I’ve been at Cisco a few years and seen a number of transformational occasions have an effect on community engineers. After I first joined Cisco in 1998, IP telephony got here on the scene and disrupted not solely classical circuit-switched communications however these of us in networking as effectively.
Extra lately, automation, programmability, and software program grabbed everybody’s consideration, promising to upend the way in which we take into consideration, design, and function networks.
In the present day, you’ll be able to’t go 10 ft (3 meters for these of you exterior the US) with out somebody speaking about AI. CES 2024 has simply wrapped up, and Synthetic Intelligence (AI) was in all places.
Whereas IP telephony and automation might have been reserved to particular locations in IT, folks from all walks of life are discussing and utilizing AI. The networking trade particularly is shifting quickly round know-how and expertise acquisition in addition to product integration and focus. All this variation and hype can depart folks confused and questioning how they’ll wrap their head round what AI is and the way to greatest make use of it.
Earlier than going an excessive amount of additional, it’s vital to say that AI isn’t just one factor. It additionally isn’t new. AI is a category of machine studying that makes use of math to investigate massive units of knowledge to make predictions, present automated classification, and summarize massive knowledge units.
Actually, a number of years in the past, when this notion of machine studying (and AI) began to first emerge in networking, I took a web based course that targeted closely on the maths behind machine studying to do issues like cluster numerous take a look at outcomes to determine anomalies and analyze a big set of quantity photos, offering new photos so AI might appropriately determine the quantity it receives. This identical form of predictive and classification AI is already baked into a number of Cisco merchandise. Have you ever ever been on a Webex with somebody mowing exterior, but nobody on the decision can hear something? That adaptive noise canceling is a type of AI.
Capturing all the eye at present, nevertheless, is generative AI, a brand new use case that, as its title suggests, is about producing new content material by analyzing an array of present content material. This generative AI receives a immediate that specifies what sort of content material to create and in what vogue, and it makes use of its underlying mannequin to fulfill the request. Totally different generative AI instruments exist for various media and totally different use instances. Listed below are some methods of utilizing it:
- Ask numerous generative AI methods to jot down a brief story within the type of Ernest Hemingway.
- Create an image of you shaking Albert Einstein’s hand.
- Convert a video you shot in your cellphone right into a silent film.
Due to the viability and availability of generative AI, the excitement surrounding it has grown to an all-out symphony. Folks have been making an attempt out totally different use instances for it in addition to performing immediate engineering to fine-tune the outcomes. As with all highly effective instruments although, we should take care. ChatGPT, for instance, warns you that it could possibly make errors.
Community engineering is an attention-grabbing use case for generative AI. You possibly can ask a software, resembling ChatGPT, to construct a easy OSPF configuration for a Cisco router, and it’ll not solely generate a pattern config, however it is going to additionally clarify a part of it. The preliminary config it generates can have placeholders in it, however you’ll be able to immediate ChatGPT to then regenerate the config with pattern values within the placeholders. This makes for a pleasant technique to bootstrap a community testbed or assist crystalize new networking applied sciences by means of instance.
How will you study extra about AI, its use instances, and its caveats? Cisco U. has you lined. Cisco U. All-Access customers can discover programs and studying paths on the foundations of generative AI, moral and privateness issues with generative AI, and others.
For these seeking to study extra about how generative AI can form community engineering and operations, we simply launched two tutorials included in Cisco U. Free, “Create a ChatGPT Shopper with Python” and “Chatting with Cisco IOS XE”. These are nice if you wish to learn to construct your individual integrations with generative AI.
Bear in mind I stated that Cisco has been folding AI into our merchandise for some time now? You can begin to get worth from AI at present in your networks utilizing quite a lot of Cisco options.
Take into account your WiFi community. I’m positive it shares traits much like all WiFi networks, however I’m additionally positive it has distinctive facets of how and the place to deploy it. As a substitute of utilizing static, manually tuned thresholds, you need to use predictive AI in Catalyst Heart Assurance to supply extra correct alerts when issues happen. Within the picture beneath, the inexperienced band is the predicted baseline, calculated over an prolonged time interval, whereas the blue line is the precise worth. The purple bands sign anomalies, which can generate alerts.
Cisco U. has content material you need to use to study extra about Catalyst Heart Assurance in addition to ThousandEyes Web Insights and Meraki Perception. On January 22, 2024, we introduced a brand new CCNP focus, Enterprise Community Assurance (ENNA) that covers abilities you’ll want to make use of those AI-enabled capabilities. The training coming this spring for ENNA will delve deeper into these analytics and assurance options.
I began out speaking about how, if you happen to’re on this trade lengthy sufficient, you’ll see your share of disruptive applied sciences. A previous concern related to automation was it could make community engineering out of date. As a substitute, it’s made community engineering extra scalable and thrilling. Likewise, AI is a software that helps push these pleasure and functionality boundaries farther. Now could be your time to study, experiment, and focus on.
If you’re prepared, put up within the feedback what you need most from AI and networking.
Comply with Cisco Studying & Certifications
Use #CiscoU and #CiscoCert to affix the dialog.