Graham Robinson, Chief Technology Officer at Data#3, and Carl Solder, Chief Technology Officer at Cisco ANZ, discuss the role of AI in technology, its challenges, and its future potential.
Watch the video in full or jump ahead to the questions linked below.
Graham: Carl, there’s a lot of talk about AI. Many people say it’s at the top of the Gartner hype cycle. There’s a growing number of people now thinking that AI is going to plummet into that trough of disillusionment and it’s going to be years until we actually see any real value from the technology. What are your thoughts?
Carl: We’ll eventually reach a point where AI will address and solve many of the challenges customers are facing today. With the advent of current technologies like ChatGPT, Gemini, and Chord, many people are exploring how to integrate these tools into business operations.
That said, let’s not forget that AI itself is not new. From Cisco’s perspective, we’ve been developing AI for over a decade. In fact, our first AI toolset was launched nearly seven years ago. Cisco has been on this AI journey for quite some time, consistently focusing on how to leverage AI effectively to address customer use cases. I believe we have some fantastic technology solutions available today that are already achieving this.
The release of ChatGPT brought AI to the forefront of public awareness. It essentially announced to the masses that AI is here. However, it’s important to recognise that a broader foundation of AI technologies has been around for a while. These mature solutions are already solving real business problems and are ready for organisations to take advantage of right now.
Graham: When you read the Harvard Business Review, they highlight that 80% of AI projects fail, but only 30% of generative AI projects fail, that means 70% of generative AI projects are not being cancelled and are proceeding through to production. Compared to the stats on digital transformation projects where only 12% succeeded – that’s a huge improvement. AI initiatives are performing multiples better then digital transformation projects. So, I fail to see how we can really say that there’s no value when we’re already seeing progress.
Carl: I agree. On the AI front, there’s generative AI, but we also need to consider how customers are consuming AI. You have the toolsets—like generative AI with chat-style interfaces where you type, interact, and drive workflows. But there’s also generative AI embedded within existing solutions. Often, customers don’t even realise it’s there, yet it’s quietly doing its job and delivering value behind the scenes. That’s where Cisco has been focusing for years. Many of our solutions already have AI built in, operating in the background. Customers may not always notice, but that embedded AI is making a difference.
Graham: The real value of AI adoption lies in moving beyond the hype and turning AI into something that delivers tangible, meaningful outcomes. It’s not about the flashy, in-your-face AI like ChatGPT—it’s about the AI that works quietly in the background, embedded within code and applications, where people use it without even realising they’re engaging with AI.
Carl: Correct. A classic example is the AI-powered radio resource management in the Cisco Meraki dashboard. Imagine a customer with a Meraki Wi-Fi network—there’s an AI engine running in the background, analysing user behaviour, network settings, and configurations. It optimises and fine-tunes the network to deliver a better Wi-Fi experience. Users benefit from this improved performance without necessarily realising AI is at work. It’s seamless, purpose-driven, and delivers clear outcomes.
This kind of embedded AI is something we’ve been leveraging for years. While you could debate whether it qualifies as generative AI, it’s still AI, serving a critical function. Now, with generative AI, we’re seeing entirely new possibilities, like virtual AI systems powered by large language models. These systems represent a shift in how operators interact with infrastructure. Instead of navigating a GUI and clicking buttons, operators can type commands and engage in a conversational workflow.
For example, a security operator could use such a system to craft an optimised security policy. The AI could analyse existing rule sets, identify redundancies, and consolidate rules to make the policy leaner and more efficient. These types of use cases highlight how generative AI can enhance productivity, improve decision-making, and ultimately drive better outcomes for organisations.
There’s also generative AI embedded within existing solutions. Often, customers don’t even realise it’s there, yet it’s quietly doing its job and delivering value behind the scenes.
Carl Solder – CTO A/NZ, Cisco
Graham: AI helps solve a lot of problems, but it raises the question: is it more trouble than it’s worth? One of the most common concerns I hear revolves around the potential threats associated with AI—security risks, overexposure to technology, and fears about AI being used against us. What’s your perspective on AI’s impact on security?
Carl: From our standpoint AI serves as the foundation for how we’re evolving our security portfolio. AI appears in a number of different ways. First, on the backend, every security device needs up-to-date information to identify and respond to current threats. We have an organisation within Cisco actively scouring the internet tracking billions of artifacts daily—emails, file attachments, web links—all in search of emerging threats.
It’s impossible for humans alone to sift through that volume of data, so AI steps in to scale the process. AI analyses those artifacts, identifies new threat vectors, and creates signatures to inform our security solutions. For example, when a solution encounters a flagged instance, it knows to pay closer attention.
AI also works within the technology itself, embedding functions that serve specific purposes. For instance, an AI system might monitor for potential threat signatures and alert SecOps teams when something warrants further investigation.
Beyond these applications, we’re also seeing AI, particularly large language models, redefine how operators interact with networks. This represents a new paradigm—operators moving from GUIs to conversational interfaces powered by AI. This shift is going to be driving a really different way in which operators are going to be doing the job.
Graham: We’ve been in the industry long enough to witness major technology disruptions – the evolution of the PC, the internet and cloud. Each wave has brought significant reskilling. With the rapid acceleration of AI, how do you see this impacting the current skills gap? Do you think AI will exacerbate the issue, or can it be leveraged to help close the gap?
Carl: First thing I’d say is that we’ve both chosen a career where you never stop learning. You take a break, you come back and oh that’s something new that didn’t exist before. The mere fact of being in IT means that you’re going to be on this continuous learning journey because technology is always changing. I see the advancement of technology picking up pace and the generation of new IT practitioners today have got it a little bit tougher than maybe when I started.
Graham: I used to say the only thing constant is change but even the rate of change is no longer constant – it’s accelerating.
Carl: Absolutely, and that means IT careers are inherently built around continuous learning. AI is just another evolution, another chapter in that journey. That said, AI does change the game in terms of how operators will work with infrastructure that they manage.
When I look back, I see an evolution in IT operations. From the traditional NetOps – when I started everything was CLI (Command Line Interface), beautiful CLI! We went through this motion of DevOps where we started using software defined networking tool sets to automate. We also went through this with AIOps a few years ago where we had AI engines embedded in the solutions to enhance operations and streamline specific tasks. Now we’re moving into this era of large language model operations. Instead of just running scripts or using GUIs, operators can now use natural language to implement workflows, troubleshoot, or optimise systems. This shift not only demands new skills but also offers tools to bridge gaps, making IT professionals more efficient in managing modern infrastructure.
Graham: Every period of technological advancement is getting shorter. What does it mean for our people today because technology has accelerated out of the gate over the last five years. There’s also been a number of things happening around the world that have stopped us from upskilling them cross-skilling our people at the same pace. It feels like the gap between technological advancement and education is widening. What role does AI play in addressing that gap?
Carl: When you look at the AI solutions in Cisco’s portfolio today, most IT professionals will find them relatively easy to adopt. For instance, many people have already used tools like ChatGPT, and we’re starting to see similar interfaces integrated into administration panels now. It becomes intuitive, but there will definitely be a learning curve, especially in understanding how to frame questions or commands to leverage AI’s full potential.
The one area that Australia will get to at some point is when you actually build and run your own custom AI workload tailored to your specific business needs. What’s exciting is these tools have the potential to help all employees perform their roles more effectively. It’s not just an IT transformation; it’s a workplace evolution where AI becomes a ubiquitous tool for problem-solving and productivity across all functions.
Graham: When you mention a custom AI workload, you’re referring to a specific application tailored for that customers’ needs. One that would probably be leveraging a third-party large language model or small language model or micro language model to perform a business function. We’re not just talking about the data centre. This extends to edge computing, specific devices, even down to mobile devices. Is that what you’re thinking?
Carl: Absolutely. In Cisco’s portfolio right now, we have an AI chat bot in our contact centre. This chatbot, powered by a large language model, enables true conversational interactions. If you think back a few years, chatbots were rigid, relying heavily on syntax and specific keywords to function. If you didn’t phrase something exactly right, they simply didn’t work.
But now, these chatbots can handle fluid, natural conversations, hugely improving the user experience. It’s not just a better customer interaction, it’s a transformative leap in how businesses engage with their users.
This also extends to IT practitioners. It’s a new world that our operators are starting to go into. Whether it’s automating routine tasks or assisting in daily workflows, these AI-driven systems are opening up new opportunities for efficiency and innovation in business operations.
Graham: Final question because I know we’re well and truly out of time that I could sit here and we have this conversation all afternoon. AI – Are we heading to utopia or dystopia?
Carl: I am a glass half full person, so I’m hoping it’s utopia. Within Cisco we have a group of engineers dedicated to ethics and how we can use AI in an ethical way. I’m optimistic that our engineering teams build AI tool sets that are going to serve mankind in a better way. That they’re going to produce better outcomes for businesses. They’re going to drive better productivity. better profitability, a better customer experience. That’s my hope for this.
That being said there’s always the darker side of human nature who might look to use those tool sets in in a bad way and I don’t think that we can avoid that. The only thing we can do is to continue to build out this technology to help mitigate those threats.
Graham: I share that hope. Going back to our conversation regarding security we know there are malicious actors out there. My hope is that AI for the first time really gives us an opportunity to harness the good in people and scale it to a level that we can actually provide a better future.
Within Cisco we have a group of engineers dedicated to ethics and how we can use AI in an ethical way.
Carl Solder – CTO A/NZ Cisco
Data#3 is a Cisco Gold Partner dedicated to helping organisations build secure, connected, and future-ready operations. As a Master Security Specialised, Master Collaboration Specialised, and Master Networking Specialised partner, Data#3 combines deep expertise with Cisco’s leading technologies to deliver tailored solutions. Recognised as Cisco’s APJC Customer Experience Partner of the Year for two consecutive years, we are committed to driving exceptional outcomes for our customers. Learn more at www.data3.com/cisco.
