Time For Telecom Execs To Learn About AI

We hear about them all the time – the wondrous promises of an AI-infused future. In your home, your car, your factory, your baseball team. Analytics, powered by AI, are helping make business, and life, better. As a telecom executive, you’re hearing all this, and you’re also hearing requests for more investment, and for patience, while you wait for the magic to happen.

Where exactly is that magic? And how exactly does it happen? And what can you do to help it along?

Those are some questions you’ve been asking, or are about to ask. As a response, here’s a short guide to help you find the answers, and to help you formulate your next sets of questions.

At its most basic, artificial intelligence works by making decisions: if the answer is this, go here; if the answer is that, go there. In operation it follows extremely complex decision trees, supported by blinding levels of processor speed. Along the way, it draws on supporting knowledge from data stores, model algorithms and other sources. Also, importantly, it is able to learn from experience by newly acquired information to its base of knowledge.

Serious academic research began in the 1950s, with computers learning. among other things, how to play checkers. In the 1960s, government research kicked in with funding by the US Department of Defense. Commercial ventures began to take hold in the 1980s, when an industry of “expert systems” grew out from numerous startups. But the hoped-for profitability didn’t materialize.

What moved AI back into popularity in the late 1990s was the fast-increasing power and sophistication of parallel-processing computers and the software to run them.

Cognitive Technologies

Today AI is still considered to be in the early stages of realizing its potential, but it, along with the related computer-processor power, is advancing quickly. Professional services firm Deloitte views AI as a growing set of cognitive technologies. Deloitte then groups these technologies according to how they support the needs of business:

Robotics and cognitive automation – These include Robotic Process Automation bots that replicate human actions and judgments. They can free up workers from performing repetitive, mundane tasks, while promoting process consistency and quality.

Cognitive insights – These include machine learning (ML) and natural language processing (NLP), and help derive patterns from large, complex data sets. Used with embedded sensors and cameras, these technologies can provide tracking and reporting of information in real time.

Cognitive engagement – These include intelligent agents, or chatbots, that can perform a range of services from answering customer queries to providing technical support to employees.

Telecoms Leading

Deloitte also points out that telecom companies are finding themselves on the cutting edge of AI implementation, because they are “dealing with users who are increasingly willing to change suppliers, intensifying familiar problems such as engaging customers and managing churn. They also face vexing infrastructure management problems as new networks, both wired and wireless, increase in complexity.”

All of this is true, but that’s not stopping leading telecoms from putting AI to use for competitive differentiation. Customer service chatbots and voice-recognition services are becoming commonplace, as is the use of predictive maintenance to forestall potential outages.

According to AI authority TechEmergence, these services are in play by AT&T, Verizon and others, as well as, for instance, the analysis of drone-captured video data for infrastructure maintenance and voice-activated television navigation.

Verizon offers an “out of the box” AI platform to B2B customers that can be used for personalized marketing campaigns and other advertising and sales functions. One research team at Comcast uses machine-vision analysis to find what types of content appeal most to customers. And DISH Network announced a collaboration with Amazon to let customers control their DVRs through Amazon’s Echo or Echo Dot Alexa devices.

Needed: Technology…And Skills

All experts agree that applications such as these represent nothing more than a starting point on the journey to fuller maturation of AI. Experts agree, too, that success won’t come from technology alone, but will require applications of human skills, not just at the data scientist level, but in executives as well. And these skills can be hard to find.

“Communications service providers (CSPs) need technically trained people who know how to handle data, create data models and algorithms, and implement and maintain AI and machine learning applications. However, finding these people is not easy, especially when data scientists and AI/ML specialists are in such high demand and there are so few workers available with the necessary skills,” says Aaron Boasman-Patel, writing in TM Forum Inform.

A recent multi-industry survey by Infosys found that 67 percent of telecoms and CSPs are finding it difficult to attract qualified staffers for AI integration projects.

That same report pointed out that business leaders in all industries also will benefit from AI training.

“C-level executives likewise called out training the leadership team on AI as a top priority—47 percent of business leaders put leadership training in their top three priorities compared to 40 percent who put employee training in their top three priorities.”

Training in AI-related fields will be essential to success for telecoms, as it will in all industries. As for technology, the best solutions will be those that facilitate human interaction with the models, menus, and development tools of AI applications. User-friendly tools such as quick-start process templates and low-code/no-code development environments, all delivered via high-performance cloud technologies, will help data scientists, executives and others make the most of their AI skill sets.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>