• AI systems that have human-level cognitive abilities. These systems can understand, learn,
and apply knowledge across various domains, similar to human intelligence.

• Scope: General AI is not limited to a specific task and can handle a wide range of activities,
just like a human being. It has the potential to excel in multiple intellectual tasks.

• Learning: A true General AI would possess the ability to learn and adapt autonomously
across various domains, not just through supervised learning like Narrow AI but through
unsupervised and self-supervised learning as well.
Unit 1: Introduction to Artificial Intelligence

• Machine Learning

• Machine learning is a subset of artificial intelligence that focuses on the development of

algorithms and statistical models that enable computers to learn and improve their performance on
a specific task without being explicitly programmed for that task. Instead of following predefined
rules, machine learning systems use data to identify patterns, make decisions, and improve their
performance over time.

Traditional programming Machine Learning

Unit 1: Introduction to Artificial Intelligence

• AI Ethics and Impact to society

AI Bias:

Privacy Concerns:

Job displacement:
Unit 1: Introduction to Artificial Intelligence

• Advantages & Disadvantages https://www.tesla.com/AI

Unit 1: Introduction to Artificial Intelligence

• Home Work 1

Q: Imagine you are a billionaire with interest in giving back to society. You have to build 5 AI
systems that can help mankind…
Write down 5 ideas that you would like to bring in, how would you make the AI (source for training,
testing, validation), safety considerations and other considerations for a successful product
Unit 2: AI project cycle

• What is training a model?

https://code.org/oceans
Unit 2: AI project cycle

• Designing an Autonomous Car

Unit 2: AI project cycle

• Success Metrics

• Accuracy and Performance: Measure the accuracy and performance of the AI model. This
could be the percentage of correct predictions, the reduction in error rate, or the speed at
which the AI system processes data and provides outputs.
• Efficiency and Cost Savings: Evaluate whether AI implementation has led to increased
efficiency and cost savings compared to traditional methods. Calculate the time and resources
saved by using AI to perform tasks that were previously done manually.
• Return on Investment (ROI): Measure the financial impact of the AI project. Compare the
costs of implementing AI against the benefits and monetary gains achieved as a result of AI
implementation.
Unit 2: AI project cycle

• Success Metrics

• Automation Ratio: Determine the percentage of tasks or processes that have been
successfully automated through AI. A higher automation ratio indicates a more successful AI
implementation.
• Business Impact: Identify and quantify the positive impact of AI on the business or project
goals. This could be improved customer satisfaction, increased revenue, or enhanced
decision-making.
• Ethical and Fairness Metrics: Evaluate the ethical implications and fairness of the AI
system. Measure how well the AI avoids biased decision-making and adheres to ethical
guidelines.
Unit 2: AI project cycle

• Success Metrics

• Training Time and Maintenance Effort: Analyze the time and effort required to train and
maintain the AI model. A successful AI system should be manageable and sustainable in the
long term.
• Competitive Advantage: Determine whether the AI implementation has provided a
competitive advantage to the project or organization compared to competitors who haven't
adopted AI
Unit 2: AI project cycle

• Correlation vs Causation
Unit 2: AI project cycle

• Correlation vs Causation
Unit 2: AI project cycle

• Correlation vs Causation
Unit 2: AI project cycle

• Correlation vs Causation
Unit 2: AI project cycle

• Correlation vs Causation
Unit 2: AI project cycle

• Correlation vs Causation

• To establish casualty, three conditions must be met:

Temporal precedence: The cause (independent variable) must precede the effect
(dependent variable)
Empirical association: There must be an observed correlation between the two variables

Nonspuriousness: The observed correlation between the two variables cannot be due to a
third or outside variable
Unit 2: AI project cycle
Unit 2: AI project cycle
Unit 2: AI project cycle

• Train, Test and Validate

Unit 1: Introduction to Artificial Intelligence

• Home Work 1

Q: Imagine you are a billionaire with interest in giving back to society. You have to build 5 AI
systems that can help mankind…
Write down 5 ideas that you would like to bring in, how would you make the AI (source for training,
testing, validation), safety considerations and other considerations for a successful product
Unit 2: AI project cycle

• Data Acquisition

• Build an AI bot that can identify the song from the first few seconds of the clip

• Build autonomous driving cars

Unit 2: AI project cycle

• Data Acquisition

• Build a seat belt detection AI

Unit 2: AI project cycle

• Data Acquisition

• Build a seat belt detection AI

• 1. Data Collection: The AI system needs data to learn how to recognize seatbelt usage. This data could
include images or videos captured from cameras inside the vehicle. These images show people in different
positions, some wearing seatbelts and some not.

• 2. Data Annotation: Human annotators mark or label the images to indicate whether each person is wearing
a seatbelt or not. This labelled data helps the AI learn the difference between the two states.

• 3. Training the AI Model: The AI system uses the labeled data to train a machine learning model. This
model learns to identify patterns and features that distinguish between a person wearing a seatbelt and a
person not wearing one. It learns to recognize visual cues like the presence of the seatbelt strap, its position,
and its connection to the buckle.
Unit 2: AI project cycle

• Data Acquisition

• Build a seat belt detection AI

• 4. Prediction and Detection: Once the AI model is trained, it can analyze new images or video frames in
real-time. It scans the images for the same visual features it learned during training. If it detects a person
without a seatbelt, it can flag that detection.

• 5. Integration with Vehicle Systems: The AI system can be integrated with the vehicle's onboard computer
or safety systems. If the system detects that a person is not wearing a seatbelt, it can trigger alerts such as
sounds, visual warnings, or notifications to the driver. In some cases, it could even interact with vehicle
controls to limit functionality until the seatbelt is fastened.

• 6. Continuous Learning: To improve accuracy over time, the AI system can continue to learn from new
data. As more images and videos are collected, the system can be periodically retrained to become better at
identifying seatbelt usage.
Unit 2: AI project cycle

• Data Acquisition

• Data: Quality vs Quantity

• What is data and why is it important?

• Different types of available data • Ethical considerations

• Sources of Data • Data pre-processing

Unit 2: AI project cycle

https://datavizcatalogue.com/
• Data Visualization
Unit 2: AI project cycle

• Data Visualization

BAR CHARTS PIE CHARTS SCATTER CHARTS

Unit 2: AI project cycle

• Data Visualization

• Bar charts: Comparison of Values –

Show periodical changes

• Line charts: Comparison of Trends –

Show changes over a period of time

• Histogram: Distribution of Data according to

categories – Show data according to category

• Pie chart: Highlight a portion of a whole –

Highlight data according to value