AI and automation empower groups to develop into extra environment friendly by supporting duties associated to community monitoring, change administration, efficiency, safety, and a lot extra.
Nevertheless, the ability consumption required by the know-how will proceed to overpower present networks. Efficiency challenges, longer processing instances, and rising safety threats are just a few hurdles you’ll face due to bigger workloads.
On this archived keynote session, Neha Kaul, head of automation and AI know-how for Guardian Life, explores the ever-evolving challenges and crucial variations for community leaders as a result of rise of generative AI and bigger workloads.
This section was a part of our dwell digital occasion titled “Tips on how to Make Community Administration Simpler in 2024-25.” The occasion was offered by Community Computing on September 19, 2024.
A transcript of the video follows beneath. Minor edits have been made for readability.
Neha Kaul: For GenAI and bigger workloads, we’re going to see an increasing number of stress on our present networks as enterprises. We’re going to see efficiency challenges, longer processing instances, and new threats arising, particularly with AI.
With the black field fashions, we actually do not know what they carry. There was a number of emphasis on open-source predictive AI fashions, together with open-source GenAI fashions. We do not know the place they get injected with what.
So, there’s going to be a bigger scope on the threats. Now, what do I like to recommend? In a time when community leaders are wanting ahead to the altering calls for of shoppers and their very own companies relating to use instances, I like to recommend they relook at their foundations.
On the base of the muse is our community structure. If now we have an optimum community structure, scaling that structure turns into very simple. Scaling that structure would imply including newer parts that may simply be built-in into your present structure.
It is necessary to relook at these architectures, particularly from the safety perspective with zero belief. As we implement that, we additionally increase all of that with automation. We’ll speak extra about AI within the additional slides, however that is how I have a look at the longer term course. Setting the muse for that future course.
It would not be an overstatement if I mentioned that in 2025, community automation and AI instruments, together with an optimum structure, is not going to solely be a luxurious, however a necessity that almost all community and know-how leaders should have a look at.
So, on this slide, I simply wished to the touch base on what all of us have seen taking place within the final 10 years. In 2010, we had IPv6 adoption taking place, and it continued by way of the 2020s as nicely. However that is the pivotal level the place we have been in a position to get bigger IP addresses, areas, and improved community capabilities to take care of most of our networking wants.
From 2014 to 2016, it was additional augmented by LPWAN applied sciences arising. These have been low-power, wide-area networks that helped us develop to the broader areas and but had low energy consumption. As a result of on the finish of the day, if you run your information facilities and all these high-load networks, you need to take into consideration the associated fee financial savings.
That is why low energy consumption helped. This additionally helped us connect with the Web of Issues (IoT), which is all of your web gadgets. These are your cellphones and plenty of different gadgets which might be community enabled. Houses are even community enabled now. So, all that noticed huge progress in 2014-2016.
Someplace round 2015 to 2017, we had SDNs and NFVs, once more, shaping the way forward for cloud. There may be extra virtualization taking place. SDNs are increasing and decoupling a number of our software program from the {hardware}.
One level I wish to make is that every one these know-how developments have been serving to set a base for the following wave of developments. Once I was placing the slides collectively, it gave me an opportunity to take a look at this evolution and made me understand that none of it was in silos.
The whole lot that occurred as part of the development had one thing previous to it, asking it to occur, and one thing sooner or later the place these applied sciences have been setting and filling these gaps.