Ray Le Maistre, TelecomTV (00:05):
It is FutureNet World 2025 in London. I'm here with Roy Chua from AvidThink, who has been in the chair seat today overseeing the part of the program or very much of the program today. And Roy, you've been here for both days. The overall show has really been very much focused on AI and automation. Are there some key takeaways from this event that you are taking away and that you think others might be taking away as well?

Roy Chua, AvidThink (00:35):
Yeah, sure. I think fundamentally there are a few things that I would say a lot of people already know, but I think it helps to have it emphasized. I think one of the key things that it's hard to avoid is data, data, data, data. So on top of ai, ai, ai, AI is all about data, and I think that is true. I think fundamentally what the telcos are realizing is that in order to extract maximum value from ai, everyone's realizing data's important. And so if you didn't take the steps earlier where we told them to go fix your data, take them out of silos, modernize your data platform, what's your data strategy, right? We said that a couple of years ago, it was important for cloudification,

(01:16):
It's still just as important, and so you can't get to step three, which is the AI step without step one data or step two cloudification or making things more agile. And so I think they're realizing that there is no sort of shortcut. You have to go back and start step one, step two to get to step three, or at least in order to get full value from step three, we have to do that. I think that's one element. I think the second element is we understand that AI is a transformation, and just like any other transformation, changing the culture, changing the people reskilling, upskilling is important. We said almost exactly the same thing. We did the cloud transformation. If we remember 3, 4, 5 years ago, we said the same thing. We go to reskill, we go to upskill, we can transform. Who are the biggest blockers?

(01:58):
Well, it's people, it's culture. We're repeating exactly the same points again with ai. I think the one difference or subtlety here is that we never said anything about cloud taking a job, but people here are concerned that AI is going to take their job. So there's a more personal elements to it, but hopefully that drives them to want to understand it and embrace it and use it better. I don't think it's going to take away your job immediately, but certainly you should transform to adapt to it and learn how to use it as part of trying to figure out the role of humans in the new world where AI gets better and better and better. I think that's the second element. The third one, which I think we haven't explored a lot, is that the cloud, so to speak, or past technologies necessarily self-improving ai, on the other hand with reinforcement learning with maybe help from humans can improve and will improve with more data, with more time in the seat as it were, it will improve.

(02:58):
I don't think it's clear where that takes us yet, but just the early signs of the reasoning models and the ability to be autonomous, which we like the autonomous network ability to reason. I'm not sure where it's going to take us, right? But I think certainly we need to understand that AI will self improve, and it may take us down directions that we're not sure of. I think the one thing that I heard here was, well, we're giving low hanging fruit, but maybe we should not be looking at things like just customer experience or network troubleshooting or root cause analysis anymore. But what else can AI do if given the leeway to maybe it looks at data sets that we would normally look at, maybe it comes with correlations or moves that we don't

Ray Le Maistre, TelecomTV (03:40):
Expect. And I wonder if that's going to come maybe from the shift that's been hinted at here of a switch of mindset from the service providers to thinking about, okay, what does that group of users there want and how can we deliver it now in a way we didn't before? And maybe that will help to unlock some of that, I think would. I

Roy Chua, AvidThink (03:58):
Think if you give AI the resources and the ability to use skills tools, tool usage, connect our adjunct ai, where we give them the ability to connect the systems to pull information from multiple systems and the likewise program and act on those systems, I think we might see something different. And just like I think we saw Alpha go with that deep learning model. I think there was move 37 when it played and people were like, oh my God, move 37. No human would've thought of that. I don't know what is a move 37 movement in telco when AI is supplied? I hope you'll be happy when that happens, as supposed to freak it out. But I think fundamentally the journey is on. It's just like any transformation. We have to do the basics, the data layer, the cloudification layer, all have to be in place. But AI is different from anything that's come before. And we need to realize that. And I think we need to be very careful at the same time cautiously exploring it, guard railing it, understanding it, but that shouldn't stop us from trying to embrace it at the same time. So I think that's sort of my takeaways.

Ray Le Maistre, TelecomTV (05:09):
Okay. Well, Roy, you've got the name for your next podcast, which has got to be Move 37, of course. So that can be one takeaway from this conversation at least. So Roy, thanks so much for joining us.

Roy Chua, AvidThink (05:15):
It's a pleasure. You're very welcome. Thank you, Ray.