How Generative AI Reshapes the Business Landscape

Views have ranged from a prediction of an AI-dominated Skynet/Terminator world where humans barely exist, to a future similar to Disney’s Wall-E Sky Liner world where humans grow so dependent on technology that they are physically and mentally hapless. Is this just another piece of over-hyped technology? Signs point otherwise as G-AI applications have reached one million users faster than any other digital tool in modern history.

Tuck Associate Professor of Business Administration Alva Taylor is faculty director of the Tuck Glassmeyer/McNamee Center for Digital Strategies. In the MBA program, Taylor teaches Design Thinking for Strategic Innovation; Strategic Change in the Turbulent Digital Age; and the sprint course Generative AI and the Future of Work. Taylor is also faculty director for Growing an Established Diverse Business and for the Digital Excellence Program in Tuck Executive Education.

Generative AI is a catch-all term for deep-learning algorithms, also known as large language models, trained on large quantities of data and parameters to discern patterns and structures within data. Once trained, these models generate new outputs based on prompts (input questions) that mirror their training data. These out-puts can be anything from coherent and contextually relevant text to intricate pieces of music, graphics, or computer programs. What makes the models unique is that both the inputs and outputs are conversational and contextual in ways that mimic human expression and interaction. This feature allows for previously unobtainable ease of use, understanding, and feedback.

As we think about the business implications of G-AI, it is important to understand what it is not. The models and applications are not computational or fact-finding; they are simply predictive of the next relevant response or string of words, given the prompt. If asked, “What is 2+2?”, it responds with “4” not due to calculation but because the model trained on data where the common response is “4.” Moreover, it cannot say “no” unless programmed to do so. Hence, it might provide contextually appropriate responses, even if incorrect—often referred to as “hallucinations.”

Generative AI has the potential to transform many sectors of business. Already, marketing teams use it to create ads, email campaigns, and social media posts, and development teams use it in new product development to write software code. Other functions seeing early impact include customer service, where it is used to answer customer questions and resolve complaints; and operations, where it automates tasks and optimizes supply chains.

While the technology holds great promise, business leaders will face significant new challenges in implementing it. As managers harness these tools to improve their organizations, they must be mindful of the dangers of misuse and the increasing difficulty in recognizing nefarious applications. As an example, G-AI can create realistic synthetic media which can be both an innovative tool for content creation, but also a threat to information integrity in the form of “deepfakes.”

As we entrust more of our calculation and knowledge recall tasks to G-AI, our perception of intelligence is undergoing a seismic shift. The rise of G-AI will redefine the concept of being smart. It’s no longer about memory capacity or computational speed—areas where AI has us beat. Instead, intelligence will be defined by the ability to ask insightful questions, frame problems, make nuanced decisions, and motivate people.

In the health care sector, G-AI can sift through medical literature and patient data at lightning speed, offering potential diagnoses. However, it is the doctor’s role to ask the right questions, interpret the AI’s suggestions, and make the final call.

In the corporate world, G-AI can analyze market trends, predict consumer behavior, and even suggest strategic moves. However, the onus remains on the human leader to frame the strategic questions, interpret the AI’s predictions, and make decisions that align with the organization’s values and goals.

Patrick Wheeler is executive director of the Tuck Glassmeyer/McNamee Center for Digital Strategies.

Most organizations are woefully unprepared for this shift. Success lies in identifying, screening, and choosing talent based on these new criteria. Organizations that hire and train managers to be adept in those skills and alter their processes to reflect this shift in value will have an advantage in both value creation and long-term organizational success.

Tuck takes this paradigm shift seriously, integrating generative AI and its implications into the school’s courses, experiential learning opportunities, internal training, and cross-Dartmouth linkages on AI activities. Tuck’s Center for Digital Strategies has been one of the leaders of these efforts, where executive director Patrick Wheeler developed and led student, faculty, and administrative “generative AI 101” sessions to provide different constituents with foundational AI knowledge. Professor Taylor, as faculty director of the Center, developed and taught a three session Sprint Course on Generative AI and the Future of Work this spring.

Whatever the future of generative AI holds, we should remember two points. First, the systems are continually getting better, meaning many of the criticisms of system capabilities and limitations will soon be moot. Second, and most important, generative AI done well is not a replacement for human capital, but a tool to free up individuals, managers, and organizations to focus more of their efforts on high-value creation activities.

Tuck intends to be at the forefront of helping shape leaders that can guide the technology’s use and development in a positive direction.

This opinion piece originally appeared in print in the summer 2023 issue of Tuck Today magazine.

Tuck School of Business

Please visit the firm link to site