Guy Daniels, TelecomTV (00:13):
Hello and welcome to another seasonal edition of Extra Shot with Added Marshmallows. This is Telecom TV's wrap up of the Telcos and AI event, which was held in London during the first week of December alongside the great Telco debate. If you have yet to view all of the sessions, then hopefully we can persuade you to do so with the following carefully curated highlights. So let's get started with the first session of the day AI deployment scenarios for telcos. In fact, it proved so popular in the runup period that we decided to devote all morning to the topic, splitting the discussion into two parts, and for the first half we wanted to explore where Telcos plan to deploy AI in their networks and learn more about some of the promising use cases. We also wanted to find out who should be responsible for AI within telcos and who's making the decisions.

Ray Dolan, Cohere Technologies (01:24):
I think if it's cost reduction, it's probably the CTO and I think if it's your revenue generating, it should be the CMO and hopefully those two roles exist in telcos sometimes. Sometimes you've got an engineer running the marketing organization, and if you don't divide it that way, the revenue side's never going to show up because the engineering side is just wired to reduce cost and then if possible, drive revenue. So I think you got to bifurcate it. We're much more focused on the cost saving performance process almost. We have no focus right now on the revenue generating side of ai,

Prashant Agarwal, Intel (01:57):
So we need to as industry come together and see, okay, we have convinced the CTO, we know that it works. It'll provide this optimization, but how can we convert that in terms of the monetary values? That's solid bit struggle, to be honest. Still seeing

Andy Burrell, Nokia (02:11):
Very often it is the CTO, whether that's the right place for that to be made is another thing. Perhaps it should be Chief Digital Officer or someone else. But yes, pretty much CTO.

Chetan Narang, Colt Technology Services (02:22):
Yes, I think there should be an AI center of excellence within the organization, which is basically a cross-functional team kind of aspect, which touches the product, the it, the data office, and also the governance side because with the governance, you get into the you need from a risk and a legal aspect also to be part of that.

Guy Daniels, TelecomTV (02:40):
This conversation also led onto the thorny issue of procurement, perhaps not the most obvious of subjects for an AI discussion, but it demonstrates just how much groundwork is still necessary before telcos can reap the benefits of ai. These are the issues that are still causing concern within the industry. Well, for part two of our session on AI deployment scenarios for telcos, we invited a new panel of guests to the stage and this session covered use cases including those related to energy efficiency, something we will be looking at more closely in our next Green Network summit, but it also came back to selecting the right trusted partners and ensuring the reliability of AI solutions in mission critical networks. In other words, creating the optimal conditions for AI deployment scenarios.

Rajwinder Singh, Ericsson (03:37):
If as a telco you need a simple enterprise virtual assistant to write ops for you, you can do it yourself or with an SA vendor. If you need expertise into analytics or into O-S-S-B-S-S or ran, then you go onto the vendors like Ericsson Nuia and all the reason for that is we know our data, we know our data catalogs where the schema is, you can develop on your own, but then say goodbye to TCO.

Manish Singh, Dell Technologies (04:08):
What data do you have and where is that data and how good the quality of that data is. From there in, you start to build who are the right set of partners or partner or partner that you bring in into realizing that use case.

Chris Voudouris, Airband (04:26):
Also, systems integrators, they need to change their approach. Look to help you out to make more transition to Gentech, and I would say probably you have to work with all of them and pick the leaders that start to invest more and dominate this specific segment as the game develops and it's not set right now.

Graeme Oxby, Community Fibre (04:46):
Customers have to take ownership of the problems they're trying to solve fundamentally is what you're saying. You've got to know what problem you're trying to solve. You might not quite know how to solve it, but if you know what problem you're trying to solve, you also know the risks of solving it in a bad way, like turning something off. So you can't delegate your accountability for the problem solution and the outcomes that might emerge from the wrong deployment or bad deployment of ai. Exactly same as any other technologies.

Vess Bakalov, IBM (05:14):
To your point about ais and reliability, I think we're setting a very high standard for reliability. Ais to be accepted in production are going to be a lot better than junior employees. I think they're going to be better than senior employees by the time we trust them to do what they do.

Guy Daniels, TelecomTV (05:34):
Just like the first session, this discussion highlighted the work that still needs to be done before we can talk sensibly about actual deployments and results. And so to our third session, which looked at the network challenges of implementing ai, and this was a tricky one. We wanted to find out more about the issues faced by telcos and their vendor partners in the implementation of AI network capabilities, but more to the point where exactly is AI being used in the network today?

Jeremy Chelot, Netomnia (06:09):
Currently for us, the only real use case that we've seen at the network level is really inside the home. So if you look from a network perspective, because being able to improve customer experience in the home by looking at wifi, looking at speed, looking at a handover between several access point or nodes that you've got inside the home to try to understand basically when there is a problem that you can be proactive rather than reactive is the only place so far where from a network standpoint where we found basically a use case

Aaron Boasman-Patel, TM Forum (06:48):
When we talk about AI in the network, for us it's all about level four because that's where you having autonomy and that's where AI is actually doing what it's supposed to do rather than just say using advanced analytics for example. So for us, really what we are working to now, we are working with Orange, we're working with Deutsche Telecom and we're working with China Mobile to prove out some of these high level what we call value scenarios. And we've got four of those 15, which we're targeting actually for DTW in June to be able to showcase them actually working on live networks. And those four areas around single domain troubleshooting, it's around home broadband service assurance, just very much like Joan was talking about private line service provisioning and around optimization. And we think that those are the first four we're actually going to see live in operating networks using genuine AI to solve some of these challenges and problems.

Guy Daniels, TelecomTV (07:33):
So very much a developing area and highly dependent upon what type of telco we are talking about. And the session went on to look at the security issues of AI in the network along with organizational readiness. Well, for the final session of this year's event, we attempted to bring together all of the conversation threads of the day and look at how best to build the telco AI ecosystem. Specifically we wanted to discuss what new partnership models and relationships we can expect, and whether telcos need a center of AI excellence.

Rahul Atri, Rakuten Symphony (08:10):
Surely somebody can play that advocate because there are enough smart people who can be the chief AI officer, but if not this kind of centralization in terms of budget, authority, autonomy, then everyone should have the target because without that autonomy, you're going to add one more layer and that poor guy will have KPIs for everyone else, and I'm very sure others would not support because others would have their own budget and they'll buy their own tools and they'll have their favorite partners to work with, and everyone will create an AI platform and everyone would want to eat away somebody else's platform. So take the call to have one platform and one team to drive it,

Jeet Singh, Tech Mahindra (08:47):
And that person really has to believe that this is the future, this is what the company will pivot on. Once you've got that, you will then empower again, like Rahul said, have that one unit which has the mandate from the CEO to drive and infuse AI across the organization because look, you have to infuse it across the organization or it doesn't work because otherwise you do POCs and that's where it'll be at because if you've got to scale it at an enterprise level, different parts of the organization need to come together.

Laura Murphy, BT (09:25):
I think there's certainly a role for champions, center of excellence, so on and so forth to really drive out those horizontal streams of value across an organization. If you have a central team or a central champion that sits above or alongside those kind of organizational building blocks that can cut across them, they're certainly a very useful role for that in terms of driving those horizontal end-to-end patterns of value.

Guy Daniels, TelecomTV (09:54):
Well, we also featured a fireside chat with our special guest, Lord Chris Holmes of Richmond in conversation with my editorial colleague, Ray Le Maistre. And Lord Holmes gave us an exclusive update on the progress of the UK AI bill, the new government's approach to AI and the potential implications for the telecoms industry. Here's just a little of what he had to say.

Lord Chris Holmes (10:20):
The only way we are going to make the success of this that we can is if we think cross economy, cross sector, cross society, horizontal approach. Because what you would want, if you're an innovator, an investor, a citizen, a consumer, you would want to know that wherever you bump up against ai, be it telcos, be it in buying your weekly shopping, whatever environment, what would you want you would want clarity, consistency, coherence, and the certainty that comes from that. And that will only come if you take a horizontal cross sector cross economy approach rather than doing a specific bill on large foundational models and then saying beyond that, it's really up to each individual sector and each individual regulator to determine how they're going to approach this.

Guy Daniels, TelecomTV (11:21):
Really interesting comments from Lord Holmes showing the different approaches being taken towards AI regulation. Of course, all of these highlights represent only a very small percentage of the debates and presentations at the Telcos and AI event. If you want to see and hear more than they are all available to watch on demand in full at the telecom TV website. Before we go, my three quick observations from this year's event, one agents 2025 is going to be all about agentic AI and developments are happening thick and fast. Now, one of our speakers, Chris Voudouris, CTO of Airband, gave a presentation on this very topic during session two, but we're going to have to wait to see how telcos react to this. Two conversational ai. And I highlighted this at the beginning of the day, voice prompts turn detection, interruption management is going to open up how we all use and interact with AI services.

(12:27):
It's an opportunity for telcos to use itself to support third parties and to offer as a service. And three network ai. Where is it? As we heard, there are precious few real examples of AI being used in the network today. Telcos have to step up and fast, so this time next year, we want to be hearing a lot more about this. We want to be hearing about telco success stories. Well, we are back in February for our next online summit, the Green network in which AI features quite extensively, and of course a full strength extra shot. For now though, thank you for watching and goodbye.