Join our book community on Discord

https://packt.link/EarlyAccess/

Laying a solid foundation is an essential part of understanding anything, and the frontier of artificial intelligence (AI) products seems a lot like our universe: ever-expanding. That rate of expansion is increasing with every passing year as we go deeper into a new way to conceptualize products, organizations, and the industries we’re all a part of. Virtually every aspect of our lives will be impacted in some way by AI and we hope those reading will come out of this experience more confident about what AI adoption will look like for the products they support or hope to build someday.

Part 1 of this book will serve as an overview of the lay of the land. We will cover terms, infrastructure, types of AI algorithms, and products done well, and by the end of this section, you will understand the various considerations when attempting to build an AI strategy, whether you’re looking to create a native-AI product...

Definitions – what is and is not AI

In 1950, a mathematician and world war II war hero Alan Turing asked a simple question in his paper Computing Machinery and Intelligence – Can machines think?. Today, we’re still grappling with that same question. Depending on who you ask, AI can be many things. Many maps exist out there on the internet, from expert systems used in healthcare and finance to facial recognition to natural language processing to regression models. As we continue with this chapter, we will cover many of the facets of AI that apply to products emerging in the market.

For the purposes of applied AI in products across industries, in this book, we will focus primarily on ML and deep learning (DL) models used in various capacities because these are often used in production anywhere AI is referenced in any marketing capacity. We will use AI/ML as a blanket term covering a span of ML applications and we will cover the major areas most people would consider...

ML versus DL – understanding the difference

As a product manager, you’re going to need to build a lot of trust with your technical counterparts so that, together, you can build an amazing product that works as well as it can technically. If you’re reading this book, you’ve likely come across the phrase ML and DL. We will use the following sections titled ML and DL to go over some of the basics but keep in mind that we will be elaborating on these concepts further down in Chapter 3.

Learning types in ML

In this section, we will cover the differences between supervised, unsupervised, semi-supervised, and reinforcement learning and how all these learning types can be applied. Again, the learning type has to do with whether or not you’re labeling the data and the method you’re using to reward the models you’ve used for good performance. The ultimate objective is to understand what kind of learning model gets you the kind of performance and explainability you’re going to need when considering whether or not to use it in your product.

LLMs, NLP, GANs and how they relate to machine learning and generative AI

Just as “AI” is an umbrella term, “generative AI” follows suit. As you might have inferred from the its naming, generative AI is an area of AI that’s all about generating new content whether that’s text, an image or even code. The machine learning models that power generative AI are creating outputs that closely resemble the training data they learn from.

When you think of generative AI, I want you to primarily think of advanced deep learning models that are grouped into two categories: GANs and large language models (LLMs). GANs are adversarial neural networks that are primarily used for image generation. You can think of them as two neural networks fighting each other for the “best” or “most correct” image. LLMs are neural networks that have been trained on terabytes of data. So much data, in fact, that it’s almost prohibitively expensive...

The order – what is the optimal flow and where does every part of the process live?

Companies interested in creating value with AI/ML have a lot to gain compared to their more hesitant competitors. According to McKinsey Global Institute, “Companies that fully absorb AI in their value-producing workflows by 2025 will dominate the 2030 world economy with +120% cash flow growth.” The undertaking of embracing AI and productionizing it – whether in your product or for internal purposes – is complex, technical debt-heavy, and expensive. Once your models and use cases are chosen, making that happen in production becomes a difficult program to manage and this is a process many companies will struggle with as we see companies in industries other than tech starting to take on the challenge of embracing AI. Operationalizing the process, updating the models, keeping the data fresh and clean, and organizing experiments, as well as validating, testing, and the storage...

Managing projects – IaaS

If you’re looking to create an AI/ML system in your organization, you’ll have to think about it as its own ecosystem that you’ll need to constantly maintain. This is why you see MLOps and AIOps working in conjunction with DevOps teams. Increasingly so, we will start to see managed services and infrastructure-as-a-service (IaaS) offerings coming out more and more. There has been a shift in the industry toward companies such as Determined AI and Google’s AI platform pipeline tools to meet the needs of the market. At the heart of this need is the desire to ease some of the burdens from companies left scratching their heads as they begin to take on the mammoth task of getting started with an AI system.

Just as DevOps teams became popular with at-scale software development, the result of decades of mistakes, we will see something similar with MLOps and AIOps. Developing a solution and putting it into operation are two different key...

Deployment strategies – what do we do with these outputs?

Once you’re happy with the models you’ve chosen (including their performance and error rate), you’ve got a good level of infrastructure to support your product and chosen AI model’s use case; you’re ready to go to the last step of the process and deploy this code into production. Keeping up with a deployment strategy that works for your product and organization will be part of the continuous maintenance we’ve outlined in the previous section. You’ll need to think about things such as how often you’ll need to retrain your models and refresh your training data to prevent model decay and data drift. You’ll also need a system for continuously monitoring your model’s performance so this process will be really specific to your product and business, particularly because these periods of retraining will require some downtime for your system.

Deployment is going...

Succeeding in AI – how well-managed AI companies do infrastructure right

It’s indicative of the complexity of ML systems that many large technology companies that depend heavily on ML have dedicated teams and platforms that focus on building, training, deploying, and maintaining ML models. The following are a few examples of options you can take when building an ML/AI program:

• Databricks has MLflow: MLflow is an open source platform developed by Databricks to help manage the complete ML life cycle for enterprises. It allows you to run experiences and work with any library, framework, or language. The main benefits are experiment tracking (so you can see how your models are doing between experiments), model management (to manage all versions of your model between teammates), and model deployment (to have a quick view of deployment in view in the tool).
• Google has TensorFlow Extended (TFX): This is Google’s newest product built on TensorFlow and it’s an end...