A recap of Telco AI Summit Americas and the key takeaways from the event
San Francisco is a popular tourist destination, known for its cool summers, fog, steep rolling hills and eclectic mix of architecture. As Joe Flower once said, “Money lives in New York. Power sits in Washington. Freedom sips cappuccino in a sidewalk cafe in San Francisco” (Now adjusted slightly to “Developers write AI algorithms at a sidewalk cafe in San Francisco” to suit Telco AI Summit Americas.)
The Telco AI Summit Americas is a 2-day event, and a sister event to the Telco AI Summit in Europe that we also attended, with attendees ranging from service providers to vendors in the Telco AI and Analytics domain. The summit brings together the telecoms community to discover the new opportunities within AI, analytics and automation. This more intimate event boasted some heavy hitting speakers from BT, Google, Telefonica, IHS Markit and the TM Forum.
Now, onto the key takeaways from the event.
Analytics is the first step in the evolution to AI-driven networking ecosystems
Analytics have become widespread and are rather well understood and used successfully in many organizations. Analytics is all about figuring out why did an event happened and uses techniques such as drill-down, data discovery, data mining, and correlations. But, where things get really interesting is when trying to use predictive analytics to project what will happen in the future. Typically this is done by using existing data to train predictive machine learning (ML) models. Analytics is thus part of the evolution that leads to AI.
And don’t forget about automation!
Network automation is the second most important AI domain. Without automation, the telecom business model is at risk of breaking down. For networks to support the billions of devices that are expected to be connected to the internet within the next decade, they must be self-optimizing and self-healing. This requires machine learning.
The below are some concise facts and recommendations Aaron Richard Earl Boasman-Patel, the Vice President of AI, Customer Experience and Digital Transformation at TM Forum, presented in his keynote titled Are you ready to exploit the AI opportunity?
- Telecom networks are complex, and the transition to cloud, virtualized functions, and software defined networking is increasing the complexity. 5G will add even more complexity.
- As internet of things (IoT) expands, network and service management must be zero-touch.
- A network servicing 10 million endpoints and 10,000 nodes could see these numbers increase by up to five times by 2020.
- 5G will have the potential to add to network complexity
Here comes AI!
The world we live in is a mobile world and in today’s mobile network, AI is starting to be more present. From flying drones performing cell site inspections (checking RF signals and detecting any malfunctions) to performing traffic load balancing, such as video, to using ML to detect coming congestion in a network, and even for RAN optimization, etc. For telecoms service providers, AI will be a key component of the whole telco digital transformation process leading to automation, intent-driven networking, and ultimately, AI-driven networking ecosystems. That’s got a nice ring to it, doesn’t it?
You may recognize the above slide from our recent webinar with IHS Markit on Exposing the cost tradeoffs of cloud-native NFV as well.
But, back to AI. According to Boasman-Patel at TM Forum (and the McKinsey Global Institute), AI could contribute an additional 1.2% to annual GDP growth for at least the next decade. In addition, 70% of firms will adopt at least one form of AI by 2030, and a significant portion of large firms will use a full range of that technology. Maybe the money is moving to San Fran…
However, there are also huge challenges that need to be overcome in terms of AI, such as:
- Dealing with fragmented or inconsistent data in order to enable better automation. As Bill Gates stated, “Automation applied to an inefficient operation will magnify the inefficiency. Automation applied to an efficient operation will magnify the efficiency”.
- In order to move forward with AI in the telco space, service providers need to hire individuals with the right skills (data analytics and SW expertise). This is easier said than done. Hiring AI capabilities is currently a big pain-point as telcos are competing against the Googles and Facebooks for the same skillset.
- In order to make AI more acceptable by consumers and specifically regulators, service providers will need to make AI explainable. AI algorithms are viewed as very complex black boxes, their decision making will need to be explained in a language that everyone can understand. In addition, they will need to address concerns about security when deploying AI OPs.
Today, AI is mostly used in 2 network domains; cloud services offered to customers and network operation and management. The biggest driver of AI for CSPs is “The Customer”, or more specifically, improving customer centricity.
The Generation C mindset drives AI forward
Boasman-Patel also covered Generation C in his keynote, a powerful new force in consumer culture. It is a term used to describe people who care deeply about creation, curation, connection and community. It is not an age group; it is an attitude and mindset. Latest statistics show that 91% of Gen C’ers sleep next to a smartphone. Within the last three to five years, improving customer centricity has become the single biggest strategic priority for telecom operators. AI is needed to give customers the kind of digital experiences they are demanding, and it can deliver these capabilities through chatbots and voice assistants.
Stéphane Téral’s final word at the AI Summit? “Telcos can no longer live without AI.” But what about AI bias? According to a recent Heavy Reading, 60% of communications service providers (CSPs) are concerned about the
potential impact of AI bias. Learn more about AI and breaking the bias barrier in this white paper from Heavy Reading.