1

READING COMPREHENSION
Elon Musk, a tech entrepreneur, says artificial intelligence (AI) is a bigger threat to humanity
than nuclear weapons. On the other hand, Mark Zuckerberg, the CEO of Facebook, thinks AI
will save lives. Regardless of whether you agree with Musk’s or Zuckerberg’s argument, what is
clear is that AI is reshaping the world we live in. It's leading to improvements in medicine and
self-driving cars, shaking up businesses from manufacturing to marketing. And although we may
not be aware of it, we already use it in our Facebook news feeds when we talk to Siri on our
iPhones or ask our Alexa-enabled speakers to play a track. Yet, although it powers products and
services we use every day, AI remains a mystery to many. So Facebook AI experts Yann LeCun
and Joaquin Quiñonero Candela have set about simplifying this complex field of computer
science in a series of educational videos and blog posts. ‘Not magic, just code’ “Artificial
intelligence is not magic,” write LeCun, head of Facebook’s AI research, and Candela,
Facebook’s Director of Applied Machine Learning, in their blog. "But we've already seen how it
can make scientific discoveries that seem like magic and help us do everyday things like identify
objects in photos, recognize speech, drive a car, or translate an online post into dozens of
languages." How does it do this? Many intelligent machines and systems use algorithmic
techniques loosely based on the human brain. These neural networks can learn to recognize
patterns, translate languages, do simple logical reasoning, create images, and even come up with
ideas.
“All of this happens at blinding speed through a set of coded programs designed to run neural
networks with millions of units and billions of connections,” write LeCun and Candela.
“Intelligence emerges out of the interaction between this large number of simple elements.” If,
for example, you want to teach a computer to tell the difference between a car and a dog, instead
of programming it to carry out the task, you can train it to recognize objects in images so that it
learns for itself. So what is deep learning?
Deep learning is a type of machine learning that structures neural networks into multiple
processing layers. This helps a computer identify what is in an image or learn to recognize
speech and text. “In a park, we can see a collie and a chihuahua, but recognize them both as
dogs, despite their size and weight variations,” write LeCun and Candela. “To a computer, an
image is simply an array of numbers." Local patterns, like the edge of an object, are easy to spot
in the first layer of this set of numbers. “The next layer would detect combinations of these
simple motifs that form simple shapes, like the wheel of a car or the eyes in a face." "The next
layer will detect shape combinations that form parts of objects, such as a face, a leg, or an
airplane wing." “The last layer would detect combinations of parts that form objects: a car, an
airplane, a person, a dog, etc. The depth of the network, with its multiple layers, allows it to
recognize complex patterns in this hierarchical fashion.” Deep learning is helping to push
forward research in fields including physics, engineering, biology, and medicine. It is also at the
heart of developing autonomous systems such as self-driving vehicles. What about AI’s impact
on jobs? Despite growing anxiety over automation eliminating jobs, LeCun and Candela believe
that AI will create new roles for humans in manufacturing, training, sales, maintenance, and
management of intelligent robots. “AI and robots will enable the creation of new services that are
difficult to imagine today." "But health care and transportation will be among the first industries
to be completely transformed by it,” they write. “Increasingly, human intellectual activities will
 2

be performed in conjunction with intelligent machines." "Our intelligence makes us human, and
AI is an extension of that quality.”
Answer the following questions:
(a) How does artificial intelligence (AI) facilitate scientific discoveries and everyday tasks
and what fundamental principles underlie its functioning?

(b) What role does deep learning play in AI, particularly in tasks such as image recognition
and language processing, and how does it mimic the human brain's processing mechanism?

(c) Give a suitable title of the passage with justification.

(d) Give the meanings of the following words as used in the passage.

Manufacturing, emerge, detect, hierarchical, eliminating

Sample Answers

(a) Artificial intelligence (AI) facilitates scientific discoveries and everyday tasks through
algorithmic techniques, particularly neural networks loosely based on the human brain. These
networks can learn to recognize patterns, translate languages, perform logical reasoning, generate
images, and more. Fundamentally, AI operates through the interaction of a large number
of simple elements, such as neural network units and connections, which execute coded
programs at high speeds. For instance, instead of explicitly programming a computer to
differentiate between a car and a dog, AI can be trained to recognize objects in images, allowing
it to learn autonomously. This approach enables AI to accomplish tasks like identifying objects
in photos, recognizing speech, driving cars, or translating text into multiple languages.
(b) Deep learning, a subset of machine learning, plays a crucial role in AI, particularly in tasks
such as image recognition and language processing. Deep learning structures neural networks
into multiple layers of processing, mimicking the hierarchical processing mechanism of the
human brain. In this architecture, each layer detects increasingly complex features or patterns
within data. For example, in image recognition, the first layer may identify local patterns like
 3

edges, while subsequent layers detect combinations of these patterns to recognize shapes and
objects. This hierarchical processing allows deep learning models to recognize intricate patterns
and features, contributing to advancements in various fields like physics, engineering, biology,
and medicine. Additionally, deep learning is pivotal in developing autonomous systems such as
self-driving vehicles, as it enables these systems to understand and interpret complex data inputs.

(a) AI facilitates tasks by using neural networks to learn patterns and execute coded programs.
Fundamental principles include neural network interaction and autonomous learning.
(b) Deep learning structures neural networks into layers, mimicking the brain's hierarchical
processing. It recognizes complex patterns in data, crucial for tasks like image recognition and
language processing.

(c) Title: "Demystifying (clarifying) Artificial Intelligence: From Everyday Tasks to Scientific
Breakthroughs"
Justification: The passage covers various aspects of artificial intelligence, from its practical
applications in everyday tasks like image recognition and language processing to its role
in facilitating scientific discoveries. Additionally, it delves into fundamental principles
underlying AI functioning and discusses the impact of deep learning, a key component of AI, in
mimicking the human brain's processing mechanism. Therefore, the title "Demystifying
Artificial Intelligence" reflects the aim of simplifying this complex field for readers, while
highlighting its broad scope and relevance.

● Manufacturing: The process of producing goods, typically in large quantities, through

machinery, labor, and raw materials. The context of the passage refers to the production
of goods, which AI is expected to impact through automation.
● Emerge: To come into existence or become apparent. In the passage, it refers to
intelligence arising from the interaction between simple elements within AI systems.
● Detect: To identify or discover something, often through observation or analysis. In
the passage, it refers to the ability of neural networks to recognize patterns or features
within data.
● Hierarchical: Arranged in a graded or ranked series. In the passage, it describes the
structure of deep learning, where neural networks are organized into layers that process
data in a hierarchical manner.
● Eliminating: To remove or get rid of something completely. In the context of the
passage, it refers to the concern over automation in AI potentially leading to job loss by
replacing human labor.
4