Hyped to Death: AI Must Avoid Becoming a Cliché
AI use by telcos and providers can potentially improve customer service, speed onboarding, and reduce outages by pro-actively identifying problems. Kailem Anderson examines whether this is all hype, or if can AI deliver benefits to the enterprise. This article was originally published in Network Computing.
Artificial intelligence (AI) is in vogue. It’s almost impossible to read an article in any media outlet that doesn’t mention AI and the possibility it will reshape the world in which we live. In fact, according to research conducted by AT&T, AI has the potential to double GDP growth across geographies by 2035. Consumers are already interacting with a variety of low-level AI assistants, such as Siri, Cortana, and Alexa.
With respect to the telecom sector, AI – supported by machine learning (ML) – is fundamental to controlling and operating communications networks of the future. With AI, future networks will be more predictive and intelligent. They will be programmed to automatically make recommendations, implement policies and respond to changes instantly. However, it is essential to understand the characteristics of AI in telecom. Otherwise, it is likely to become another overused, overhyped, and underwhelming tech term that fails to deliver.
With AI, future networks will be more predictive and intelligent. They will be programmed to automatically make recommendations, implement policies and respond to changes instantly. However, it is essential to understand the characteristics of AI in telecom. Otherwise, it is likely to become another overused, overhyped, and underwhelming tech term that fails to deliver.
Defining AI
Talk to people in the telecom industry, and each one will give a different answer of what AI means to them. The fact of the matter is, AI does not have a single purpose or meaning. While AI in a basic sense can help describe what is currently happening or going to happen, a more mature level can identify why it is happening and take corrective action.
A clearer purpose of AI in the telecom industry makes it easier for businesses, decision-makers, and customers to determine how useful the technology will actually be, and how it could help them accomplish their goals. This presents an opportunity for those leading the charge to define certain AI standards and definitions.
Opportunity knocks
If you refer to the 2017 Gartner Hype Cycle, it is obvious that AI and ML are both somewhat over-hyped in the market. AI is already popular in ‘upcoming tech’ predictions with Gartner still marking it as 10 or more years away, and though ML may be coming sooner (two to five years, according to the report), it is currently in the peak of inflated expectations.
It is really the traditional telcos that will most benefit from AI in the future. Much of what service providers and telcos do involve manual tasks that take up the majority of a network technician’s day. AI will help these providers improve their operational efficiency and provide a much better and tailored customer experience.
It is clear that AI presents many business opportunities. Over-the-top (OTT) content and service providers such as Google, Facebook, Netflix, and Amazon are leveraging big data analytics and AI-based automation in their services to provide every customer, a highly customized experience.
That said, it is really the traditional telcos that will most benefit from AI in the future. Much of what service providers and telcos do involve manual tasks that take up the majority of a network technician’s day. AI will help these providers improve their operational efficiency and provide a much better and tailored customer experience.
While many are unsure what exactly AI will mean to a traditional telco, there is an opportunity in defining clear adoption strategies. Above all, the apprehension of giving total control of a complex network to a machine has led to a slower adoption through a crawl, walk, run strategy for implementing AI. By first providing intelligent insights and letting humans translate them into specific actions, then enhancing this with supervised changes done by machines. This will eventually lead to enough confidence that we let the machines ensure highly available networks all by themselves.
The most important factor at this point is having a clear vision of what a successful AI/ML adoption strategy looks like and defining which use cases will lead the market in providing the most ROI and OpEx savings.
Avoiding the AI cliché
As AI is more widely adopted, it will be built around next-generation transformational technologies, like the internet of things (IoT) and 5G. Many network providers are already taking steps to integrate AI with existing services. For example, AT&T is already incorporating AI and ML into customer interactions, as well as software-defined networking and even tasks like fleet management.
As they adopt AI, service providers will see many benefits which will also translate to better customer service and enhance end-user experiences. True AI will have to react faster than a human and make accurate decisions based on the situation and different variables, learning as it goes with no human input. The telcos that are able to define what a successful AI integration looks like will be the leaders of the industry, and now is the time for these companies to begin to figure out what works best for them. There is no doubt: AI will be one of the most disruptive technologies to enter the industry, and we are just starting to see a glimpse of all the benefits it will bring.
