Digital transformation must become a core competency for organizations, and this is a key piece of advice for CIOs and IT leaders.

Strategic priorities change significantly every two years or less, from growth in 2018, to new crown epidemics and telecommuting in 2020, to hybrid office models and financial constraint issues in 2022.

The impact of generative AI, including ChatGPT and other large-scale language models, will be a major transformation driver in 2024.

As CIOs begin to prepare for 2024 budgets and digital transformation priorities, it will be necessary to develop a strategy to find opportunities to improve the business model, look at near-term operational impacts, prioritize projects where employees should test the waters, and develop a risk mitigation plan related to AI.

But with all this excitement and hype, it's easy for employees to invest their time in AI tools that will compromise confidential data, or for managers to choose shadow AI tools that haven't been vetted for security, data governance and other vendor compliance. The bigger challenge is to develop a realistic strategy and respond to the "impossible dreamers". In this case, the "impossible dreamer" is a business leader who "wants to get to the top" and a hell of a business executive.

Transformation priorities should be fundamentally linked to business priorities and what the organization wants to achieve," says Abhijit Mazumder, CIO, Tata Consultancy Services. In most organizations, leadership is equally focused on growth and operational efficiency, but without forgetting to prioritize resilience, cybersecurity and technical debt elimination programs."

Here are just a few of the drivers of generative AI that CIOs need to consider when setting their digital transformation priorities.

Develop a game-changing strategy for large-scale language modeling

How generative AI and large languages will impact every industry, for example:

Accelerating drug discovery by leveraging the intelligence that comes from unstructured data Enabling front-line manufacturing assembly workers to solve problems faster and more reliably Enabling healthcare providers to offer patients personalized solutions to health issues Assisting in the development of new insurance, banking and other financial service products based on customer conversations Transforming education by providing teachers with new ways to improve students' creative thinking, collaboration and problem-solving skills Now, CIOs and CTOs must not only be creative and do more with less, but also make thoughtful investments to outperform their competitors, who may be delaying or cutting back on their own transformation programs," said Jeremiah Stone, CTO, SnapLogic. Prioritize transformation initiatives that create new revenue streams, advance technology adoption, or can reduce technology debt, and especially consider the opportunities presented by generative AI."

CIOs may recognize that transformation initiatives of this magnitude are multi-year-long plans that require assessing the capabilities of large language models, experimenting, and finding customer offerings that are minimally viable and sufficiently secure. But failing to strategize at all can lead to confusion, and one of the key mistakes IT leaders can make when attending board meetings is failing to develop a plan at all for world-changing emerging technologies such as generative AI.

Cleaning and preparing data for private large-scale language models

Generative AI will increase the importance and value of unstructured data in the enterprise, including documents, videos, and content stored in learning management systems. Even if organizations are not yet ready to leverage generative AI to transform their industries and businesses, proactive transformation leaders will take steps to centralize, cleanse, and prepare unstructured data for use in large language models.

Kjell Carlsson, Head of Data Science Strategy and Evangelism at Domino, said, "With a strong demand from users across the organization to make generative AI functionality part of their daily activities, the top priority for CIOs, CTOs, and CDOs is to enable secure and scalable access to an increasing number of generative AI models, and for data science teams to develop and implement large-scale language models tailored to organizational data and use cases."

There are now 14 large language models out there outside of ChatGPT, and if you have large datasets, you can customize proprietary large language models using platforms like Databricks Dolly, Meta Llama, and OpenAI, or build your own large language model from scratch.

Customizing and developing large language models requires a strong business case, technical expertise, and funding.Peter Pezaris, Chief Design and Strategy Officer at New Relic, said, "The cost of training large language models can be extremely high and the outputs are not yet perfect, so leaders should prioritize investing in solutions that help to monitor the cost of usage and improve query solutions that improve the quality of results."

Increasing Efficiency Through Improved Customer Support

McKinsey predicted back in 2020 that AI could create $1 trillion in value annually, with customer support being a major opportunity. Today, thanks to generative AI, that opportunity has become even bigger, especially as CIOs centralize unstructured data in large language models and enable service agents to ask and answer customer questions.

Justin Rodenbostel, Senior Vice President at SPR, said, "Look for opportunities to leverage GPT-4 and large language models to optimize activities such as customer support, especially in automating tasks and analyzing large amounts of unstructured data."

Improving customer support is a fast track to providing a short-term return on investment through large language models and AI search capabilities. Large-scale language models require centralization of the enterprise's unstructured data, including data embedded in CRMs, file systems and other SaaS tools. Once IT centralizes this data and implements large-scale language models, there is also the potential to improve areas such as sales lead conversion and HR onboarding processes.

Gordon Allott, president and CEO of GetK3, said, "Organizations have been populating SharePoint and other systems with data for decades, and by cleansing that data and using large-scale language models, it's actually likely to be very valuable."

Reducing Risk by Communicating Around Large Language Models

There are more than 100 tools in the generative AI space, covering categories such as testing, images, video, code, voice, and more. So what's stopping employees from trying out a tool and pasting proprietary or other confidential information into the content of their prompts?

Rodenbostel suggests, "Leaders must ensure that their teams only use these tools in approved and appropriate ways by researching and developing acceptable use policies."

There are three departments that the CIO must work with the CHRO and also the CISO to collaborate, communicate policies and create a governance model that supports smart experimentation. First, the CIO should assess how ChatGPT and other generative AI will impact coding and software development. it is important for IT to lead by example and be clear about where, how to experiment, and when not to use tools or proprietary datasets.

Marketing is the second department to focus on. marketers can use ChatGPT and other generative AI in content creation, lead generation, email marketing, and more than a dozen common marketing practices. there are already more than 11,000 marketing technology solutions available, so there are plenty of opportunities to experiment when testing SaaS with new large-scale language modeling capabilities and make unintentional mistakes.

CIOs at leading organizations are creating a registry to incorporate new generative AI use cases, defining a process for reviewing methodologies, and centrally managing the impact of AI experiments.

Re-evaluating decision-making processes and authorizations

One other important area to consider is how generative AI will affect the decision-making process and the future of work.

Over the past decade, many businesses have aimed to become data-driven organizations by democratizing access to data, training more business people in data science for all, and instilling proactive data governance practices. Generative AI has unleashed new capabilities that allow leaders to prompt and get answers quickly, but timeliness, accuracy, and bias are key issues for many LL.

Putting humans at the center of AI and building strong frameworks around data use and model interpretability will go a long way toward reducing bias in these models and ensuring that all AI outputs are ethical and responsible," said Erik Voight, vice president of enterprise solutions at Appen. The reality is that AI models cannot replace humans when it comes to critical decision-making and should be used as a complement rather than allowing them to take over completely."

CIOs should seek a balanced approach to prioritizing generative AI initiatives, including defining governance, identifying short-term efficiencies, and pursuing opportunities for long-term transformation.

Digital transformation must become a core competency for organizations, and this is a key piece of advice for CIOs and IT leaders.

Strategic priorities change significantly every two years or less, from growth in 2018, to new crown epidemics and telecommuting in 2020, to hybrid office models and financial constraint issues in 2022.

The impact of generative AI, including ChatGPT and other large-scale language models, will be a major transformation driver in 2024.

As CIOs begin to prepare for 2024 budgets and digital transformation priorities, it will be necessary to develop a strategy to find opportunities to improve the business model, look at near-term operational impacts, prioritize projects where employees should test the waters, and develop a risk mitigation plan related to AI.

But with all this excitement and hype, it's easy for employees to invest their time in AI tools that will compromise confidential data, or for managers to choose shadow AI tools that haven't been vetted for security, data governance and other vendor compliance. The bigger challenge is to develop a realistic strategy and respond to the "impossible dreamers". In this case, the "impossible dreamer" is a business leader who "wants to get to the top" and a hell of a business executive.

Transformation priorities should be fundamentally linked to business priorities and what the organization wants to achieve," says Abhijit Mazumder, CIO, Tata Consultancy Services. In most organizations, leadership is equally focused on growth and operational efficiency, but without forgetting to prioritize resilience, cybersecurity and technical debt elimination programs."

Here are just a few of the drivers of generative AI that CIOs need to consider when setting their digital transformation priorities.

Develop a game-changing strategy for large-scale language modeling

How generative AI and large languages will impact every industry, for example:

Accelerating drug discovery by leveraging the intelligence that comes from unstructured data Enabling front-line manufacturing assembly workers to solve problems faster and more reliably Enabling healthcare providers to offer patients personalized solutions to health issues Assisting in the development of new insurance, banking and other financial service products based on customer conversations Transforming education by providing teachers with new ways to improve students' creative thinking, collaboration and problem-solving skills Now, CIOs and CTOs must not only be creative and do more with less, but also make thoughtful investments to outperform their competitors, who may be delaying or cutting back on their own transformation programs," said Jeremiah Stone, CTO, SnapLogic. Prioritize transformation initiatives that create new revenue streams, advance technology adoption, or can reduce technology debt, and especially consider the opportunities presented by generative AI."

CIOs may recognize that transformation initiatives of this magnitude are multi-year-long plans that require assessing the capabilities of large language models, experimenting, and finding customer offerings that are minimally viable and sufficiently secure. But failing to strategize at all can lead to confusion, and one of the key mistakes IT leaders can make when attending board meetings is failing to develop a plan at all for world-changing emerging technologies such as generative AI.

Cleaning and preparing data for private large-scale language models

Generative AI will increase the importance and value of unstructured data in the enterprise, including documents, videos, and content stored in learning management systems. Even if organizations are not yet ready to leverage generative AI to transform their industries and businesses, proactive transformation leaders will take steps to centralize, cleanse, and prepare unstructured data for use in large language models.

Kjell Carlsson, Head of Data Science Strategy and Evangelism at Domino, said, "With a strong demand from users across the organization to make generative AI functionality part of their daily activities, the top priority for CIOs, CTOs, and CDOs is to enable secure and scalable access to an increasing number of generative AI models, and for data science teams to develop and implement large-scale language models tailored to organizational data and use cases."

There are now 14 large language models out there outside of ChatGPT, and if you have large datasets, you can customize proprietary large language models using platforms like Databricks Dolly, Meta Llama, and OpenAI, or build your own large language model from scratch.

Customizing and developing large language models requires a strong business case, technical expertise, and funding.Peter Pezaris, Chief Design and Strategy Officer at New Relic, said, "The cost of training large language models can be extremely high and the outputs are not yet perfect, so leaders should prioritize investing in solutions that help to monitor the cost of usage and improve query solutions that improve the quality of results."

Increasing Efficiency Through Improved Customer Support

McKinsey predicted back in 2020 that AI could create $1 trillion in value annually, with customer support being a major opportunity. Today, thanks to generative AI, that opportunity has become even bigger, especially as CIOs centralize unstructured data in large language models and enable service agents to ask and answer customer questions.

Justin Rodenbostel, Senior Vice President at SPR, said, "Look for opportunities to leverage GPT-4 and large language models to optimize activities such as customer support, especially in automating tasks and analyzing large amounts of unstructured data."

Improving customer support is a fast track to providing a short-term return on investment through large language models and AI search capabilities. Large-scale language models require centralization of the enterprise's unstructured data, including data embedded in CRMs, file systems and other SaaS tools. Once IT centralizes this data and implements large-scale language models, there is also the potential to improve areas such as sales lead conversion and HR onboarding processes.

Gordon Allott, president and CEO of GetK3, said, "Organizations have been populating SharePoint and other systems with data for decades, and by cleansing that data and using large-scale language models, it's actually likely to be very valuable."

Reducing Risk by Communicating Around Large Language Models

There are more than 100 tools in the generative AI space, covering categories such as testing, images, video, code, voice, and more. So what's stopping employees from trying out a tool and pasting proprietary or other confidential information into the content of their prompts?

Rodenbostel suggests, "Leaders must ensure that their teams only use these tools in approved and appropriate ways by researching and developing acceptable use policies."

There are three departments that the CIO must work with the CHRO and also the CISO to collaborate, communicate policies and create a governance model that supports smart experimentation. First, the CIO should assess how ChatGPT and other generative AI will impact coding and software development. it is important for IT to lead by example and be clear about where, how to experiment, and when not to use tools or proprietary datasets.

Marketing is the second department to focus on. marketers can use ChatGPT and other generative AI in content creation, lead generation, email marketing, and more than a dozen common marketing practices. there are already more than 11,000 marketing technology solutions available, so there are plenty of opportunities to experiment when testing SaaS with new large-scale language modeling capabilities and make unintentional mistakes.

CIOs at leading organizations are creating a registry to incorporate new generative AI use cases, defining a process for reviewing methodologies, and centrally managing the impact of AI experiments.

Re-evaluating decision-making processes and authorizations

One other important area to consider is how generative AI will affect the decision-making process and the future of work.

Over the past decade, many businesses have aimed to become data-driven organizations by democratizing access to data, training more business people in data science for all, and instilling proactive data governance practices. Generative AI has unleashed new capabilities that allow leaders to prompt and get answers quickly, but timeliness, accuracy, and bias are key issues for many LL.

Putting humans at the center of AI and building strong frameworks around data use and model interpretability will go a long way toward reducing bias in these models and ensuring that all AI outputs are ethical and responsible," said Erik Voight, vice president of enterprise solutions at Appen. The reality is that AI models cannot replace humans when it comes to critical decision-making and should be used as a complement rather than allowing them to take over completely."

CIOs should seek a balanced approach to prioritizing generative AI initiatives, including defining governance, identifying short-term efficiencies, and pursuing opportunities for long-term transformation.