Exploring the Marvels of Artificial Intelligence: A Beginner's Guide
Artificial intelligence, often abbreviated as AI may sound like something out of a science
fiction movie, but it's actually a fascinating field of technology that's shaping our world in
many incredible ways.
Whether you're a curious child or a newcomer to the concept, understanding AI can be both
exciting and rewarding. In this article, we'll embark on a journey to demystify AI, exploring
its basic principles, fun applications, and how it impacts our daily lives.
Before we deep dive into understanding what exactly is artificial intelligence, and how does it
work and where we need it ,let us understand what we mean by “ intelligence”?
As mentioned by Max Tegmark, in Life 3.0: Being Human in the Age of Artificial
Intelligence “Intelligence: the ability to accomplish complex goals”.
Or as told by Dr. Jeff Hawkins, Neuroscientist, Intelligence is “The ability to see patterns
and predict outcomes based on previous experiences”.
For most of us though it’s imperative to “get on with it ” still a definition remains important
and Nils J. Nilsson has provided a useful one:
“Artificial intelligence is that activity devoted to making machines intelligent, and
intelligence is that quality that enables an entity to function appropriately and with
foresight in its environment.”.
Source: Photo by David Hanson. Wikimedia Commons.
Lets try to understand this by an example , say we ask a human to distinguish between a cat
and a dog, some may say cats are smaller in size compared to dogs, the fur coating is
different between the two, dogs outnumber cats in breed etc etc , but it’s very difficult to
explain these attributes to a machine or for it to differentiate between the two. Hence we feed
multiple images of cats and dogs and label the difference along with the images. Its quite
similar to teaching a child to recognize the difference between the two with a picture . After
few iterations as the child understands and calls it out so is the machine trained now to
recognize a cat and a dog by its picture.
If we think on this example in detail ,AI goes through three processes that includes learning
(the acquisition of information and rules for using the information), reasoning (using rules to
reach approximate or definite conclusions), and self-correction.
AI systems are designed to perform tasks that typically require human intelligence, such as
visual perception, speech recognition, decision-making, and language translation. These
systems are programmed to analyze data, identify patterns, and make decisions with minimal
human intervention.
There are two types of AI:
Narrow AI: ability to accomplish a narrow set of goals, aka “weak AI”. e.g. a model that
excels at classifying cat pictures cannot play online chess, and vice-versa.
Artificial general intelligence (AGI): ability to understand and learn any intellectual task
that a human being can, aka “strong AI”.
Fun Applications of AI:
AI isn't just for serious tasks – it can also be a lot of fun! Have you ever played a video game
where the characters seem to react intelligently to your actions? That's AI at work, creating
immersive and challenging gaming experiences. AI is also behind voice assistants like Siri
and Alexa, which can answer questions, play music, and even tell jokes!
AI in Everyday Life:
Believe it or not, AI is all around us, making our lives easier and more convenient. When you
use social media, AI algorithms help determine which posts you see in your feed based on
your interests and behaviour. In healthcare, AI can assist doctors in diagnosing diseases and
recommending treatment plans. Even when you search the internet, AI algorithms help
deliver relevant results quickly.
Now let’s talk about what happened in Nov 30,2022….
ChatGPT(Chat Generative Pre-trained Transformer)
On Nov 2022 the world was knocked their socks off by the launch of this product . It’s a
form of Generative artificial intelligence (generative AI, GenAI, or GAI).
ChatGPT is an AI-powered conversational agent developed by OpenAI. ChatGPT is designed
to engage in natural language conversations with users, offering assistance, information, or
simply engaging in dialogue on various topics. It's capable of understanding context,
generating coherent responses, and adapting to different conversational styles. This helps
with content creation and information retrieval in form of text, image and voice.
Limitations of AI:
Hallucination. Often large and complex queries as an input from human can generate
plausible sounding yet false answers. It can also provide different “facts” depending on the
language in which the prompt was given.
For [Link] the below example you can see 2 incorrect answers for the same query generated by
ChatGPT
Source: arXiv 2307.10169. Challenges and Applications of Large Language Models.
Up-to-date and specific knowledge. The knowledge of AI model is frozen in time from
when it was trained. In contrast to world general knowledge, business information changes
quickly.
For eg. If you ask them about the result of a cricket match from last night it will not be able to
return you a result as its trained dataset date is frozen and is subjected to patch updates at
intervals.
Being Probabilistic in nature: AI is probabilistic is nature , each and every time it can yield
a new result even though the query is same. Try it 😊
Getting Started with AI: If you're interested in learning more about AI, there are plenty of
resources available to help you get started. You can find books, online courses, and even
coding kits designed specifically for beginners. Experimenting with simple AI projects, like
building a basic chatbot or training a virtual pet, can be a fun and educational way to explore
the possibilities of AI
Conclusion
I will end this article by asking the readers this question. This picture has gone viral over the
internet but its not what you think.
What did you see? And how certain are you on a scale of 1-10?
Concert... or cotton harvest?
Source: BBC. 9. November 2017.
As amazing as human intelligence is, people are far from 100% accurate.
A computer can learn to execute tasks... often faster and more accurately than people, but
usually not with 100% accuracy.
Humans + AI working together.
Signing Out for this year
Swarnali Basu
.
Disclaimer: This article featured in this magazine is based on experiences and insights gathered from various internet sources and work experience. While efforts have been
made to ensure the accuracy and reliability of the information presented, it is important to note that the perspectives shared in this article are those of the author alone.
Readers are encouraged to verify the information and consult with relevant professionals before applying any advice or recommendations provided herein to their own lives.
The magazine and its publishers do not endorse any specific views or opinions expressed in this article, and shall not be held liable for any consequences resulting from
interpretation and implementation
