What is generative AI? Will it pave the way for AIOps?

Digital face on digital background.

Generative AI uses machine learning to generate multiple forms of content, such as text, images, or audio, based on natural language cues. Its widespread adoption began with the emergence of GPT, an OpenAI initiative, which has inspired a plethora of innovative applications across industries.

If you haven't explored GPT for yourself, I suggest you should ask it some questions. For example, what it knows about you or a celebrity, or whether it can explain how something works. While be careful about the reliability of the information when asking GPT questions – as it is not always correct – it was an eye-opening experience because it was a new one.

What the platform does is provide a proof of concept for generative AI. It gives us an idea of the possibilities of AI and how it can bring a new paradigm to the way we work. For us at the Office of the Chief Technology Officer at F5, some interesting explorations into the potential of the technology for application delivery and security could lead to the rise of AIOps.

The shift from imperative to declarative to generative

One of the challenges of IT infrastructure is configuring the large number of devices, services, and systems required to deliver and secure a single application. Without counting as-a-service offerings, enterprises typically rely on an average of 23 different application services.

I don't have to tell you that configuring web applications and API protection services is different from configuring plain old load balancing services. This means that the individual responsible for configuring and operating the application service may have to be an expert in more than a dozen different languages.

The industry has struggled to meet this challenge for years. While APIs become the primary way to configure everything, application delivery and security services are no exception. Everyone relies on imperative APIs, which essentially changes the way commands are issued. For example, instead of typing commands on the CLI, you send API commands over HTTP.

Digital face on digital background.

Generative AI uses machine learning to generate multiple forms of content, such as text, images, or audio, based on natural language cues. Its widespread adoption began with the emergence of GPT, an OpenAI initiative, which has inspired a plethora of innovative applications across industries.

If you haven't explored GPT for yourself, I suggest you should ask it some questions. For example, what it knows about you or a celebrity, or whether it can explain how something works. While be careful about the reliability of the information when asking GPT questions – as it is not always correct – it was an eye-opening experience because it was a new one.

What the platform does is provide a proof of concept for generative AI. It gives us an idea of the possibilities of AI and how it can bring a new paradigm to the way we work. For us at the Office of the Chief Technology Officer at F5, some interesting explorations into the potential of the technology for application delivery and security could lead to the rise of AIOps.

The shift from imperative to declarative to generative

One of the challenges of IT infrastructure is configuring the large number of devices, services, and systems required to deliver and secure a single application. Without counting as-a-service offerings, enterprises typically rely on an average of 23 different application services.

I don't have to tell you that configuring web applications and API protection services is different from configuring plain old load balancing services. This means that the individual responsible for configuring and operating the application service may have to be an expert in more than a dozen different languages.

The industry has struggled to meet this challenge for years. While APIs become the primary way to configure everything, application delivery and security services are no exception. Everyone relies on imperative APIs, which essentially changes the way commands are issued. For example, instead of typing commands on the CLI, you send API commands over HTTP.

Again, it's not ready to deploy, although it's already very close to functionality and only takes fifteen seconds to build. What's more, it doesn't require any training from me. Moving from generation to automation. But it's an easy thing to do. I should be able to instruct it further, "Oh, by the way, deploy it." "When I enjoy my morning coffee, this technology should be able to do that. If I ask for it too, sing me a little song.

But it didn't end there! If I also want to tell the generative AI system later, "Hey, Green Bay has a lot of user logins and slow performance, please clone application A and move it to our site in Milwaukee." "What to do?

Indeed it is. Because if we look deeper, all of this is just a network of APIs, configurations, and commands that can now and often are automated via scripts. These scripts are usually parameterized and loosely related to the parameters in my AI prompts: Green Bay, Milwaukee, App A. So what changes is the generator, and how quickly it is generated.

I often say that AI and automation are force multipliers. Because technology doesn't know what it needs to do, but we do. But AI and automation can automate tasks through a network of APIs, configurations, and commands to complete tasks faster and more efficiently. Here, AI can effectively increase productivity, shorten time to value, and give experts time to focus on strategic decisions and projects while AI can learn from them. Over time, AI can further enhance our capabilities and open up new possibilities.

This is no longer science fiction, but the reality of computer science.

Generative AI will enable tomorrow's AIOps

Many of today's AIOps solutions rely heavily on pre-existing configurations and provide insights that are missing from 98% of organizations.

It is important to remember that they only solve yesterday's problems, not tomorrow's needs.

In the field of AIOps platforms, AIOps platforms with higher levels of autonomy, such as security services, increasingly rely on pre-existing configurations and well-formed responses. It typically does not use artificial intelligence to enable operations to be performed more autonomously across heterogeneous application delivery and security layers. Here, AI is used to analyze data and discover insights beyond human capabilities and time constraints. But that's usually where it ends, at least for layers above the network and well-known security issues.

That's where generative AI comes into play, and that's why I'm fully committed to exploring the huge potential of this technology to streamline application delivery and security processes. Think of this as the frontier of the AI revolution.