Week 1: Introduction and Capabilities of Generative AI

[MUSIC] Let’s get started with
an introduction to generative AI. Imagine a world in which AI is going
to make us work more productively, live longer, and have cleaner energy. Well, that world is already here. Generative AI has made a big impact
on the way we go through life. Generative AI models can mimic human
thinking and creativity to create novel content and perform complex tasks,
just like you and me. Organizations can leverage generative AI
to become more productive and profitable. Individuals can use generative AI
tools to increase their efficiency, add tangible value to their work, save
money, and maximize their brand value. If you’re not there yet, this course
is for you, inviting all professionals, enthusiasts, practitioners and students with a genuine interest in the
rapidly developing field of generative AI. Regardless of your background or
experience, this is a course for everyone. This course is designed to give you
a solid understanding of the capabilities, applications, and common models and
tools of generative AI. By the end of this course, you’ll be able
to describe the capabilities of generative AI and its use cases in the real world. Identify the applications of generative
AI in different sectors and industries, and explore common generative
AI models and tools. As this is a focused course
comprising three modules, you’re expected to spend 1-2 hours
to complete each module. In Module 1 of the course, you’ll learn
about the core concepts of generative AI, view use cases of how it is applied
across different domains, and understand its capabilities for generating
text, images, code, audio and video. In Module 2,
you’ll discover how different sectors and industries such as information technology,
entertainment, education, finance and healthcare leverage generative AI. Further, in this module,
you’ll learn about the capabilities and features of common models and tools for
generating text, images, code, audio, and video such as ChatGPT,
DALL-E, and Synthesia. Module 3 requests your
participation in a final project and presents a graded quiz to test your
understanding of course concepts. You can also visit the course glossary and receive guidance on the next
steps in your learning journey. The course is curated with a mix of
concept videos and supporting readings. Watch all the videos to capture the full
potential of the learning material. You’ll enjoy hands-on labs and a final project
that demonstrates the common use cases of generative AI across multiple domains. There are practice quizzes at the end
of each lesson to help you reinforce your learning. At the end of the course,
you’ll also attempt a graded quiz. The course also offers discussion forums
to connect with the course staff and interact with your peers. Most interestingly,
through the Expert Viewpoint videos, you’ll hear from experienced practitioners
who will share their perspectives on different aspects of generative AI. Well, don’t just stand by. As generative AI strengthens the creative
expression and professional capabilities of individuals, organizations,
and communities across the world, this course presents an opportunity for
you to create new experiences. [MUSIC]

Generative AI is on the minds of
every leader in every organization, businesses, or governments, and
with interest comes opportunity. Organizations are looking for
people who understand the technology and, most importantly,
have the skills to apply it. Unlike many of the previous
trending technologies, generative AI touches almost
every role and every profession. Generative AI skills are expected
to become important not just for computer scientists, for everybody,
which is why they will be essential as Word processing, spreadsheets,
even basic business literacy. This is why these programs
are generative AI for everyone. There’s a lot of new interest
happening right now in AI, and businesses are looking beyond consumer AI. A chatbot interface is a great way to
demonstrate generative AI potential. Real-life use cases are embedding
generative AI into existing processes and making it an integral function of
nearly every business workflow. IBM is proud to be helping businesses
integrate generative AI into their operations. These are skills you will gain
as part of these programs, that should help you with your career and
be very applicable to your job instantly. Businesses are excited by
the potential of generative AI, but they’re also apprehensive
of potential dangers. >> This mission is too important for
me to allow you to jeopardize it. >> These programs will give you the skills
to deal with AI ethics that are grounded in a responsible approach,
really started at IBM.

Did you know that marketers
across the world are already using generative
AI to create content, write copy, inspire
creative thinking, analyze market data,
and generate images? According to Bloomberg, the
generative AI market is estimated to reach 1.3
trillion dollar by 2032. In that case you
definitely want to get better acquainted
with generative AI. However is generative
AI for everyone? Yeah, it is, and
you can leverage its potential to create a better career and
life for yourself. This specialization
is for anyone who’s passionate about discovering
the power of generative AI. Prior technical knowledge or background in AI
is not required. Even beginners can benefit
from this specialization as it provides a comprehensive
understanding of the fundamental concepts, models, tools, and
applications of generative AI. By the end of this
specialization, you’ll be able to, explain the
fundamental concepts, capabilities, models,
tools, applications, and platforms of generative
AI foundation models. Describe prompt
engineering and apply powerful prompt engineering
techniques to write effective prompts and generate desired outcomes from AI models. Discuss the limitations of
generative AI and explain the ethical concerns and considerations for the
responsible use of generative AI, and recognize the ability
of generative AI to enhance your career and help implement improvements
at your workplace. Spread across five focused
self paced courses, each requiring three to
five hours to complete. Course one is your
first step toward understanding the capabilities
of generative AI, which span different
domains such as text, image, audio, video, virtual
world, code, and data. You’ll understand how different sectors and industries apply common generative AI models
and tools such as GPT, DALL-E, Stable Diffusion, IBM Granite, and Synthesia. Course two introduces
the concept of prompt engineering
and how it can help you unlock the full potential of generative AI tools
like ChatGPT. You’ll explore the
techniques, approaches, and best practices for
developing effective prompts and work with commonly used
tools such as IBM watsonx, Prompt Lab, Spellbook, and Dust. Course three focuses on the core concepts and building
blocks of generative AI, such as deep
learning, transformer based large language models, diffusion models, and
foundation models. You’ll also learn about the different
generative AI platforms like IBM watsonx.ai
and Hugging Face. In course four, you’ll explore ethical considerations
related to generative AI. How does it impact data
privacy and security, copyright infringement, the workforce, and
the environment? You’ll also describe
limitations such as data bias, lack of explain
ability, transparency, and interpretability
and identify common misuses of generative AI, such as deep fakes
and hallucinations. Lastly, course five discusses the future of generative AI. Don’t you want to know what your career opportunities
are in that future? You’ll learn how
generative AI can impact and enhance existing
functions, skills, and job roles in different
sectors and industries, and how you can use
generative AI to build your own applications to create new business
opportunities. The content in this
specialization is designed to engage
and empower you. By viewing curated
concept videos, hearing AI experts share
their insights and tips, and practicing techniques and hands on labs and projects, you’ll feel more
confident in using generative AI tools and
applications in your daily life. Currently, 65% of
generative AI users are millennials or Gen Z, and 72% are employed. With this specialization, you’ll be ready to join the
generative AI change makers. Generative AI is for everyone.

Welcome to Introduction
to Generative AI. After watching this video, you’ll be able to
describe generative AI and its evolution. You’ll also be able
to explain how generative AI differs
from discriminative AI. Artificial intelligence, or
AI, has been around for years, shaping almost every sphere of our lives and revolutionizing
how we live and work. At its core, AI
can be defined as the simulation of human
intelligence by machines. AI models learn from vast
amounts of existing data. The process of learning from
data is called training. There are two fundamental
approaches to AI, discriminative AI,
and generative AI. Discriminative AI
is an approach that learns to distinguish between
different classes of data. A discriminative AI
model is given a set of training data where
each data point is labeled with its class. The model then predicts the
class of a new data point by finding the side of
the decision boundary that the data point falls on. Discriminative AI models use advanced algorithms
to differentiate, classify, identify patterns, and draw conclusions
based on training data. An example of how a
discriminative AI model works can be seen in how email spam
filters can differentiate between spam and non-spam
emails. Discriminative AI models are best applied to
classification tasks. They cannot, however,
understand context or generate new content based on a contextual understanding
of the training data. And this is where generative
artificial intelligence, or generative AI comes in. Generative AI models learn to generate new content based
on the training data. They can capture the
underlying distribution of the training data and generate
novel data instances. Generative AI starts
with a prompt. This can be text, an image, video, or any other input
that the model can process. As an output, the model generates new content
including text, images, audio, video
code, and data. Generative AI can produce output in the same form in which
the prompt is provided. For example, text to text, or in a different
form from the prompt, such as text to image
or image to video. Here is a simple example to understand the
difference between discriminative or traditional
AI and generative AI. Discriminative AI would be best suited to answer
questions such as, is this image a drawing
of a nest or an egg? Generative AI would respond
to prompts such as, draw an image of a nest
with three eggs in it. While discriminative AI mimics our analytical and
predictive skills, generative AI goes a step further to mimic our
creative skills. As implied by this comment from the Harvard
Business Review, AI can not only boost our analytic and
decision-making abilities, but also heighten creativity. Generative models can take
what they have learned and create entirely new content
based on that information. Both discriminative
and generative models are created using deep
learning techniques. Deep learning involves training artificial neural networks to learn from vast amounts of data. An artificial neural network is a collection of smaller
computing units called neurons, which are modeled in a
manner that is similar to how a human brain
processes information. The creative skills of generative AI come from
generative AI models, such as generative
adversarial networks or GANs, variational
autoencoders or VAEs, transformers, and
diffusion models. These models can
be considered as the building blocks
of generative AI. Generative AI is
not a new concept. Its roots trace back to the
origins of machine learning. In the late 1950s, when scientists proposed
machine learning, they explored using algorithms
to create new data. During the 1990s, the rise of neural networks infused
advancements in generative AI. Further, during the early
2010s, deep learning, supported by the availability of large data sets and
enhanced computing power, further advanced the
development of generative AI. In 2014, generative AI
was transformed with the introduction of GANs by Ian Goodfellow and
his colleagues. GANs and other models such as VAEs and transformers
set the stage for generative AI’s growth and the development of
foundational models and tools. Foundation models
are AI models with broad capabilities
that can be adapted to create more
specialized models or tools for specific use cases. A specific category of foundation models called
large language models, or LLMs, are trained to understand human language and can process and generate text. In 2018, OpenAI introduced a transformer-based LLM called generative pretrained
transformer, or GPT. Over the years,
different LLMs such as GPT-3 and GPT-4 in
the GPT series, Google’s pathways
language model or PaLM, and Meta’s large language model, Meta AI or Llama have
significantly enhanced generative AI to generate
coherent and relevant text. There have been
similar developments and models for other use cases. For example, Stable
Diffusion and DALL-E are models for
image generation. The development of a variety of generative models has led to a growing market for
generative AI tools for diverse use cases. For instance, you have ChatGPT and Bard for text generation, DALL-E 2 and Midjourney
for image generation, Synthesia for video generation, and Copilot and AlphaCode
for code generation. The rapidly emerging models
and tools have generated a wide scope for generative AI applications across domains. To quote from
McKenzie’s report on the economic potential
of generative AI, “Generative AI has the potential to change
the anatomy of work, augmenting the capabilities
of individual workers by automating some of their
individual activities.” The report also predicts that
generative AI’s impact on productivity could add
trillions of dollars in value to the global economy. In this video, you learned
that generative AI models can generate new content based on the data
they are trained on. Further, you learned that
the creative skills of generative AI are built
from models such as GANs, VAEs, transformers,
and diffusion models. Foundation models can
be adapted to create specialized models or tools tailored to specific use cases. Finally, you learned that generative AI models
and tools have a wide scope for applications across different
domains and industries.

The Capabilities of Generative AI: A Hands-on Tutorial

Have you ever dreamed of conversing with a virtual bard, creating captivating artwork instantly, or composing your own masterpiece of music? Generative AI is making these possibilities a reality, and this tutorial will guide you through its fascinating capabilities.

Step 1: Demystifying Generative AI

Generative AI models leverage sophisticated algorithms to learn from vast amounts of data, enabling them to create entirely new content. Think of it as a skilled artist, trained on years of observing the world, now able to paint original landscapes or sculpt compelling characters.

Step 2: Exploring Diverse Capabilities:

• Text Generation: Imagine chatting with a witty AI companion, or having your next blog post automatically written. Large Language Models (LLMs) like GPT-3 and PaLM can generate human-quality text, translate languages, and even write different kinds of creative content.

Let’s try it! Play with online platforms like Google AI’s “Bard” or OpenAI’s “Jukebox” to test text generation, and see how LLMs can mimic different writing styles or answer your questions creatively.

• Image Generation: Forget expensive photo shoots! Tools like DALL-E 2 and StyleGAN can conjure up realistic images based on your descriptions or even combine existing ones into surreal new creations.

Challenge yourself! Use online demo platforms like NightCafe Creator or Dream by WOMBO to describe your dream vacation destination or a mythical creature, and see what AI paints for you!

• Audio Generation: Want to compose your own soundtrack or hear a poem recited in a unique voice? Generative models like WaveGAN and MuseNet can create music, convert text to speech, and even modify or enhance existing audio.

Experiment! Explore tools like Jukebox or Google’s Tacotron 2 to generate music in different styles or hear your favorite poem read aloud with different emotional tones.

Step 3: Beyond the Playground:

The applications of Generative AI extend far beyond mere entertainment.

• Revolutionizing Industries: In medicine, AI can create realistic simulations for surgeon training or generate synthetic data for drug discovery. In design, AI can suggest innovative product prototypes or personalize user experiences.
• Boosting Productivity: Software developers can use AI to autocomplete code or generate boilerplate text, while writers can benefit from AI-powered brainstorming tools.
• Unlocking Creativity: Artists can collaborate with AI to explore new styles or generate variations on their existing work, while musicians can use AI to create unique backing tracks or remixes.

Step 4: The Future of Creativity:

Generative AI is still in its early stages, but its potential is boundless. As technology advances, we can expect even more sophisticated models and integrations into our daily lives.

Get involved! Stay updated on the latest advancements in Generative AI and explore how it can empower you to become a co-creator, not just a consumer, of the world around you.

Remember, Generative AI is a powerful tool, and with responsible use, it can unlock unimaginable possibilities for creativity, innovation, and progress. So, let’s dive into this exciting world together and explore the future of AI-powered creation!

Bonus Resources:

• OpenAI: https://openai.com/
• Google AI: https://ai.google/
• DeepMind: https://deepmind.google/
• Papers with Code: https://arxiv.org/list/cs.CL/recent

Start your Generative AI journey today, and who knows, maybe you’ll be the next AI-powered artist or inventor!

[MUSIC] Welcome to The Capabilities
of Generative AI. After watching this video,
you’ll be able to describe some of the capabilities of generative AI
and explore their use in the real world. Let’s start with a high level overview of
some of the capabilities of generative AI that we’ll discuss. First is the text generation
capability of generative AI, that is, its ability to generate clear lucid and
contextually relevant textual responses. The second capability is image generation,
that is, synthesizing artistic and realistic images
that are very similar to real ones. The third capability is audio generation. Generative AI enables music composition
and synthetic audio generation. The fourth capability we’ll
discuss is video generation. Generative AI enables
the generation of dynamic films and small videos based on textual
descriptions and even images. The fifth capability is the code
generation capability of generative AI. Generative models can generate
code functions and programs. We’ll also discuss the data generation and
augmentation capability of generative AI. This helps generate synthetic data
to create and augment datasets. Finally, we’ll explore generative
AI’s capability to create real and immersive virtual worlds. These are just some of
the capabilities of generative AI. Essentially, whatever the human mind is
capable of conceiving is a potential use case for the application of generative AI. Now, let’s delve deeper into
some of these capabilities. Let’s begin with the text generation
capabilities of generative AI. At the core of generative AI’s text
generation capability are advanced AI-powered large language models or LLMs. LLMs are trained on large datasets and can generate human like
text in various contexts. These models learn patterns and structures
within the data to generate coherent and contextually relevant responses. These models generate text, converse and
provide explanations, summaries, and more. Some popular LLMs are Open AI’s
generative pre-trained transformer or GPT and Google’s pathways language model,
or PaLM. These models can perform various language
related tasks such as text completion, summarization, question answering,
translation, code generation, and image and text pairing. Conversational interactions
with chatbots and virtual assistants are powered by LLMs. Let’s look at the image generation
capabilities of generative AI. Generative models can generate high
quality, convincing images based on deep learning techniques such as generative
adversarial networks or GANs and variational autoencoders or VAEs. These generated images exhibit
realistic textures, natural colors, and fine grained details, giving
the impression of a real camera capture. StyleGAN, for example,
can generate high quality, high resolution new images of
imaginary faces, animals, or nature. While DeepArt can create comprehensive
artwork from a simple sketch. DALL-E can generate entirely new
images as described by the users. Apart from applications in art,
design, entertainment, gaming, and research domains, generated images can
augment training data and aid medical, imaging and scientific visualization. In the context of audio generation,
generative models can generate new musical compositions, convert text into
audio using text to speech or TTS, and create synthetic voices and
natural sounding speech. Generative models can convert,
modify, and transform and clean up voices, and also reduce noise and
enhance audio quality. These models also have the capability to
mimic a human voice to a fair amount of likeness. WaveGAN, for example, can create new and
realistic raw audio waveforms, including speech, music,
and natural sounds. MuseNet from OpenAI can combine
various instruments, styles, and genres to generate novel
musical compositions. Google’s Tacotron 2 and Mozilla TTS use
advanced TTS systems to create synthetic speech resembling human tone, pitch,
modulation, pronunciation, rhythm, and expressions. Audio generated by generative models
has applications in media, creativity, entertainment, training, education,
gaming, virtual reality, and several other domains. Now, let’s look at the video generation
capabilities of generative AI. Generative AI models
can create dynamic and lucid videos ranging from basic
animations to complex scenes. These models transform images into
dynamic videos by incorporating temporal coherence. In natural language processing, temporal
coherence refers to the consistency and continuity of meaning or
context over time. This enables these models to
exhibit smooth motion and plausible transitions in videos. For instance,
a popular AI model, Video GPT, follows textual prompts users
provide to generate new videos. Users can specify the desired content and
guide the video generation process, including completion, editing, synthesis,
prediction, and style transfer. These generated videos can be used in
domains such as art, entertainment, education, gaming, medicine, and research. Now, let’s talk about generative
AI’s code generation capabilities. Generative models can generate
new code snippets functions or complete programs based
on desired functionality. Trained on existing code repositories,
these models can complete or create code, synthesize or refractor code, identify and
fix bugs in code, test software, and create documentation including
comments, function descriptions, and usage examples. For instance, GitHub Copilot and
IBM Watson Code Assistant are AI-based programming assistants that help
autocomplete code, accelerate hard tasks, and generate code for provided input. AI generated code can be used
in software and web development, machine learning and
natural language processing, data science and analytics, robotics and
automation, virtual game and AR/VR environment development, and audio,
video, and speech processing. Software developers can benefit from
leveraging code generation capabilities to write, debug, and test their code. Now, let’s explore the data generation and augmentation capabilities
of generative AI. Generative models can generate new
data and augment existing datasets. Generating synthetic datasets
helps increase the diversity and variability of the data, leading to
more robust and effective performance. These models can generate new samples and
augment data sets for images, text, speech tabular, data and statistical
distribution, time series, data finance, and more. The data generation and augmentation capabilities of generative AI
have applications in medicine, healthcare, gaming, education and
training, art and creativity. Self-driving automobiles, and many more. Another powerful capability of
generative AI models is their ability to create highly realistic and
complex virtual worlds. You can create avatars that simulate
realistic behavior, expressions, conversations, and even decisions. You can also create complex virtual
environments with realistic textures, sounds, and objects that follow
the principles of the physical world, metaverse platforms use generative
models to create unique and personalized experiences for
individual users. Generative AI also makes it
possible to create virtual identities with unique personalities
avatars that can be fitted with specific personality traits that reflect in
their behaviors and conversations. The virtual world capability of
generative AI has applications in gaming, entertainment, education,
augmented and virtual reality. Metaverse platforms, and also virtual
influencers and digital personalities. In this video, you learned about some
of the capabilities of generative AI models and their use in the real world. Generative AI can create coherent and
contextually relevant content and generate realistic, high quality images, synthetic
voices, new audio and dynamic videos. And generative AI models can generate and
complete code and synthesize new data to augment
the existing data sets. Generative AI models are also capable
of creating highly realistic and complex virtual worlds, including virtual
avatars and digital personalities. [MUSIC]