The famous AI pioneer Oren Etzioni once said: “AI is a tool. We have the choice of how it is used.”
Businesses decide how to use AI and theirs Use cases continue to evolve. Network teams typically use AI as a safeguard to protect against potential security threats, as a network management tool, or as a way to enable automation. As network environments become more complex and distributed, they produce vast amounts of data that humans alone cannot handle.
Enterprises may consider implementing AI to manage complex systems such as 5G networks, or collect data analysis. AI can monitor network performance and alert managers to potential problems before they occur. Some forms of automated AI can also fix problems without requiring human intervention.
According to a 2021 Gartner reportAdoption of AI for IT Operations (AIOps) – describes the process of IT teams using AI technologies, like machine learning (ML), to automate tasks – increasing in companies. Gartner estimates that the AIOps market was between $900 million and $1.5 billion in 2020, and Gartner expects it to grow at a compound annual growth rate of 15% through 2025.
One reason for the increasing acceptance of AIOps is that companies are at the beginning of the digital transformation. As operations become more digitized, it becomes increasingly difficult for humans to analyse, monitor and manage the newly accumulated data. Actually the Top trends of digital transformation of the past year included the deployment of ML operations.
Organizations are using AIOps strategies to “replace traditional surveillance tools” as they ultimately plan for a “post-COVID-19 pandemic environment dominated by practical outcomes,” the report says. New business requirements and pressures brought on by the pandemic have prompted companies to adopt AI.
While interest in AI is growing, not all organizations are adopting it quickly. Benefits aside, AI is still a new, advanced technology that has yet to reach its full potential and is turning organizations into leaders reluctant to use – and they’re not the only ones.
Network teams are also among those who choose not to implement AI. Read below what three networking analysts have to say about the state of AI in enterprise networks and how they believe networks will embrace AI in the future.
Editor’s note: Responses have been edited for length and clarity.
What role does AI currently play in networks that have deployed it?
John Burke
John Burke, Research Analyst, Nemertes Research: The use of AI and networks is still in its infancy. But, [in areas] Where it currently matters, the rules focus on visibility for administrative purposes. What is actually happening in my network? What is strange about what is currently happening on my network? What are the anomalies? The situation is very similar with security: What is actually happening in my network? Which of the weird things going on in my network should I be concerned about?
In both cases, the goal of the AI is to get between the raw data and the people, weeding out the normal wobble and the benign, uninteresting anomalies [so teams can] focus their attention on important things, either performance or safety. On the performance side, it’s, “Where do I have brewing issues that I need to address?” And longer term, “How should I plan for capacity going forward?” This is a bigger problem in data centers than anywhere else, but it is by no means limited to data centers.
Juniper just announced that it is adding AI capabilities to its SD-WAN [software-defined WAN]. There will be many more in this direction from everyone else, including Network-as-a-Service providers – particularly SASE [Secure Access Service Edge] Vendors – about using AI to better direct traffic, optimize delivery, and troubleshoot issues as they arise.
Are there use cases for AI in 5G networks?
John Early
John Fruehe, Independent Analyst: 5G carrier networks are environments where many variables are constantly changing. AI makes sense because you’re inputting a lot of data that you’re feeding into how things work. AI has little value on stable networks where network traffic looks the same every day and doesn’t change much. But in a carrier network, things are always changing, and AI can help with many deployments. As we deploy more 5G and we can get more granular, a lot of those connections and switches will happen, and that’s where AI makes sense.
Is AI likely to change the roles of network teams in any way?
early: Part of what has happened is that the pandemic has absolutely decimated all traditional roles [networking]. All the big plans that [network teams] had for 2019, 2020 and 2021 — [such as] Network migrations and rollouts – got a giant wrench. I think teams have probably spent the last two years in scramble mode now trying to figure out how to overcome all these issues and get back to a stable position.
I don’t think we’re at a point now where things are stable enough for people to start thinking about doing things on a higher level. There’s still a lot of blocking and tackling to do. Networks have become more distant than in the past. The biggest changes that have taken place in the networking space are in end-user location, which has driven a lot of software-defined WAN and VPN.
How do you see the influence of AI on networking in the future?
Bob Laliberte
Bob Laliberte, Principal Analyst, Enterprise Strategy Group (ESG), a division of TechTarget: As the environment becomes more complex, more and more data is transmitted over the network. This adds up to more complexity that is beyond human understanding to manage effectively. This is where AI and machine learning technologies will play a role.
However, it is important to note that AI/ML is not intended to replace humans. There will certainly be times when AI/ML warns and recommends, but cannot make any change. If something goes physically wrong with a switch or a cable, there isn’t a lot of AI/ML or automation that will fix it. A human needs to replace a cable or replace a switch or power supply.
Advancement in the use of technology is vigilant, vigilant and recommending and automating. [ESG research showed] that 20% of companies are fully automated. About 60% are in this warning and recommendation phase, which helps them use this information to do their jobs effectively. The last 20% are organizations that want AI to simply alert them to a problem so they can fix it themselves. Over time, as environments become more distributed and complex, it becomes harder to do this and troubleshoot problems in a timely manner. This is where the AI engines come into play.
