Jeff Orr's Analyst Perspectives

A lot has been written about the definition of generative artificial intelligence (AI) and large language models (LLMs), though less has been written about the business considerations for an organization to evaluate adopting and implementing these technologies. And more importantly, does the technology align with the Office of the CIO objectives and the goals of the business? The value of generative AI software must be put into terms that all stakeholders can relate to. And organizations cannot afford to jump into a new technology without making a proper assessment of the risks and benefits. A pair of applications where generative AI software shows promise are improved collaboration to improve workflow and the addition of conversational computing interfaces.

New technology must also align with the objectives of IT leaders. Six objectives that we observe intersecting with collaboration and conversational computing and generative AI software are:

• Optimize operational efficiency to drive cost savings and streamline processes;
• Improve customer experience (CX) through self-service tools and 24/7 client access to personalized interactions;
• Enhance decision-making from insights and recommendations using larger datasets, including patterns that lead to more accurate decisions and better business outcomes;
• Foster a more collaborative work environment to bring about innovation and creativity, knowledge sharing, and employee engagement and satisfaction;
• Deliver on governance, risk management and compliance (GRC) initiatives by ensuring that ethical decisions are made, governing the performance of models, and mitigating risk to the organization;
• Prepare the organization for the future by embracing digital technology that can adapt to changing market conditions and scale into the future.
• The interest in generative AI technologies to further business goals cannot be understated. We assert that through 2025, one-quarter of organizations will deploy generative AI embedded in one or more software applications.

Generative AI software has several applications that highlight benefits from collaboration and conversational computing. Four use cases that benefit organizations are individual and team collaboration, conversational computing for internal resources and conversational computing for external audiences.

Generative AI is often touted as a creative partner for individual collaboration. New ideas and concepts are generated that expand the narrative conversation and potential of individual contributors. The existing workflow benefits from suggestions and recommendations that enhance the quality of tasks and storytelling. The efficiency of workflow increases as generative AI tools automate repetitive tasks and certain parts of the content creation process.

Benefits from the use of generative AI for team collaboration include facilitating brainstorming sessions and co-creating content. Software takes on the role of moderator, which enables teams to seed discussions with relevant starting points, generate diverse ideas, effectively combine inputs from the group and lead refinement exercises. Collaborative writing is possible by integrating inputs from the group while adhering to the organization’s style guidelines and incorporating references to existing assets.

Communicating with software interfaces in a more natural way, such as replicating human conversation, remains an opportunity for the organization and its workforce. Interactions using natural language processing (NLP) offer a voice-activated experience. Voice-first applications allow inclusive experiences for those with accessibility needs as well as provide multiple interfaces to simplify user interactions. Most applications are written with English as the only supported language. Generative AI expands the potential for multiple language inputs, generating translations to enable a greater diversity of workers to engage with systems. Finally, LLMs introduce NLP application environments with a larger dictionary of understanding than previously possible, which broadens the types of inquiries and range of routine tasks that can be automated.

Multiple use cases are emerging for LLM-powered conversational computing with audiences external to the organization as well. These range from a new generation of chatbots to personalized customer support. Early virtual assistant applications were limited by the ability to match user inquiries to knowledge base articles. Generative AI software expands the dictionary of inquiry phrases as well as support for multiple languages, which enables effective communications with customers that simulate natural language conversation. As a result, chatbots that offer 24/7 instant response and can respond to multiple, simultaneous inquiries are possible. Self-service knowledge expands to personalized customer support typified by understanding the customer sentiment and providing responses tailored to user preferences that enhance CX.

These are representative use cases for generative AI related to collaboration and conversational computing. We assert that by 2026, more than one-half of organizations will have invested in newly formed digital platforms to unify collaborative and conversational technology to simplify the work experience. Evaluate your business objectives and how technology can help achieve the organization’s goals.

The CIO and IT leaders are also responsible for the integration of technologies across the organization. Generative AI must be compatible with existing infrastructure and systems to increase the likelihood of user adoption. While generative AI provides a user interface that does not require specialized training in the data sciences, the deployment of AI and machine learning (ML) by line of business managers and organization-wide programs has revealed a skills gap for AI/ML talent that cannot be ignored. The success of generative AI software applications can make a difference in addressing the shortage of AI/ML talent and the opportunity to scale use of these technologies for future needs.

Organizations must establish governance frameworks and demonstrate a commitment to responsible and ethical use of AI technologies. Collaboration with other stakeholders, including legal, compliance and data privacy, is necessary to identify and mitigate potential risks to the organization.

Generative AI is susceptible to the same risks as AI/ML adoption, including bias in data and model training as well as privacy and data security. These systems learn from vast amounts of data, and if that data is biased or lacks diversity, it can amplify those biases in the generated output. Unfair content could harm an organization’s brand or tarnish its reputation with customers. IT leaders can evaluate data sources to ensure they are representative and implement processes to detect and mitigate bias during the training and deployment of generative AI models.

Generative AI software requires access to significant amounts of sensitive business data. The IT team has a responsibility to protect it and ensure compliance with data privacy regulations. If proper security measures are not put into place, the organization is at risk of unauthorized access, data breaches or misuse of the information. The Office of the CIO must assess the security considerations for all generative AI software before moving ahead with integration and deployment. If not already in place, the organization should establish data management and privacy policies and procedures. Similarly, generative AI requires ongoing evaluation of the model’s performance. Errors will occur and automated monitoring to maintain the reliability and accuracy of the model helps mitigate risk to the organization.

Technology has its benefits, and they can only be realized when addressing the business challenges and opportunities mapped to the organization’s business goals. Generative AI software integrated into collaboration and conversational computing applications can help optimize operational efficiency and improve CX. As with any new technology, a proper evaluation of its risks — from data bias to integration and support challenges — is necessary. Involve stakeholders from legal, information security and data privacy to understand risks to the organization. Verify compliance with existing data security and privacy policies. Mitigating risks during the evaluation process builds confidence across the organization and helps inform decision-making.

In today’s evolving business landscape, collaboration and conversational computing are critical for driving innovation, creativity, increasing productivity and achieving organizational goals. CIOs recognize the value of creating a work environment where teams can exchange information and work together towards shared objectives regardless of location.

Regards,

Jeff Orr