McKinsey: What Every CEO Should Know About Generative AI

Combining generative AI with all other technologies, work automation could add 0.5 to 3.4 percentage points annually to productivity growth. However, workers will need support in learning new skills, and some will change occupations. If worker transitions and other risks can be managed, generative AI could contribute substantively to economic growth and support a more sustainable, inclusive world. A company optimizes a foundation model for customer service conversations, fine-tuning it on high-quality customer chats and sector-specific Q&A. Operating in a sector with specialized terminology, the company introduces a generative AI customer-service bot to handle most inquiries, aiming for swift, brand-aligned responses.

Learn more about how we’re redefining the possibilities of AI and enterprise solutions! The company decided to create a tool that would help scientists understand the relationship between drug chemistry and the recorded microscopy outcomes to accelerate R&D efforts. Since such multimodal models are still in infancy, the company decided to train its own instead. To build what every ceo should know about generative ai the model, team members employed both real-world images that are used to train image-based foundational models and their large internal microscopy image data set. Supporting the new tool is a small cross-functional team focused on selecting the software provider and monitoring performance, which should include checking for intellectual property and security issues.

The solution, utilizing a foundation model accessed through an API, scans large documents and provides synthesized answers to RMs’ questions. Additional layers ensure a streamlined user experience, integration with company systems, and application of risk controls. The generative AI accelerates the RM’s analysis process, potentially capturing overlooked insights and improving job satisfaction.

Even though so many companies have already achieved real cost savings and revenue improvements with gen AI, these are still the early days of the technology. Once new models are deployed, telcos will need to monitor model outputs daily to ensure quality and accuracy do not waver as models learn and adapt their responses based on user queries and feedback. Large language model operations (LLMOps) is an emerging practice that aims to streamline the daily management and monitoring of gen AI models. A key component of LLMOps is a dedicated operations team to oversee all deployed gen AI models, continuously monitoring for issues and rapidly adapting solutions when needed, just as a network operations team might do for network performance. Organizations can start small now and build capability in this area as the field of LLMOps develops.

Momentum among workers for using gen AI tools is building

The company decided to introduce a generative AI customer-service bot to handle most customer requests. The goal was a swift response in a tone that matched the company brand and customer preferences. Some may see an opportunity to leapfrog the competition by reimagining how humans get work done with generative AI applications at their side.

Once the decision is made, there are technical pathways that AI experts can follow to execute the strategy, depending on the use case. With proper guardrails in place, generative AI can not only unlock novel use cases for businesses but also speed up, scale, or otherwise improve existing ones. A specially trained AI model could suggest upselling opportunities to a salesperson, but until now those were usually based only on static customer data obtained before the start of the call, such as demographics and purchasing patterns. A generative AI tool might suggest upselling opportunities to the salesperson in real time based on the actual content of the conversation, drawing from internal customer data, external market trends, and social media influencer data.

The economics and technical requirements to start are not prohibitive, while the downside of inaction could be quickly falling behind competitors. Some CEOs may decide that generative AI presents a transformative opportunity for their companies, offering a chance to reimagine everything from research and development to marketing and sales to customer operations. Once the decision is made, there are technical pathways that our AI experts can follow to execute the strategy, depending on the use case.

Generative AI is rapidly evolving: Organizations should seize the business imperative now

Our experience working with clients indicates the potential for telcos to achieve significant EBITDA impact with gen AI. However, generative AI’s impact is likely to most transform the work of higher-wage knowledge workers because of advances in the technical automation potential of their activities, which were previously considered to be relatively immune from automation (Exhibit 13). Generative AI’s ability to understand and use natural language for a variety of activities and tasks largely explains why automation potential has risen so steeply. Some 40 percent of the activities that workers perform in the economy require at least a median level of human understanding of natural language. You can foun additiona information about ai customer service and artificial intelligence and NLP. While we have estimated the potential direct impacts of generative AI on the R&D function, we did not attempt to estimate the technology’s potential to create entirely novel product categories.

• They could also have an impact on knowledge workers whose activities were not expected to shift as a result of these technologies until later in the future (see sidebar “About the research”).
• Within this new partnership, Deloitte Middle East will offer access to Moderator so enterprises can overcome common regional security and governance hurdles.
• Generative AI has the potential to revolutionize the entire customer operations function, improving the customer experience and agent productivity through digital self-service and enhancing and augmenting agent skills.
• In general, fine-tuning foundation models costs two to three times as much as building one or more software layers on top of an API.
• Customer service channels using cognitive chatbots could seamlessly answer complex queries in real time while taking into account privacy and fairness concerns, thereby revolutionizing efficiency while offering customers a human-like experience.

As a result, they possess impressive and unprecedented power and can perform very sophisticated functions, like passing the law bar exam and advanced sommelier test. A new wave of artificial intelligence systems, generative AI solutions, like Open AI’s ChatGPT, have tremendous potential to change how work gets done and even how life is lived. They have the potential to streamline business processes, reduce expenses and boost revenue. CEOs need to take them seriously as new solutions are quickly working their way from development labs into the corporate DNA.

While the combination of generative AI and no-code platforms offers many benefits, it’s crucial to approach this union with flexibility, seeing challenges as growth opportunities. Like any technology, no-code solutions have their own set of challenges and pitfalls. Thus, organizations should balance the rapid development of AI and no-code with strategic planning, testing and IT protocol adherence to mitigate risks and maximize benefits.

All of this is possible because generative AI chatbots are powered by foundation models, which are expansive neural networks trained on vast quantities of unstructured, unlabeled data in a variety of formats, such as text and audio. In contrast, previous generations of AI models were often “narrow,” meaning they could perform just one task, such as predicting customer churn. One foundation model, for example, can create an executive summary for a 20,000-word technical report on quantum computing, draft a go-to-market strategy for a tree-trimming business, and provide five different recipes for the ten ingredients in someone’s refrigerator.

For the purposes of this article, we focus on ways generative AI can enhance work rather than on how it can replace the role of humans.

The right partnerships with the right experts will help companies move quickly to create value from generative AI, though they will want to take care to prevent vendor lock-in and oversee possible third-party risks. To fully harness the potential of generative AI, companies must invest in evolving their operations and training their workforce. This includes developing technical competencies and ensuring employees are equipped to work effectively with AI-enhanced processes​​. Generative AI technology represents a quantum leap in not only processing power but also intelligence compared to previous versions of AI. One reason why today’s solutions are gaining so much attention is they work with much larger volumes of information (hundreds of billions of words) and larger data models (hundreds of billions of parameters) than previous AI systems.

In this example, research scientists in drug discovery at a pharmaceutical company had to decide which experiments to run next, based on microscopy images. They had a data set of millions of these images, containing a wealth of visual information on cell features that are relevant to drug discovery but difficult for a human to interpret. In this case, generative AI freed up service representatives to focus on higher-value and complex customer inquiries, improved representatives’ efficiency and job satisfaction, and increased service standards and customer satisfaction. The bot has access to all internal data on the customer and can “remember” earlier conversations (including phone calls), representing a step change over current customer chatbots. Structured data are tabular data (for example, organized in tables, databases, or spreadsheets) that can be used to train some machine learning models effectively.

Joint customers will also have access to Deloitte’s critical knowledge and expertise around successfully deploying the Moderator solution in business contexts. Armed with this industry-leading AI security and governance platform, organizations will receive a customized deployment methodology from Deloitte that seamlessly creates the foundation needed to successfully deploy generative AI. First, they can draft code based on context via input code or natural language, helping developers code more quickly and with reduced friction while enabling automatic translations and no- and low-code tools.

To maximize value, companies are increasingly fine-tuning pretrained generative AI models with their own data. This customization allows businesses to address their unique needs, unlocking new performance frontiers​​​​. Opportunities come and go quickly, so businesses need to be structured to respond rapidly. CEOs need to not only be aware of the need to create an agile enterprise but also put the mechanisms in place to create one. Artificial intelligence systems become a foundational piece in such transformations because they provide companies with much-needed flexibility.

In fact, more-experienced engineers appear to reap the greatest productivity benefits from the tools, with inexperienced developers seeing less impressive—and sometimes negative—results. A known risk is that the AI-generated code may contain vulnerabilities or other bugs, so software engineers must be involved to ensure the quality and security of the code (see the final section in this article for ways to mitigate risks). The first foundation models required high levels of investment to develop, given the substantial computational resources required to train them and the human effort required to refine them.

It took only a few months before the product became a significant threat to the long-established dominance of tech giants like Google, Microsoft and IBM. In conclusion, generative AI offers CEOs transformative opportunities but also carries risks. By prioritising alignment, developing strong business cases, and addressing ethical considerations, CEOs can effectively leverage generative AI to drive innovation and stay competitive in the ever-changing business landscape.

And contrary to the notion that the rise of no-code platforms and generative AI might render professional developers obsolete, these platforms present an opportunity. Freed from routine tasks, developers can channel their expertise into pioneering new technological frontiers, exploring untouched areas, and crafting the very tools and frameworks that might be the backbone of future no-code solutions. Generative AI is developing fast, and companies will have to balance pace and innovation with caution. The board’s role is to constructively challenge the management team to ensure this happens, keeping the organization at the forefront of this latest technological development yet intensely mindful of the risks. The questions posed here are not, of course, exhaustive, and more will arise as the technology progresses. Ultimately, board members hold responsibility for how generative AI is used in the companies they oversee, and the answers they receive should help them meet that responsibility wisely.

Traditional predictive models rely heavily on historical data, which can often hinder accurate predictions when faced with complex or rapidly changing scenarios. In simple terms, an organization can only predict with any sort of confidence level if something is going to happen if there is a history of the action happening in the past. Trying to predict in areas with no precedence isn’t recommended, and leaders should make full use of any available data to drive decisions and solve problems using what is known, rather than what is thought to be known. In this instance, generative AI can speed up an RM’s analysis process (from days to hours), improve job satisfaction, and potentially capture insights the RM might have otherwise overlooked.

Managing risks and security concerns in Generative AI involves implementing robust encryption protocols, regular software updates, and strict access controls to safeguard sensitive data and prevent misuse. Generative AI is a subset of artificial intelligence that focuses on teaching machines to mimic human like creativity. It involves algorithms capable of creating content, imitating patterns, and generating new data, such as images, texts, or even music. Unlike other AI systems, Generative AI operates autonomously to produce original and unique outputs.

In practice, that can mean, for example, prioritizing the use-case pipeline, identifying opportunities for reusability, setting key performance indicators to measure and track impact at the level of both use case and enterprise, and managing suppliers and risk. Gen AI represents the latest advance in AI, and it may arguably be one of the most important. The technology’s ability to analyze more and different types of data such as code, images, and text, and to create new content, enables new levels of personalization, performance, and customer engagement. With today’s capabilities, many use cases are already possible across network operations, customer service, marketing and sales, IT, and support functions. The potential of technological capabilities in a lab does not necessarily mean they can be immediately integrated into a solution that automates a specific work activity—developing such solutions takes time.

Recently, McKinsey published an insightful article, “What Every CEO Should Know About Generative AI,” shedding light on the transformative potential of this cutting-edge technology. GPT-4, for example, was released in March 2023, following the release of ChatGPT (GPT-3.5) in November 2022 and GPT-3 in 2020. In the world of business, time is of the essence, and the fast-paced nature of generative AI technology demands that companies move quickly to take advantage of it. To improve engineers’ productivity, the company is implementing an AI-based code-completion product that integrates with the software the engineers use to code. This allows engineers to write code descriptions in natural language, while the AI suggests several variants of code blocks that will satisfy the description.

While text-generating chatbots such as ChatGPT have been receiving outsize attention, generative AI can enable capabilities across a broad range of content, including images, video, audio, and computer code. And it can perform several functions in organizations, including classifying, editing, summarizing, answering questions, and drafting new content. Each of these actions has the potential to create value by changing how work gets done at the activity level across business functions and workflows. Finally, the survey findings suggest that gen AI has already begun to influence long-standing market dynamics.

CEOs are urged to explore generative AI, viewing it as essential rather than optional. It holds value across various use cases, with manageable economics and technical requirements. CEOs should collaborate with their teams to strategize its implementation, whether as a transformative force or through gradual scaling. Generative AI’s organizational impact often stems from existing software features, enhancing productivity for knowledge workers.

Fostering a culture of self-driven research and experimentation can also encourage employees to innovate processes and products that effectively incorporate these tools. As demonstrated in the use cases highlighted above, technical and talent needs vary widely depending on the nature of a given implementation—from using off-the-shelf solutions to building a foundation model from scratch. For example, to build a generative model, a company may need PhD-level machine learning experts; on the other hand, to develop generative AI tools using existing models and SaaS offerings, a data engineer and a software engineer may be sufficient to lead the effort. However, some characteristics set foundation models apart from previous generations of deep learning models. To start, they can be trained on extremely large and varied sets of unstructured data. For example, a type of foundation model called a large language model can be trained on vast amounts of text that is publicly available on the internet and covers many different topics.

Documenting code functionality for maintainability (which considers how easily code can be improved) can be completed in half the time, writing new code in nearly half the time, and optimizing existing code (called code refactoring) in nearly two-thirds the time. Our research found that marketing and sales leaders anticipated at least moderate impact from each gen AI use case we suggested. They were most enthusiastic about lead identification, marketing optimization, and personalized outreach.

We expect that agile (or bionic) models will remain the most effective and scalable in the long term, but with centralized IT and R&D departments staffed with experts who can train and customize LLMs. This centralization should ensure that employees who work with similar types of data have access to the same data sets. When data is siloed within individual departments—an all-too-common occurrence—companies will struggle to realize generative AI’s true potential.

We then estimated the potential annual value of these generative AI use cases if they were adopted across the entire economy. For use cases aimed at increasing revenue, such as some of those in sales and marketing, we estimated the economy-wide value generative AI could deliver by increasing the productivity of sales and marketing expenditures. Managers should understand that generative AI can create new content by learning from existing data, offering innovative solutions and creative possibilities, but they must also be aware of its limitations and the importance of ethical oversight. Lenovo aims to simplify AI implementation by delivering it wherever data resides through the use of Lenovo’s network and the resources of its large network of partners. Every tech support group needs outside help from time to time, so it is most helpful that HPE GreenLake LLM also provides access to AI and performance engineering experts, for assistance with optimizing supercomputing systems and software resources.

Overall, the integration of Generative AI in business operations holds the promise of increased productivity, creativity, and competitiveness in the marketplace. However, roadmaps varied as to when each company planned to offer foundation models and GenAI as stand-alone products or to create advanced features in existing products. By 2024, Gartner predicts that 60% of data for AI will be synthetic for simulating reality and future scenarios. GenAI can create synthetic data sets that imitate real-world data which can train machine learning models for fraud detection, customer segmentation, and demand forecasting. As a CEO, this is relevant as your business can be relieved from the burden of collecting real-world data. Artificial Intelligence, machine learning, and data science are more than just buzzwords in the current business landscape.

What CEOs need to know about gen AI

Two scenarios are shown for early and late adoption of automation, and each bar is broken into the effect of automation with and without generative AI. The addition of generative AI increases CAGR by 0.5 to 0.7 percentage points, on average, for early adopters, and 0.1 to 0.3 percentage points for late adopters. In the overall average for global growth, generative AI adds about 0.6 percentage points by 2040 for early adopters, while late adopters can expect an increase of 0.1 percentage points. McKinsey has found that gen AI could substantially increase labor productivity across the economy. To reap the benefits of this productivity boost, however, workers whose jobs are affected will need to shift to other work activities that allow them to at least match their 2022 productivity levels.

Embracing a mindset of perpetual learning enables CEOs to stay abreast of emerging trends, technologies, and ever evolving market demands. CEOs enhance their leadership strategies by learning continuously, attending workshops, and seeking diverse insights for a broader knowledge base. Collaboration and integration are crucial when incorporating generative AI into existing systems. It involves merging the innovative capabilities of generative AI with established structures to enhance efficiency and productivity. Collaborating, these systems adapt, learn, create harmony, maximizing AI tech and existing framework’s potential through seamless integration. High performance hardware, such as GPUs or TPUs, is necessary to process and learn from the data, a process that can take hours, days, or even weeks.

It will then be critical to use sound data practices and in-house expertise to safely train models for maximum business value. Like other companies, Cisco has been experimenting with AI for years but hasn’t incorporated generative AI into products until now. The new Webex AI features will increase user productivity by making the product easier to use and automating meeting follow-up activities. According to a 2023 Cisco study, IT professionals ranked generative AI as the technology most likely to impact business significantly.

• CEOs face the decision of whether to embrace generative AI now or proceed cautiously through experimentation.
• The company found that major updates to its tech infrastructure and processes would be needed, including access to many GPU instances to train the model, tools to distribute the training across many systems, and best-practice MLOps to limit cost and project duration.
• Seeking expert guidance and investing in compatible infrastructure can mitigate integration obstacles.
• HPE GreenLake for LLM offers customers the performance characteristics of a supercomputer and a cloud service.
• This approach could enable companies to promote broader AI adoption and create the culture of innovation that is essential to maintaining a competitive edge.

After some initial indecision, Google followed Microsoft’s example by adding generative AI tools to all its productivity and search products. I wrote an earlier article on Google’s full range of actions to shore up its competitive position. This trend started after OpenAI released ChatGPT in November 2022, creating one of the fastest and most extensive industry disruptions in modern business history. By January 2023, ChatGPT had set a record as the fastest-growing consumer software application in history, gaining over 100 million users.

The CEO has a crucial role to play in catalyzing a company’s focus on generative AI. In this closing section, we discuss strategies that CEOs will want to keep in mind as they begin their journey. Many of them echo the responses of senior executives to previous waves of new technology. However, generative AI presents its own challenges, including managing a technology moving at a speed not seen in previous technology transitions.

Generative AI has the potential to revolutionize the entire customer operations function, improving the customer experience and agent productivity through digital self-service and enhancing and augmenting agent skills. The technology has already gained traction in customer service because of its ability to automate interactions with customers using natural language. Crucially, productivity and quality of service improved most among less-experienced agents, while the AI assistant did not increase—and sometimes decreased—the productivity and quality metrics of more highly skilled agents. This is because AI assistance helped less-experienced agents communicate using techniques similar to those of their higher-skilled counterparts. Generative AI could have an impact on most business functions; however, a few stand out when measured by the technology’s impact as a share of functional cost (Exhibit 3).

Companies that have not yet found ways to effectively harmonize and provide ready access to their data will be unable to fine-tune generative AI to unlock more of its potentially transformative uses. Equally important is to design a scalable data architecture that includes data governance and security procedures. Depending on the use case, the existing computing and tooling infrastructure (which can be sourced via a cloud provider or set up in-house) might also need upgrading. A clear data and infrastructure strategy anchored on the business value and competitive advantage derived from generative AI will be critical. Gen AI is a big step forward, but traditional advanced analytics and machine learning continue to account for the lion’s share of task optimization, and they continue to find new applications in a wide variety of sectors.