2023 in review: How to go beyond the buzz and scale generative AI across the enterprise

As the year draws to a close, I find myself looking back on the milestones, challenges, and moments that have defined the past 11 months. There’s no doubt that the buzz surrounding generative AI dominated headlines and conversations throughout 2023, often hailed as a “must have” technology.

But, beneath the buzz, there is a reality, which is that generative AI will transform the anatomy of work, with its potential to automate work activities that absorb 60 to 70 percent of employees’ time today, according to McKinsey. It’ll impact industries across all sectors, ranging from banking to life sciences and beyond. Unlike other buzzy topics like Crypto and the metaverse, generative AI is here to stay, and it will revolutionize the way businesses operate.

In fact, it already is.

I spent most of 2023 traveling to meet with customers around the world and discuss the promises and pitfalls of generative AI. For me, this is a huge privilege because I get to learn firsthand what companies are experiencing, what problems they’re trying to solve, and how they’re trying to adopt this new technology.

The truth is, despite generative AI’s promise, most enterprises can’t get beyond the buzz. According to McKinsey, 60% of leaders claim they rarely or never using generative AI for commercial activities. That’s because, in order to put these technologies into production in large enterprises, business leaders need to consider much more than just the technology.

I think my biggest takeaway from my time spent with customers, is that there’s a big difference between trying to adopt generative AI, and actually putting it into production and scaling it across the enterprise. The technology aside, you must be willing to change your approach to your company’s problems, people, processes, and protection.

In this blog, I want to detail some of what I’ve learned not just from my conversations with our customers, but also from my market research, in an effort to help those business leaders who are looking to go beyond the buzz and scale generative AI across their enterprise.

Problem: Identify what you’re trying to solve

Companies find themselves in varying places on their AI journey. Some are ahead, some are in the middle, and some are just getting started. Regardless of where you stand in your AI journey, the biggest question you have to ask yourself is what problem you’re trying to solve. A lot of people get enamored by technology, but if you don’t know what problem you’re trying to solve, then it’s very easy to get carried away and spend millions of dollars trying to jump on the generative AI bandwagon without knowing where you’re going.

Now, based on my work with our customers, I’m seeing a number of different problems that companies are trying to solve for — from asset management, to supply chain, to content creation. But, for those who are just starting on their AI journey, here are the 3 use cases for generative AI that I would say I see most often based on their high impact potential:

1. Talent. Your people are your biggest asset. Yet, between “The Great Resignation,” a crushing skills shortage, and the growing complexity of data analysis tools, the work of finding and retaining top talent can be challenging. With generative AI, you can give your team the tools they need to onboard and support great hires. Our customers are already seeing a 40% increase in HR productivity. Learn more here →

2. Customer Service. Time is money, especially in customer service. It’s no surprise that customer service has leapfrogged other functions to become CEOs’ #1 generative AI priority. That’s because generative AI can help you understand customers in the right context, and provide fast, consistent, and accurate answers while freeing up your agents for the more complicated problems. In fact, our customers are having 70% of contact center cases be contained by conversational AI. Learn more here →

3. Application Modernization. 89% of C-Suite executives agree that generative AI in app modernization projects will drive growth by improving existing products and services. By increasing developer productivity, our customers are seeing up to a 30% productivity gain in application modernization. Learn more here →

You can find more top industry use cases, like AI for aerospace, AI for sports, AI for real estate, and more by visiting my colleague, Armand Ruiz’s blog. Go now →

People: Make your employees central to your strategy

As I stated earlier, your people are your greatest asset. And, despite all the talk about generative AI taking jobs, what I’m seeing is that generative AI isn’t replacing people, but people who use generative AI are replacing those who don’t.

77% of entry-level workers will see their jobs shift by 2025 — as will more than one-in-four senior executives. As a leader, you need to re-think your strategy and actions around talent and skills to ensure you’re making people central to your generative AI strategy. You can do that by:

1. Taking a skills first approach. One of the common challenges I’m seeing is finding the right skills to scale AI across the enterprise. To stay competitive your workforce must be continually learning. You need to upskill your existing workforce to make generative AI an advancement opportunity for everyone, at all levels, including the C-Suite.
2. Create a culture of curiosity. What is true today might be false tomorrow because this space is constantly, and rapidly, changing. The best way to overcome that is to be curious. Whether you’re an engineer, a product manager, a seller, or a marketer, it doesn’t matter. This space requires a culture of curiosity to be successful. As a leader, you must accelerate creativity and curiosity, which will help your employees find productive and innovative ways to interact and leverage generative AI within your organization. At IBM, for example, we hosted a company-wide watsonx challenge designed to bring our IBMers together to drive creativity and experiment with what AI can do, while getting hands-on experience with the watsonx platform.
3. Elevate HR from being a purely administrative function. Your HR team will have a strategic role in building the generative AI-enabled workforce of the future. So, it’s a good idea to start by reskilling the HR professionals who will need to lead this effort. Check out how we did it at IBM →

Process: Reinvent your workflows

Many of the companies who were late to join the mobile revolution (think taxis or food delivery as prime examples), were unwilling to do the deep rethinking of their business models and workflows. The same is going to be true of this AI revolution. Because the reality is, the actual technology is only a very small piece of the generative AI discussion.

To successfully scale generative AI across the enterprise, you can’t try to put new AI technology on top of existing business models and workflows. You must reinvent your processes and workflows from the ground up. The enterprises that will enjoy the most success from their AI projects are those that change their mindset from “adding AI,” to “AI first.” To be truly impactful, here are some of the common workflows and processes we’re seeing companies reinvent with AI and automation…

1. IT Automation. Make your IT systems more proactive, processes more efficient and people more productive. Rethink and automate your IT processes →
2. Security. Expand visibility and accelerate response times as security teams are often short-staffed and stretched thin. Read the power of AI in security report →
3. Sustainability. Operationalize your sustainability goals and increase transparency. Read the guide for sustainability leaders →

Protection: Put AI and data governance at the heart of your AI lifecycle

As I discussed in length during my last blog series, generative AI brings great promise, but it is also forcing business leaders, governments, and everyday people alike to seek assurances that the AI systems they use and interact with are trustworthy. While businesses recognize AI’s power, about 80% of business leaders acknowledge ethical concerns with adopting generative AI.

This is why protections — or governance — is so important for enterprises to address, and address now. Traditionally an IT concern, AI and data governance must have a firm footing in C-suite conversations. To gain a competitive edge, companies will need to cut through the red tape. At the same time, they must take a strategic approach to AI ethics and embed ethical principles into their end-to-end AI development and deployment process to ensure it’s transparent, trusted, and fair. You must be able to answer critical questions like:

• How was it trained? ChatGPT, for example, was trained on a large body of data from a variety of sources (like Wikipedia, Reddit, and more). This is essentially a garbage in-garbage out scenario. If the data used to train the model hasn’t been vetted, the contents can be inaccurate, full of bias and mistakes.

• Can it detect bias and hallucinations? Models built with inherent biases will produce biased results. A hallucination occurs when the model simply makes up answers based on its best guess.

• Is it transparent? Can you explain how the model is getting its answers?

• Does it support regulatory compliance? There are many regulations and laws that already exist, and many more coming. You must not only remain compliant, but be able to keep up with the new regulations that are coming next.

• Is it safe? You must consider how your enterprise can control the quality of the output and be sure it’s not disclosing or leaking confidential information.

• Can it be customized? Most enterprises need to be able to customize their models. Whether that’s to run on multicloud or train it with your own data.

The trusted AI conversation is a complicated but essential one — so, here are some practical first steps to get started:

1. Designate a lead AI ethics official. Hire or assign one of your company leaders to be responsible for your trustworthy AI strategy.

2. Stand up an AI ethics board, or similar function. Successful enterprises need a centralized clearinghouse, like IBM’s AI Ethics Board, for resources to help guide implementation of the trustworthy AI strategy defined by your lead AI ethics official and shared with your people across the enterprise.

3. Adopt strong internal governance practices. At IBM, we offer enterprises a number of different products and services specifically designed to help organizations adopt strong internal governance practices to ensure their AI outputs are transparent and explainable. Learn more here →

Speaking of protections, a few months ago, we announced the general availability of the first models in the watsonx Granite model series, which is a collection of foundation models to advance the infusion of generative AI into business applications and workflows. Learn more about the power of our Granite foundation models →

What makes the watsonx Granite models different is that they’ve been trained on enterprise-relevant datasets across five domains — internet, academic, code, legal and finance — and have been curated for business use by IBM. Training data was filtered for objectionable content and benchmarked against internal and external models to help enable responsible deployment and address key issues including governance, risk assessment, privacy concerns and bias mitigation. IBM believes in the creation, deployment and utilization of responsible AI models, which is why our standard IP protection applies to our watsonx models.

This is just one example of IBM’s commitment to helping enterprises develop a trusted, end-to-end AI lifecycle management process. If you’re looking to scale and accelerate the impact of AI with trusted data across your business, I encourage you to check out our new AI and data platform, watsonx →

Looking ahead to 2024

As we enter 2024, it’s evident that generative AI’s revolutionary impact on enterprises is just beginning. It will continue to reshape industries, foster innovation, drive efficiency, and deliver highly personalized experiences. However, to be successful, it is crucial that organizations consider their problem, people, process, and protections as they embrace this transformative technology.

By considering their problem, people, process, and protections when it comes to generative AI, companies will be able to tell what I refer to as their AI story. Read my blog on the importance of an AI story here →

The AI revolution is just beginning, and I am excited to see what lies ahead as companies worldwide move beyond the buzz and start to scale AI across the enterprise to tell their story.