AI Strategy Manifesto

Overview

Generative AI has sparked the world's imagination. It's the technology of the moment. It has made coding fun again and made college students rethink their majors.

ChatGPT made AI visible, tangible, and most importantly—accessible. AI went from this complicated thing for PhDs and data scientists to something everyone could play with and therefore understand its potential. Especially those in the boardrooms.

Cisco is well positioned to leverage this AI moment to benefit our customers and our own employees.

We're in the sweet spot to be a leader in the AI for the Enterprise space.

With our capabilities across Silicon One, networking, and optics, we are well positioned to build out the AI scale infrastructure needed to move the world to an AI-enabled future, especially as ethernet increasingly displaces broadband.

A bit of background: Generally speaking, there are five advanced AI capabilities: Human-AI Interaction, Decision-making, Motion, Creation, and the one that scares some—Sentience.

Creation, of which Generative AI falls under, is one of these advanced AI capabilities. All of these capabilities are amplifications of human attributes. GAI is just another form of AI as an amplifier—only on a scale we've never seen before.

But just like electricity wasn't about the kilowatts and the power grids, AI is more than just LLMs.

There is no one formula that is AI, just as there are many additional methods of AI that are—or have the potential to be—just as groundbreaking.

Many of these methods, including Generative AI, have been infused across our portfolio to enhance our products. For example, natural language processing and collaboration, AI-driven security policy management, and the ability to monitor larger AI models themselves with our observability capabilities.

We see AI as more than the method used—it's about the problems it's helping to solve for our customers that human effort couldn't accomplish on its own.

While AI is not new to Cisco, Generative AI does raise novel issues.

But it doesn't change our approach to AI—human-first.

Responsible AI is non-negotiable because responsible tech is a given for us. Always has been for Cisco. Ethics, privacy, sustainability—are just another layer of requirements of the stack. Mandatory, reviewed, and beyond reproach before anything is shipped.

Enabling Customers

At Cisco, deep listening comes before deep learning. Or machine learning. Or computer vision, etc. (pick your AI of choice) AI for the Enterprise is first and foremost about customizing and adapting technology to the customers' specific needs and priorities.

Before Predictive Networks or Generative AI Policy Assistant became a Cisco solution, they were first a customer conversation.

Understanding AI's value and achieving it are two different things. That's why our customers look to us for our expert knowledge to understand the complex dependencies and help to create the conditions for them to capture AI's exponential value. Because the businesses that take advantage of AI will dominate their vertical. Cisco can help make that happen.

Our AI Advantage Across Five Pillars

1. Silicon and Networking — Building Infrastructure for AI Workloads

Our core networking technology is already powering some of the leading AI models run by hyperscalers around the world.

Silicon One infrastructure is a fundamental advantage because of its flexibility. It unites economics and AI data compute needs of today—and tomorrow.

2. Security — Security and Intelligence Embedded in Every Connection

We secure organizations at machine-scale: AI for Security, Security for AI.

We leverage AI to deliver the most effective cybersecurity solutions, and we secure AI systems themselves from threats.

3. Observability

A simple swipe on an app puts in process thousands of actions to hundreds of computers.

The end user doesn't really care about the tool. They don't care how it happens, they simply want the experience promised: Do what I expect. Do it fast. Do it securely.

You can't automate what you can't see.

4. Networking — Everything Relies on the Network, Including AI

AI transforms the network to be more than passive infrastructure and instead turns it into a technical force with the power to positively impact business outcomes.

5. Collaboration — Create Immersive, Hyper-Realistic Hybrid Work Experiences

AI enables us to have highly effective and inclusive experiences for the global workforce by using Natural Language Processing Capabilities (NLP), Audio Intelligence and Speech Enhancements, Video Intelligence, and AI-enabled Insights.

Connected Intelligence

Today is a point in history where there isn't just one digital muse igniting a technology renaissance. There are many. Cloud. Power compute. 6G. Superapps. Edge. Distributed. Blockchain. Each a muse in their own right—going around inspiring everyone and transforming everything. They wait for no one.

Each individually remarkable. But none stand alone. Including AI.

The real power we are all feeling at the moment—the magic—is when the muses team up. Are networked together, to give capabilities exponential value as we layer them, combine them and use them in ways beyond what they were originally intended.

The future will be one of compounding intelligence.

Summary

Cisco—Making the impossible happen faster with AI.

It's been called the information economy. But it's really the intersection economy. The magic is in the intersections of connecting anything to everything. AI will make this happen faster, better.

Epiphanies can be hidden in the most ordinary of intersections. AI can spot the patterns in places humans never even thought to look.

Cisco's AI Story

Brand Messaging Framework - October 2023

At Cisco, AI is anchored in the practical, the useful, the responsible.

That's just who we are. Customers trust us because we're neutral. Transparent. Safe. Grounded doesn't mean we're not groundbreaking. We anticipated the Internet for the Future and challenged how networking chips were built to create a new kind of Silicon that's powering the Gen AI workloads of today.

Then there's our data. Along with Cisco's nearly 40 years of networking expertise, we started deploying AI solutions at scale across our portfolio over a decade ago. Our data is massive. It transcends specific domains. It's near real-time. When correlated? Massive becomes meaningful. No one can come close to doing what we can. To predict. To secure. To observe.

Where security meets the network is where Cisco shines. AI will amplify this even more. It will also introduce new threats and vulnerability points during transit across the network. Multi-cloud. Multi-endpoint. Multiple applications. Multiple models. Multiple platforms. The network just took on new importance. The transport matters. Again.

Cisco is optimized for AI.

Last year we pioneered the first-of-its-kind predictive networks. This year we introduced our Full-stack Observability platform. Right on time to give our customers the assurance they need to explore the new AI landscape. Go fine-tune. Go knowledge graph. We've got you.

It's been said that it's 1995 in AI, not 1999. Maybe, maybe not. But everyone can feel we are on the cusp of something big, something enormous. We feel it too. As Prompt User Interfaces become more mainstream (you'll see them soon across our portfolio) and as Conversational UIs proliferate, apps will fade into the background as code that the end user won't give two hoots about. They just want the experience. They will get it by using their voice, or making a gesture, or even by a simple look. How all of that will happen—like most things in tech—will be in the background. In the network. That intertwined tangle that communicates to make things happen. This to that. That to this. Times a zillion.

All the nodes, codes, models—they need to be secure. They need to be connected—hyperconnected.

The sequencing of how all of this happens needs to be choreographed on a whole new level for AI. Models will be in our phones, in access points, in sensors.

AI is not something that will be a stand-alone experience. It's going to be end-to-end, it's going to be handoffs between devices, between technologies. Security and intelligence will be in each and every connection. It's going to need Cisco.