This book has sample code for a basic Python application. To recreate it, it is recommended that you have the following:

• The latest version of Python
• A local development environment on your device for your application server
• A MongoDB Atlas cloud account to host your database. You can register for one at https://www.mongodb.com/cloud/atlas/register
• VS Code or an IDE of your choice
• An OpenAI API key

For the true beginner, let’s start with defining some key terms: AI, ML, and GenAI. You will come across these terms repeatedly in this book, so it helps to have a strong conceptual foundation of these terms:

• Artificial intelligence (AI) refers to the ability of machines to perform tasks that would normally require human intelligence. This includes tasks such as perception, reasoning, learning, and decision making. The journey of AI has evolved significantly from early speculative ideas to the sophisticated technologies of today. Figure 1.1 shows a timeline of the development of AI.

Figure 1.1: A timeline of AI

• Machine learning (ML) is a subset of AI that involves the use of algorithms to automatically learn from data and improve over time. Essentially, it’s a way for machines to learn and adapt without being explicitly programmed. Most often used in fields that require advanced analysis of thousands of data points, ML is most useful in medical diagnostics, market analysis, and military intelligence. Effectively, ML identifies hidden or complex patterns in data that would be impossible for a human to see and then can make suggestions for the next steps or actions.
• Generative AI (GenAI) is the ability to create text, images, audio, video, and other content in response to a user prompt. It powers chatbots, virtual assistants, language translators, and other similar services. These systems use algorithms trained on vast amounts of data, such as text and images from the internet, to learn patterns and relationships. This enables them to generate new content that is similar but not identical to the underlying training data. For instance, large language models (LLMs) use training data to learn patterns in written language. GenAI can then use these models to emulate a human writing style.

A stack combines tools, libraries, software, and solutions to create a unified and integrated approach. The GenAI stack includes programming languages, LLM providers, frameworks, databases, and deployment solutions. Though the GenAI stack is relatively new, it already has many variations and options for engineers to choose from.

Let’s discuss what you need to build a functional GenAI application. The bare minimum requirements are the following, as also shown in Figure 1.2:

• An operating system: Usually, this is Unix/Linux based.
• A storage layer: An SQL or NoSQL database. This book uses MongoDB.
• A vector database capable of storing embeddings: This book uses MongoDB, which stores its embeddings within your data or content, rather than in a separate database.
• A web server: Apache and Nginx are quite popular.
• A development environment: This could be Node.js/JavaScript, .NET, Java, or Python. This book uses Python throughout the examples with a bit of JavaScript where needed.

Figure 1.2: A basic GenAI stack

If you want to learn more about the AI stack, you can find detailed information at www.mongodb.com/resources/basics/ai-stack.

Python and GenAI

Python was conceived in the late 1980s by Guido van Rossum and officially released in 1991. Over the decades, Python has evolved into a versatile language, beloved by developers for its clean syntax and robust functionality. It has a clean syntax that is easy to understand, making it an ideal choice for beginner developers.

Although it is not entirely clear why, fairly early on, the Python ecosystem began introducing more libraries and frameworks that were tailored to ML and data science. Libraries and frameworks such as TensorFlow, Keras, PyTorch, and scikit-learn provided powerful tools for developers in these fields. Analysts who were less technical were still able to get started with Python with relative ease. Due to its interoperability, Python seamlessly integrated with other programming languages and technologies, making it easier to integrate with data pipelines and web applications.

GenAI, with its demands for high computational power and sophisticated algorithms, finds a perfect partner in Python. Here are some examples that readily come to mind:

• Libraries such as Pandas and NumPy allow efficient manipulation and analysis of large datasets, a fundamental step in training generative models
• Frameworks such as TensorFlow and PyTorch offer pre-built components to design and train complex neural networks
• Tools such as Matplotlib and Seaborn enable detailed visualization of data and model outputs, aiding in understanding and refining AI models
• Frameworks such as Flask and FastAPI make deploying your GenAI models as scalable web services straightforward

Python has a rich ecosystem that is easy to use and allows you to quickly get started, making it an ideal programming language for GenAI projects. Now, let’s talk more about the other pieces of technology you’ll be using throughout the rest of the book.

OpenAI API

The first, and most important, tool of this book is the OpenAI API. In the following chapters, you’ll learn more about each component of the GenAI stack—and the most critical to be familiar with is OpenAI. While we’ll cover other LLM providers, the one used in our examples and code repository will be OpenAI.

The OpenAI API, launched in mid-2020, provides developers with access to their powerful models, allowing integration of advanced NLP capabilities into applications. Through this API, developers gain access to some of the most advanced AI models in existence, such as GPT-4. These models are trained on vast datasets and possess unparalleled capabilities in natural language understanding and response generation.

Moreover, OpenAI’s infrastructure is built to scale. As your project grows and demands more computational power, OpenAI ensures that you can scale effortlessly without worrying about the underlying hardware or system architecture. OpenAI’s models excel at NLP tasks, including text generation, summarization, translation, and sentiment analysis. This can be invaluable for creating content, chatbots, virtual assistants, and more.

Much of the data from the internet and internal conversations and documentation is unstructured. OpenAI, as a company, has used that data to train an LLM, and then offered that LLM as a service, making it possible for you to create interactive GenAI applications without hosting or training your own LLM. You’ll learn more about LLMs in Chapter 3, Large Language Models.

MongoDB with Vector Search

Much has been said about how MongoDB serves the use case of unstructured data but that the world’s data is fundamentally relational. It can be argued that no data is meaningful until humans deem it so, and that the relationships and structure of that data are determined by humans as well. For example, several years ago, a researcher at a leading space exploration company made this memorable comment in a meeting:

“We scraped text content from websites and PDF documents primarily, and we realized it didn’t really make sense to try and cram that data into a table.”

MongoDB thrives with the messy, unstructured content that characterizes the real world—.txt files, Markdown, PDFs, HTML, and so on. MongoDB is flexible enough to have the structure that engineers deem is best suited for purpose, and because of that flexibility, it is a great fit for GenAI use cases.

For that reason, it is much easier to use a document database for GenAI than it is to use a SQL database.

Another reason to use MongoDB is for its vector search capabilities. Vector search ensures that when you store a phrase in MongoDB, it converts that data into an array. This is called a vector. Vectors are numerical representations of data and their context, as shown in Figure 1.3. The number of these dimensions is referred to as an embedding, and the more of them you have, the better off you are.

Figure 1.3: Example of a vector

After you’ve created embeddings for a piece of data, a mathematical process will identify which vectors are closest or nearest to each other, and you can then infer that the data is related. This allows you to return related words instead of only exact matches. For instance, if you are looking for pets, you could find cats, dogs, parakeets, and hamsters—even though those terms are not the exact word pets. Vectors are what allow you to receive results that are related in meaning or context or are alike, without being an exact match.

MongoDB stores your data embeddings alongside the data itself. Storing the embeddings together makes the consequent queries faster. It is easiest to visualize vector search via an example with explanations of how it works along the way. You will learn more about vector search in Chapter 8, Implementing Vector Search in AI Applications.

When asked to list the most important capability of GenAI applications, ChatGPT, which is arguably the most popular GenAI application in existence, said the following:

Content Creation: Generative AI can craft text, images, music, and even videos. It can pen articles, generate realistic images from textual descriptions, compose music, and create video content, opening endless possibilities for creative industries.

That response took 1.5 seconds to generate, and most people would agree with it. GenAI applications can create content for you and your users with lightning speed. Whether it’s text, video, images, artwork, or even Java code, GenAI is able to easily draft foundational content that can then be edited by professionals.

But there are other key features of GenAI applications that merit calling out as well:

• Language translation: With remarkable proficiency, GenAI can translate languages in real time, preserving context and nuance, and facilitating seamless communication across linguistic barriers.
• Personalization: In the realm of marketing and customer service, GenAI can tailor experiences and content to individual users. When given proper context, it can analyze preferences and behaviors to deliver personalized recommendations, emails, and customer interactions.
• Simulation and modeling: In scientific research and engineering, GenAI can simulate complex systems and phenomena. It aids in predicting molecular behaviors, climate patterns, and even economic trends by generating realistic models based on vast datasets.
• Data augmentation: For ML, GenAI can produce synthetic data to augment training sets. This is invaluable in scenarios where real data is scarce or biased, allowing for the creation of diverse and balanced datasets to improve model performance. This is incredibly useful for testing purposes, particularly in software testing.

And perhaps most importantly, it can accept prompting in natural language (such as in English) to do these tasks. This makes performing tasks you previously found difficult incredibly easy. You may use GenAI to accomplish multiple and varied tasks in a day, such as reviewing a pull request, guiding you through some tasks for Golang, and generating illustrations for the interior artwork of a book.

Why use generative AI?

Each of the preceding abilities is compelling and important, and when used correctly and in combination, revolutionary. Put simply, there is no industry where GenAI cannot play a role. By rapidly aggregating and summarizing a wide range of content and simplifying searching, GenAI improves the user experience of finding ideas and building knowledge. It can help gather new information, summarize it, and recraft it into content. It can help speed up or even automate administrative tasks, and exponentially increase output.

But beyond all of that, the experience of using GenAI is an order of magnitude better than what is available today. Consider, for example, a customer service bot. Many of you will be familiar with this flow:

1. The customer first encounters a long menu of options: If you want to talk to sales or support, press 1. For billing, press 2. For administration, press 3. For orders, press 4.. When the customer has a question that does not neatly fit into any category, they may press 4 anyway.
2. Upon pressing 4, they are then routed to a support page that does not have the answer they seek. They click a button that says, No, this did not answer my question.
3. They search the knowledge base themselves, perhaps never finding the answer and reaching out via phone.

Imagine being able to type what you wanted and the bot responding in a natural way—not routing you to a page but just giving you the answer. Imagine even further that the user can then chat with the bot to say they want to modify the address on their order, and the bot is able to do that from within the chat window, having a multi-step dialogue with the user to confirm and record their new information.

It is a wholly new, more pleasing experience for the customer!

The ethics and risks of GenAI

Despite those benefits, there are risks and concerns about the use of AI. In some fields, the outcry against AI is substantial and has merit. Art generated by AI, for example, flooded the internet’s marketplaces, displacing artists and illustrators who make their living off their craft. There are questions about whether using AI to write a book gives a person the right to call themselves an author. There are no clear-cut answers here; from our own experience, the authors of this book believe that GenAI accelerates, rather than replaces, the existing paradigms of work done today. But that may not always remain true. As AI improves, it may be more likely to replace the humans who are using it.

The risks of GenAI are considerable, and some of them are not well understood. Even the ones that are well understood, such as hallucinations, are difficult to identify for users, and harder still to combat. You can read more about the challenges of GenAI in Chapter 11, Common Failures of Generative AI, along with recommendations on how to mitigate them in Chapter 12, Correcting and Optimizing Your Generative AI Application.

This chapter laid the background for the GenAI application, from describing the role of each component to their strengths. You learned some key definitions and were introduced to the basics of the AI stack. By now, you also know why Python is a great choice for building GenAI applications and why you will be using the OpenAI API and MongoDB with Vector Search to build your GenAI application. Finally, you also saw some significant use cases for GenAI and learned why you should use GenAI in the first place, while also remaining mindful of the ethics and risks of using it. Since you’re reading this, I’ll assume that the case was compelling—that you’re still interested and ready to explore.

In the next chapter, you will get a fast-paced, concise, and actionable overview of the building blocks of GenAI applications in more detail and learn how to get started.