The predictions for AI use cases have been prolific and wide-ranging in recent years. From humanoid robots to predictive analytics for legal institutions, hedge funds, and more, there has rarely been more excitement generated by a technology than the current buzz emanating from AI software.
Application of this technology extends to mobile and telecommunications too. Here, it has become an important next step in helping operators’ transition from Communications Service Providers into more advanced Digital Service Providers that can predict their customers wants and needs.
AI is empowering service providers with a range of new capabilities such as deep learning, natural language processing, and cognitive computing to create a digital interface that will essentially deal directly with human beings, addressing and resolving customer service issues. Sound like science fiction, but it’s the new reality. AI is a huge catalyst for change, not just in telecoms, but almost every walk of life.
So, what exactly is artificial intelligence? Is it about creating robots powered by super computers, which outperform their human counterparts? Or is it grounded in less futuristic, albeit still important, applications and in more sedate data crunching and algorithms than walking-and-talking machines? It’s all of the above and more.
AI comes of age
AI can be about simulating human intelligence, incorporating traits such as reasoning, perception, problem solving and forward planning. At its crux, though, AI is about the development and enactment of methods of transforming vast amounts of complex, often unstructured data into intelligent insights.
The key elements of artificial intelligence – machine learning, cognitive computing, natural language processing, and sentiment analysis, combined with more effective real-time data management – make this possible.
It can establish rules and use algorithms to deliver accurate analytics and predictions. Importantly, this may also reveal ‘hidden’ insights in data that would have been missed if the process were to be conducted by a human.
In turn, for service providers across all industries, this makes it possible to take much more informed actions as a result of predicting what customers will need and when, making timely, relevant, and attractive offers to drive sales and further engagement.
Findings from data can be utilised to improve services, develop new tools and technologies, and drive production or business efficiencies, in addition to wide-ranging benefits we are yet to realise. Although the terms are often used interchangeably – or confused – this more specific application of AI can be more closely defined as ‘machine learning.’
The potential challenges ahead
Much has been made of the opportunities AI can deliver to businesses and consumers, but the technology has attracted some negative publicity, particularly in relation to the potential impact of AI and robotics on the job market.
However, as with other technological and industrial revolutions which have come before it, the disappearance of jobs in some sectors coupled with the introduction of new technologies will likely spur the creation of new (human) roles in other areas of business.
New training and skills will be needed for workforces in order to adapt their jobs to the new opportunities AI presents, requiring new educators. AI technology will need maintaining and new systems will need to be developed, necessitating individuals with knowledge and experience in this new field. In addition, the automation of traditionally repetitive, administrative office jobs would arguably allow for more creativity and boost workforce morale, an advantage for any industry.
Questions over how to regulate and control AI have also become a key topic, particularly after Facebook was forced to shut down an artificial intelligence program after it created its own language.
In this instance, two bots created a series of code words and nonsensical text strings for communicating tasks. Although more efficient than a full English sentence, the phrases could not be interpreted by human controllers.
There is not enough evidence to suggest that this unforeseen development poses a threat, but it’s certainly a development that needs closer monitoring and regulation for the future of this technology.
The ethics of AI and the role it will play in our lives continues to drive debate as a result. It has even split the opinions of two of Silicon Valley’s most esteemed CEOs. Recently, Facebook’s Mark Zuckerberg and Tesla’s Elon Musk found themselves embroiled in a very public row about the viability of AI, the benefits it can offer and the potential challenges it poses.
Zuckerberg was more optimistic focusing on the breakthroughs that have been made in healthcare and the development of self-driving cars. It’s no surprise Zuckerberg holds this view given that Facebook has invested so heavily in AI. And, despite their differences, both CEOs have accepted AI will play a crucial role in improving their businesses.
A new era of communications
Despite significant developments in AI over the past decade, full automation and computer super intelligence is still a while off yet. Where the technology has been making significant inroads, though, is in communications.
In the digital-first landscape of today, consumers are more demanding than ever. They expect always-on digital services and for engagement with service providers to be immediate, reliable, and on their terms through platforms of their choosing.
Until recently this engagement meant contacting a call centre, with delays and queues to connect to an operator. AI, and chatbots in particular, have revolutionised this space. Gartner recently predicted that by 2020, 85% of all customer interactions will be handled without a human agent.
Facebook launched its chatbot creation tools in 2016, allowing platform creators to build their own version of the technology and integrate these customer service tools into their business offering.
As of January 2017, a reported 45,000 developers were using Facebook’s Wit.ai chatbot-building tool to create chatbots for Facebook Messenger.
In addition to driving customer satisfaction through quicker, more convenient interaction with companies, the use of chatbots can also prove a revenue-booster: an estimated 36% of sales representative positions in the US could be automated, meaning annual savings from salaries of at least $15 billion. However, it is not only in communications generally where AI is proving its worth, but in the telecommunications industry more specifically.
The intelligent service provider
With the rise of disruptive new digital service providers (DSPs) and the continuing shift to data usage, mobile operators and communication service providers (CSPs) face threats to their traditional revenue streams.
Access to data services is now seen by consumers and businesses as a necessity, and, as the market has opened up to more competition – and more choice – consumers are demanding more from their service providers.
Although not solving all ills, AI can be seen as a force of differentiation that will empower service providers to drive value across their businesses. And not only in the customer interaction space but also in areas like network management and optimisation, and improving subscriber experiences through more accurate data visibility and analytics.
It is within this latter area that AI will play an essential role, ensuring telcos not only survive but thrive. The imminent arrival of next generation 5G technologies and networks, coupled with the rapidly expanding IoT will vastly increase the number of end-points a CSP/DSP must manage.
More consumer connected devices, a greater number of machine-to-machine (M2M) connections, and sensors embedded in infrastructure and vehicles will result in a dizzying amount of communication between technologies and information traversing networks.
This level of data is unmanageable for the human to process, yet an AI system can provide real time visibility and management of this information, delivering intelligent data analytics to improve processes.
AI can also extract data from one part of a business in order to place it in another, linking up sectors to allow them to ‘learn’ from each other.
In the case of data from consumer devices, AI-grounded analysis can also be used to improve services for subscribers. Customer behaviour and engagement data can be gathered and processed more rapidly and in greater volumes, used to influence the development of optimal pricing models and create new services.
The granular data insight garnered by AI goes beyond that capable of humans, meaning a deeper understanding and learning of competition data, such as BSS information, advertising and voice of the consumer type feedback.
Proactive not reactive
AI will not just allow for reactive management of customer services, but, due to the predictive capabilities of the technology, it’ll also support proactive customer care. Implementing intelligence and automation will enable operators to anticipate the needs of their customers in order to engage with them via the channel of their choosing and at the time most suited to their preferences. For traditional CSPs looking to revitalise their business in order to keep up with innovative DSPs, this kind of action could prove a vital differentiator.
Artificial intelligence will also deliver huge benefits to network management. Many telcos have already introduced network functions virtualisation (NFV) and software defined networking (SDN), as well as moving processes and applications to the cloud. Artificial intelligence can be harnessed to aid efficient traffic routing, as well as managing network traffic capacity. Faults on an operator’s network can quickly be identified, and problems rectified.
Data on factors such as capacity demands and user behaviour can be analysed and networks can be automatically configured in response. Again, the use of AI technologies means these processes will be proactive rather than reactive.
Machine learning systems can be taught to recognise patterns in data and information, and networks and applications can be adjusted and altered in order to solve any problems before they impact the consumer.
Widespread implementation of such systems is a while off yet, although the aforementioned chatbots are making strides when it comes to CSPs and their customer service offerings.
Indeed, many CSPs are ready for a full transformation towards AI implementations, whilst others are taking it at a slower pace. This said, in addition to developing and fine tuning AI systems, an operator – or any business considering implementing AI – must recognise the level of risk associated with this move, and ensure they implement strategies and business models to maximise opportunity whilst reducing risk.
Artificial intelligence can help boost efficiencies, aid customer engagement and services, and drive revenue as a result. However, any potential financial gain must be weighed against the cost of investing in an AI strategy.
This cost includes not only the AI tech itself, but the training which will be needed to ensure that workforces are trained to support implementation, and any new skills required added to the business through new hires.
This may include change management committees, which can help to manage an AI implementation process, including overseeing the cultural changes which are often experienced when a business makes such a dramatic move.
Finally, many traditional CSPs will likely have already experienced challenges caused by digital transformation in the industry. Building an AI system in-house could easily exacerbate any problems, so companies wishing to make this move should consider looking to external suppliers to instead create and help deploy any new system.
Even after 60-plus years, artificial intelligence is still a while off peak maturity. Equally, the implementation by any business of an AI strategy should not be a rushed process. The benefits artificial intelligence can deliver will be great, but these must always be costed against potential pitfalls.
Intelligent systems may play the key to success through digital transformation and delivering on consumer expectations, but this will only be realised through intelligent deployment and management models.
Sourced by Jonathan Kaftzan, head of Digital & Intelligence marketing at Amdocs