The arrival of Generative AI in the latter half of 2022 and early 2023 sparked widespread fascination, marking one of the most celebrated technology debuts in recent memory. Its adoption has been swift and widespread, with staggering user numbers. For instance, OpenAI's ChatGPT garnered 100 million users within just 60 days of its release to the public, surpassing TikTok's milestone in nine months. Midjourney's image generator boasts approximately 16 million users, while Dall-E 2 attracts 1.5 million daily users. In July alone, Google's Bard chatbot racked up 10 million page views.

This exponential growth in Generative AI usage is propelled by several converging factors. Firstly, advancements in specialized AI hardware, particularly AI chips used for model training, have facilitated the development of more sophisticated models like large language models (LLMs). These models have become mainstream, offering an intuitive user experience that even non-technical individuals can engage with.

Moreover, this surge in interest has ignited a frenzy among investors, with substantial investments flowing into startups with Generative AI technology at their core. Investors are banking on the emergence of a new era in business technology, where automated insights, self-reviewing contracts, and continuous content generation redefine brand engagement strategies.

While discussions abound regarding the potential job displacement caused by AI, there's little evidence to suggest that business leaders intend to automate knowledge-based roles on a large scale. A survey of leaders reveals that the primary motivations for deploying generative AI are to enhance content quality, gain competitive advantage, and augment employee expertise. Interestingly, reducing headcount ranks low on the list of priorities. It appears more likely that AI will liberate workers from mundane, repetitive tasks, allowing them to channel their efforts into more creative endeavors.

The prevailing notion is that AI is not just on the horizon; for many, it's already here. However, savvy business leaders understand that growth cannot be achieved by simply cutting costs or minimizing risks. Therefore, the most effective use of generative AI will not involve replacing human workers but rather empowering them with tools to boost productivity, knowledge, and creativity. This, in turn, fosters innovation within the enterprise.

Executives face mounting pressure to expedite this transition and outpace their rivals. A survey reveals that 64% of CEOs feel significant pressure from investors, creditors, and lenders to accelerate the adoption of generative AI. However, these leaders also recognize the importance of aligning AI implementation with genuine business needs. Simply integrating generative AI into every process without strategic forethought is unlikely to yield substantial benefits. Instead, businesses stand to gain more from a targeted approach that leverages the unique capabilities of generative AI to solve existing challenges and distinguish themselves in the market. This strategic approach characterizes the actions of innovative enterprises today.

The true potential of generative AI lies in its ability to revolutionize business functions, drive cost reductions, disrupt product and service cycles, and achieve unprecedented process efficiencies. To unlock this value, business leaders are advised to take an evolutionary approach to their enterprise data and technology strategy.

Becoming an AI-driven organization requires meticulous attention to detail and a focus on maintaining robust systems and algorithms. Just as a rocket needs a launch pad and flight controls to reach its destination, generative AI tools require infrastructure and control systems to thrive in enterprise settings. Fortunately, many of the skills and practices developed in recent years while building data analytics and machine learning capabilities are transferrable to generative AI, albeit with some adjustments.

Generative AI typically demands vast amounts of data stored on high-performance computing clusters equipped with graphics processing units (GPUs). Given that few businesses possess this infrastructure, most opt to access it through service providers. Through application programming interfaces (APIs), engineers can seamlessly integrate generative AI capabilities into existing software without the need for significant infrastructure overhauls. While AI vendors prioritize user-friendliness, it remains crucial for enterprises to remain mindful of these engineering requirements.

Moreover, it is imperative to choose use cases wisely. Generative AI can be harnessed to reduce costs, accelerate processes, simplify complexities, enhance customer engagement, foster innovation, and instill trust. While the specific applications of generative AI vary across businesses, targeting projects that yield improvements in specific areas is a prudent starting point.

Here are some additional insights from businesses that have already embraced generative AI:

Businesses must ensure that their data is well-structured and accessible to AI applications for model training and next-generation use cases.

This lesson was learned by Enbridge, the largest natural gas utility in North America. Several years ago, during an ambitious cloud migration initiative, the company did not anticipate pioneering new uses for generative AI. Instead, its primary objectives were to modernize infrastructure and reduce technical debt by downsizing on-premises data centers. However, along the way, Enbridge established a centralized data repository that aggregates data from various sources across the enterprise, replacing hundreds of disparate data silos.

With the advent of generative AI, Enbridge recognized that this centralized data repository was the ideal foundation for driving AI-driven efficiencies. The technology team introduced a generative AI-based copilot tool to aid developers in rapidly and efficiently coding. Additionally, office staff were equipped with a copilot tool to navigate productivity applications.

Joseph Gollapalli, Director of Cloud, IT Ops, and Data at Enbridge, emphasized the goal: "to accelerate our delivery and drive innovation and efficiency. These AI solutions have the potential to enhance our operations, improve safety, elevate the customer experience, and enhance our environmental performance."

In today's landscape, governance is more crucial than ever for the successful scaling of AI initiatives. A robust governance framework should articulate the business's vision, pinpoint potential risks and capability gaps, and validate performance. These considerations not only safeguard the business but also facilitate the scaling of projects beyond the proof-of-concept stage.

At CarMax, the largest used car retailer in the United States, the effective utilization of generative AI hinges on a systematic, organization-wide approach that harnesses the technology's power while implementing guardrails to ensure employees' effective use. One notable application at CarMax is a tool that incorporates AI-generated content into vehicle research pages. These pages aggregate information from thousands of actual customer reviews, enabling shoppers to quickly access insights from other buyers.

Shamim Mohammad, Executive Vice President and Chief Information and Technology Officer at CarMax, emphasizes the importance of controlled implementation for generating the most business value. CarMax places a high priority on governance, recognizing its significance in scaling generative AI initiatives. The company has established an AI governance team tasked with ensuring appropriate AI usage across the organization. Importantly, this team doesn't solely reject new use cases; it aims to standardize model training and usage to scale impactful applications across the enterprise, beyond technology or product teams.

"We've accomplished remarkable feats through machine learning and AI," says Mohammad. "My focus now is on ensuring responsible usage and aligning our deployments with our core values as a company."

Generative AI has reshaped the copyright landscape, enabling the creation of images, videos, text, and audio with ease. However, challenges arise when models are trained on third-party content, potentially rendering AI-generated content ineligible for copyright protection. Additionally, training models on copyrighted web content may pose legal risks, including intellectual property infringement.

Yet, solutions exist to address these challenges. Shutterstock, a content provider, exemplifies how generative AI can be used while respecting copyright holders' rights and enabling the commercial use of AI-generated content. Shutterstock recently introduced an image-generating tool trained on images by third-party artists, with each artist's consent and compensation. By licensing content as data, Shutterstock enhances legal protections for end users while honoring the contributions of artists and creators.

"Content creation spans industries, from CEOs to retail workers," says Michael Francello, Director of Innovation at Shutterstock. "Recognizing the explosive need for content creation, we seized the opportunity to view our content as data to train generative AI models. It's about safeguarding our business's core while respecting the contributions of artists and creators."

For years, the crawl-walk-run-fly approach has proven effective for enterprises scaling up their service offerings. Generative AI follows a similar trajectory. In the crawl stage, applications are ad hoc and demand manual effort. Progressing to the walk stage, processes become more defined and automated. In the run stage, use cases are standardized and permeate the enterprise. When it's time to soar, organizations leverage previous work to embrace next-generation capabilities.

This approach guided chemical company Eastman in developing generative AI-based internal services. With a history of utilizing data and analytics in an industry not typically associated with it, Eastman developed an advanced intelligence service to predict thermal stability in industrial processes. Expanding on this, the company experimented with how generative AI could enhance sales processes. By creating an AI-enabled tool to extract insights from sales call notes, Eastman unlocked valuable intelligence previously overlooked.

Aldo Noseda, Eastman's chief information officer, remarks, "It allows us, a chemical company, to bring a digital service layer to differentiate ourselves and gain a competitive advantage."

Given generative AI's rapid progress, applying this framework to new enterprise use cases is prudent. Let proof-of-concept projects evolve into standardized practices that become ingrained in the enterprise's operations. Once this maturity is achieved, possibilities become limitless.

Looking ahead, businesses may find it easier to harness generative AI benefits within their industries as models trained on specific data emerge. While current tools rely on general-purpose data, the next generation of large language models (LLMs) is poised to be more specialized. This trend is already evident with tools like NVIDIA's BioNeMo, Google's Contact Center AI, and BloombergGPT, tailored for specific sectors like biotech, customer service, and finance, respectively. As demand grows for sector-specific models, enterprises are likely to seek out private LLMs, offering competitive advantages with proprietary, purpose-specific models trained on private data. This represents the next phase in the generative AI journey.

The motivational poster, once a corporate cliché now turned meme, may soon regain its significance as a rallying call for enterprises: "We’re only limited by our imagination."

Traditionally, teams and organizations faced constraints—lack of data, skeptical leadership, or the dreaded "That won’t move the needle." However, in the realm of generative AI, imagination reigns supreme. With the ability to generate continuous content, uncover operational efficiencies, or analyze regulatory filings in minutes, the question shifts to: What do you want to discover?

Enterprises embracing generative AI will value the skill of asking better questions. This shift may herald the rise of a new breed of leader, one fueled by creativity rather than sheer data-driven insights. While the past decades rewarded leaders for data-driven decisions, the future may favor those with imaginative leaps. Just as an image generator responds to a dull prompt with a lackluster image, unimaginative use of generative AI yields limited impact. As businesses strive to stand out, leaders who innovate with generative AI may surpass peers fixated on conventional data.

This doesn’t diminish the importance of data-driven decision-making. In fact, it becomes more critical as generative AI expands the range of accessible data, from natural language text files to machine logs. Leaders must harness generative AI to extract insights from overlooked data sources, asking intelligent questions at the speed of thought.

Yet, the true extent of generative AI's impact remains to be fully realized. While its seismic potential is undeniable, the precise areas of transformation remain uncertain.