What is the strategy for generative AI? A key question every CEO should be concerned about

Author: Matt McIlwain. Translated by: Cointime.com QDD
“What’s our generative AI strategy?” It’s the question facing nearly every executive team at today’s Fortune 500 and other companies. The person asking the question is almost always the CEO, who expects an answer and quick results. At the same time, the same CEO will be unhappy if they discover that generative artificial intelligence (GenAI) begins to hallucinate your customers, strategic data assets are being mismanaged, or smart applications are causing security or performance issues for your company. While this reality will keep management consulting firms thriving for years to come, some practical guidance is needed to help senior executives (including CIOs and CDOs) and their business unit partners develop a compelling plan. Through conversations with enterprise clients, GenAI companies, and cloud service providers, we’ve discovered three core questions that can guide your generative AI strategy:
1. What kind of enterprise mindset is needed to successfully adopt GenAI?
2. How to achieve “quick wins” in the short term?
3. What is the best strategy for leveraging GenAI in the medium to long term?
Successful business mindset
You may have heard that AI experienced its "Mosaic Browser moment" with the launch of ChatGPT late last year. We've been using AI for decades (Google Search, Amazon Alexa, Netflix/Spotify recommendations), but now individuals can use GenAI directly and creatively to quickly deliver value. Whether that value is drafting a note, developing software code, or completing a task, by prompting the AI ​​model with natural language, GenAI can help anyone be more efficient and effective. The opportunities are endless in business settings, and your employees are already experimenting!
Because GenAI is so accessible and flexible, it requires an agile mindset that is often hard to find in large companies. Functions will evolve, new skills will be required, and many companies will be challenged by native GenAI businesses. There is no doubt that GenAI will be disruptive and many employees will resist the change. Some employees will use areas such as security and governance to approach your experiments with caution. It is important to manage these risk areas and set safeguards and operational best practices, such as the framework developed by WhyLabs. However, many employees, including CEOs, are eager to try, learn, and apply GenAI to your business. Finding ways to leverage these new technologies with low risk and fast feedback loops is critical for every business.
Think about your company’s domain capabilities and differentiated data assets first, and how they can be leveraged to benefit your company and your customers in the long term. Established customer relationships, differentiated data and expertise, and existing workflows and interfaces can all be valuable tools for existing businesses. Your employees and customers are already experimenting with how your product and value proposition fit together with GenAI extensions. Embrace these innovators and work to understand the problems they solve and why. By fostering a climate that values ​​and encourages smart experimentation, you can quickly understand the range of possibilities and priorities in your organization, while managing risk!
GenAI is very approachable for end users, but building and operating GenAI models and applications is not easy.
A three-step approach to quick wins
By asking your employees, customers, and trusted technology partners what they are already doing with GenAI, you’ll uncover opportunities that are relevant to your business. In the short term, we recommend taking the following three steps to get started and learn quickly as you develop your company’s GenAI strategy:
1. Embrace “collaborative partners” to improve productivity.
2. Partner with large cloud service providers (CSPs) and software companies to deliver GenAI solutions in an “enhanced” manner.
3. Launch a GenAI learning lab and host a hackathon to conduct rapid and agile experiments to help identify and prioritize your GenAI capabilities. This learning lab can also help you discover emerging, innovative “native GenAI” companies and supplement your internal skills.
Collaboration Partners
One of GenAI’s early success stories has been the emergence of collaborative partners — such as Microsoft’s GitHub Copilot or Amazon’s Code Whisperer — to improve software development. Similarly, OpenAI’s ChatGPT and other tools have enabled intelligent assistants to appear in many forms, helping workers become more efficient in the creation of written texts.
Your developers are already leveraging GenAI tools to write and optimize code, knowledge workers are drafting marketing documents, and project managers are automating specific workflows. These tools are already improving employee productivity today. In particular, your engineering teams can improve software development by driving adoption of code generation tools. Many software teams are already seeing productivity gains of 20-30%. This is so significant that these teams should take time out of their current software development methodologies and learn how to leverage these new methods. Teams that don’t use code generation tools to improve productivity will fall further and further behind in delivering software projects.
Understanding what is already happening within your organization and encouraging increased productivity will help you assess your internal GenAI readiness and categorize the most likely opportunities as “do it yourself,” “do it with me,” or “do it for you” initiatives. In other words, is your company ready to bake the GenAI cake itself from the individual components (data, algorithms, GPUs), use off-the-shelf cake mix and frosting to make the cake (cloud service providers, hosted models), or buy it from a local bakery (native GenAI and enhanced SaaS solutions)? From an end-user perspective, GenAI is very approachable, but building and operating GenAI models and applications is not. Observing what your employees are already doing will create a baseline for internal capabilities and plans for GenAI solution delivery. To be successful, you need to objectively assess your internal capabilities, including collecting and preparing data, training and deploying foundational GenAI models, and building capabilities for GenAI enhancements to existing applications and workflows. Most likely, the early gains will be internal productivity gains and capability learnings from purchasing an enhanced solution.
Directly leverage partners
Look for existing application partners who have already integrated their data infrastructure and applications with GenAI. You have a business relationship with them, and they are typically security and compliance vendors. You also share context and data sets, and your employees understand the user interfaces and workflows provided by these vendors. As a result, they can quickly help you improve your business processes and workflows. These partners may have better access to emerging technologies and the skillsets to quickly enhance their GenAI stack.
In the short term, your established partners, including cloud service providers such as Amazon AWS, Microsoft Azure, Google Cloud, and software application companies such as Workday, Atlassian, Smartsheet, etc., have already provided you with enhanced solutions. There is also a data management layer, where CSPs and data service companies such as Snowflake, MongoDB, and Databricks provide the data management capabilities required for deeper GenAI work.
The key to taking an initial buy-better-than-build approach is to achieve additional quick wins by understanding and leveraging what your established partners are doing. Then, as you further clarify the direction of your company, you can find the best approach to building your own GenAI enhancements.
Building a GenAI Learning Lab
Whether it’s from internal early adopters or enhanced third-party solutions, your early successes will be primarily about improving internal productivity and creativity. However, the more fundamental need is to determine how to transform your business to better serve external customers. Many tech companies have already held hackathons to energize internal teams and develop priority opportunities to serve customers. This is a great place for your company to start, too! In addition, they have designated a senior executive as a cross-functional GenAI leader to lead experimental efforts and develop a GenAI strategic plan. You can think of these efforts as a GenAI learning lab.
The more existential need is to determine how you will transform your business to better serve your external customers.
While this team may monitor and summarize the rapid changes in GenAI, the most effective way to discover is to learn by doing. Being selected for the GenAI Learning Lab should be seen as an opportunity to engage potential company builders and leaders. They will need access to data resources and may need to hit walls in IT and data management to experiment quickly. Sponsorship from senior executives will be key to balancing experimentation and risk. The lab (and the company as a whole) should establish some rules and regulations about when and how projects can be shared in a safe and compliant manner. While the hackathon that helped launch the team can provide inspiration, it will most likely take time and some outside expertise for these teams to develop priorities and products that customers can test. An attitude of action and experimental thinking will help the lab and the company succeed.
Some specific early priority areas are generative chat and generative search for your customer service representatives and customers. Generative chat is about letting your customers and service representatives interact with the data and knowledge resources you already have in natural language, helping them get answers faster. The idea is to interact through a chat-like interface and get answers to questions quickly. Generative search is similar to generative chat, but is more about adding reasoning and insights to your search results. Google recently announced a series of features around enterprise search.
The Learning Lab is likely the best place to work with innovative startups building more native GenAI services in areas such as writing, images, video, code generation, and more. Companies like Jasper.ai, Copy.ai, HyperWrite, and of course OpenAI’s ChatGPT are already natively supporting a variety of writing use cases. Companies like Lexion (legal contracts) or Harvey (legal proceedings) focus on specific areas of drafting and document management. RunwayML, an early leader in the video creation and editing market, is very fortunate to have a team that deeply understands the entire GenAI stack. There are also companies emerging that focus on GenAI in enabling areas such as model testing and deployment (OctoML, Mosaic), data wrangling (Number Station), and data ingestion (Unstructured.io). We believe that native GenAI companies are looking at customer problems from a fresh perspective, the new technical capabilities required, and the agile approach to continuously iterate from data to end users. As Tomasz Tunguz thoughtfully shared a few weeks ago, building, deploying, and operating generative applications will require significant product development as well as organizational and cultural changes. Collaborating with your learning lab and later with certain native GenAI companies will inform your long-term approach to capability building.
Winning in the middle of a long process
With some quick wins and capability building underway, you’ll likely have earned buy-in and confidence from others in the organization (including the CEO) for longer-term projects. You’ll also be better calibrated with your internal team and process capabilities, data assets, and customer needs. Building on your quick wins, take the time to work back to develop a longer-term set of priorities for what you’re going to deliver and how you’re going to deliver it.
Many quick wins with collaboration partners and software partners will reveal insights into internal and external opportunities for you. Where can you provide more automation, personalization, and cost savings in areas such as customer satisfaction and customer success? Do you have proprietary data and/or understanding of public data sources that can help you build differentiated GenAI models and solutions? Is there a better system, enhanced by GenAI, for designing the next iteration of products and services? Will a distribution or partnership channel emerge to sell your product in the market while continuing to meet your needs for data and reinforcement learning?
One of the most important decisions you face is “How do we do this?” Your quick win will likely be to identify internal champions who have the skills and passion to embrace your GenAI transformation. Developing these GenAI champions and integrating them into long-term initiatives will greatly improve your chances of success. In particular, look for champions who are intellectually curious but don’t instinctively assume that your company must take a do-it-yourself approach. Most companies will benefit from both a do-it-yourself and a “do-it-with-me” partner approach as they seek to leverage the entire GenAI stack to better meet customer-facing and employee needs.
In general, here are some guidelines to increase your chances of success:
1. Initial focus on cloud-based deployments rather than edge use cases (edge ​​is emerging but still in its early stages).
2. Design from the data up and differentiate the application layer infrastructure at the application layer. You will most likely leverage other open source and hosted models (including language models and domain-specific models) and cloud infrastructure, so it is your data and workflow that will help you customize the models and applications that create unique value.
3. Always push to understand your data and metadata assets and manage them strategically to achieve long-term value (balancing the risk of data breach with the risk of data loss). You will most likely need to upgrade the underlying data stores, such as data warehouses and database types, to eventually build language models that leverage proprietary data and insights.
4. Employ both native GenAI innovators and augmented existing technology partners to support your GenAI journey. This should help increase internal productivity and creativity while improving customer experience and value!
Prepare for a marathon, not a sprint
Nearly 30 years ago, I was a young Corporate Vice President at a Fortune 500 company called The Genuine Parts Company (GPC). Our businesses were in diverse industries, including automotive parts (NAPA), industrial products, and office supplies, providing excellent supply chain services from manufacturers to end-user customers. In 1995, a disruptive technology called the Internet emerged, sparking the first wave of a large venture capital hype cycle and a series of business innovations that impacted every company. Every week, a new venture-backed company contacted GPC to explore how they could be a new technology enabler, a new type of customer, or a new partner (or possible competitor!) for us. My role quickly evolved into screening these companies, evaluating their fit with our strategy, and developing arrangements that would extend our value while embracing technological change.
As the dot-com bubble neared its peak in the evening, the entire business world was immersed in e-commerce, Internet companies, and “brick-and-mortar” innovation. Traditional companies like GPC saw their stock prices plummet amid concerns that they would fall into the “innovator’s dilemma.” NAPA Auto Parts, Amazon, and Madrona explored a “brick-and-mortar” venture, but they soon realized that the business market was not ready for an online auto parts store until the NASDAQ reached its then-record 5,000 points (March 2000). Others were not so lucky, and countless founders, startups, and investors suffered huge losses when the bubble burst. I will never forget attending a conference hosted by Fortune and Goldman Sachs in New York in April 2000, which was themed on the battle between Fortune 500 companies and Internet companies. Few believed at the time that Fortune 500 companies had a chance, but today, GPC is worth 8 times more than it was then, thanks to a learning mindset, excellent execution, and a long-term perspective!
In 2023 and beyond, disruptive GenAI technologies, intelligent and generative applications, and more natural user interfaces will create winners and losers among startups, incumbent technology companies, and established enterprises. Enterprises and their leaders will need agility, creativity, and a long-term perspective to navigate this highly dynamic era. They will also need strategic clarity on customer needs, objective assessments of internal capabilities and data assets, and engagement with external partners to emerge as winners. This is the deeper reason why every CEO is asking, "What is our generative AI strategy?"