AI is real, and it’s here to stay. The adoption of AI in the last five years has more than doubled, and recent breakthroughs in generative AI are having unprecedented operational and organizational impacts across key industries.
Generative AI refers to algorithms like ChatGPT, Bard, GitHub Copilot, and Dall-E that can create new content including audio, code, images, text, simulations, and videos. The potential for this technology to fundamentally alter critical industries is compelling executives to consider how generative AI can enhance or disrupt their business operations, and more importantly, what to do about it.
A recent report by Goldman Sachs suggests that productivity gains from generative AI could raise global GDP by 7%. These gains will come in various forms including increasing efficiency through automating activities, enhancing customer service and experience, and creating novel revenue generating opportunities. However, along with these gains come significant risks. Organizations will need to undergo a massive transformation across each aspect of their business including people, processes, products, partners, and platforms.
To make things more interesting, generative AI technology is evolving at a breathtaking speed. Chat GPT was released in November 2022; four months later, OpenAI released a new language model called GPT-4 with improved capabilities including higher creativity, greater visual input, and longer context. This accelerated rate of change makes adopting and implementing Generative AI even more challenging. Organizations need to get a head start and remain nimble to adapt to a constantly and rapidly evolving technology.
AI is getting smarter. We should too.
For communication service providers (CSPs), generative AI can help improve network performance, reduce cost, and enhance the customer experience, offering a boon to industry leaders who implement early. Generative AI has the potential to significantly impact a CSP’s operations across the board including optimizing network performance, implementing predictive maintenance, enhancing customer service, personalizing marketing, augmenting network security, and more.
For example, generative AI algorithms such as generative adversarial networks (GANs) can generate synthetic data for training machine learning models, including those used in wireless and wireline networks. This enables more efficient and accurate network optimization, anomaly detection, and predictive maintenance. Leveraging these benefits can give CSPs a competitive edge and position them for future growth.
However, CSPs face multiple challenges and must address various questions before they can effectively harness the potential of generative AI. Some key challenges include:
- Use Cases: Identifying and prioritizing specific use cases that can be implemented using generative AI is a challenging task. It requires a careful approach to ensure that both readily available use-cases, which can be executed quickly to keep up with competition, and transformational use cases with long-term differentiating potential, are considered and invested in.
- Data: Generative AI relies on a significant amount of data to train the AI models. The quality of the data has a substantial impact on the outcomes. CSPs not only have to identify, collect, cleanse, and prepare datasets but also need to address privacy concerns and technical limitations associated with data acquisition.
- Infrastructure: CSPs must establish suitable computing infrastructure to support generative AI use cases. This includes setting up the necessary hardware and software systems. Additionally, integration of these new systems with existing ones is crucial. CSPs need to develop or hire resources with the required technical expertise and establish new partnerships to ensure a seamless integration process.
- Execution: CSPs need to establish a comprehensive governance framework that guides the implementation and adoption of generative AI. This framework should outline the functions, processes, and groups involved in developing and managing an effective and responsible AI program. It should address the evaluation and implementation process, identify and mitigate risks, and commit to ensuring AI fairness, transparency, and explainability.
JLA has developed a holistic approach to assist CSPs in crafting and executing their generative AI strategy. JLA’s approach consists of four interrelated workstreams – Outside-In Analysis, Inside-Out Analysis, Use-Case Development, and the Execution Blueprint.
- The Outside-In Analysis seeks to understand the market dynamics, features, applications, roadmap, and cost of using generative AI by identifying and analyzing use-cases that other industries are implementing. These insights are then evaluated based on potential value to your organization. This stage delivers a market map of generative AI capabilities, key drivers impacting implementation, industry best practices, and a list of potential AI use cases that may impact the communications industry.
- The Inside-Out Analysis selects targeted business activities across various functions to map the people, processes, platforms, partners, products, and performance that will be impacted through generative AI. This stage establishes an “as-is” baseline through interviews with key executives, identifying the activities that serve as strong candidates for generative AI automation.
- The Use Case Prioritization stage identifies the most transformative use cases based on a comprehensive framework that considers critical parameters such as cost-benefit analysis, strategic alignment, and risks. Each use case is evaluated on key parameters, differentially weighted, and scored. From there, workshops with key stakeholders are conducted, allowing for feedback, refinement, and final use case identification.
- Finally, the Execution Blueprint stage profiles the finalized use cases, defining the infrastructure and tools required for automation, preparing data for training models, setting up cross-functional teams for execution, clarifying roles and responsibilities, outlining interdependencies, defining performance measurement criteria, establishing milestones, and creating a detailed timeline. Organizational teams are onboarded with detailed guidance and an execution roadmap. The Execution Blueprint offers a tangible way to monitor and measure progress, adjust the AI implementation plan as needed, and maintain consistent alignment with executives, investors, and employees.
Generative AI is here to disrupt – driving unprecedented productivity and fundamentally altering business models and ways of working. Its rapid adoption has created the potential to transform the communications industry; however, it has also generated uncertainty and ambiguity. JLA can help navigate these waters and make the artificial real.
This article was enhanced and improved with AI. Let JLA do the same for your services.
Sangit Rawlley is a Senior Advisor with JLA Advisors (www.jlaadvisors.io), a boutique firm offering our clients a comprehensive end-to-end set of services, from strategy development, technology architecture design and execution, to software operational excellence, with an emphasis on innovation. Sangit is also the co-founder of Aiberry (www.aiberry.com), an AI-powered mental health screening platform.