Gen AI: From theory to practice

The best way to get started with generative AI is to understand what it’s for: unlocking and expanding value.

A year has passed since generative AI exploded into the public conversation. For businesses, it was clear from the outset that generative AI could “change everything” about the modern enterprise landscape. Less clear, at least initially, was what that change would look like, let alone what practical steps a business should take, and in what order, to begin its gen AI journey.

While some of those blanks have been filled in, uncertainty remains. In our recent survey of senior business and technology executives in the US and UK, 99% are enthusiastic about this new technology, with 61% predicting a complete transformation of their business. But fewer than half (46%) said their organizations have managed to develop robust value propositions around generative AI, and a whopping 75% say their company is “stuck” when it comes to next steps in generative AI deployment.

What’s the problem? The specifics vary, but in my work with clients this year, I’ve found that the revolutionary features of gen AI are less of a sticking point than the challenge of integrating the technology with their existing business processes and technologies. Many also get bogged down with addressing gen AI integration as a technical problem, neglecting a deeper, vital question about this new technology, namely: What is generative AI for?

The clear answer in my conversations with clients across industries: Generative AI is a tool for unlocking and expanding value. Properly understood, this concept not only answers many of those shorter-term questions about integration but can also lead the way to truly radical gains and growth.

Running the business

Before looking at radical gains, businesses should start by unlocking value trapped within their existing business processes. To identify use cases where you can run your business more efficiently, make a list of your business’s manual processes and add in areas of known inefficiency. In many cases, generative AI could probably help tackle most of these items, by automating manual processes and further streamlining processes that are already automated (including those automated by “traditional” AI). The impact: immediate cost savings, shorter value chains that speed time to market, a better customer experience and a healthy boost to the bottom line.

Most businesses are already beginning to do this in pockets. In our survey, 77% of executives said their organization has begun integrating gen AI into the recruiting process—to analyze applications and screen applicants. Around 80% are using the technology in customer service, while fully 90% are confident that gen AI will enhance the power and efficiency of their marketing strategies.

Changing the business

But savings and efficiencies are just the beginning of what generative AI has to offer. Businesses can also mine this powerful technology for much deeper, richer streams of value. These can be found among the business’s data assets—data that has been generated or acquired by existing processes but has not yet been used or monetized. Think of these assets as the exhaust of your current operations that could expand value when applied to different markets or a different business model.

Customer data is a prime example. Many organizations are sitting on reams of customer information that generative AI could use to design personalized product offerings or package in such a way that it could be monetized to advertisers.

It gets better. Gen AI systems can do more than find value in existing data sets. They can also gather and integrate pools of unstructured data that may, today, be stranded in isolated systems, platforms and departments across a business. As a general principle, the more disparate and far-flung these “data silos” are today, the more novel and valuable the insights they could yield once they’re consolidated.

And don’t be constrained by a narrow conception of “data.” We’re working with a biopharmaceutical company to develop generative AI-powered portals to give doctors and patients greater visibility into the company’s treatment options. This data could be leveraged, through gen AI, to enhance the user experience, shrink the value chain and speed time to value.

Transforming business models

If this were all, it would be a lot. The bottom-line benefits of more efficient operations—coupled with the value gen AI can unlock within existing business models—will confer a huge competitive advantage on those companies that take a proactive but responsible approach to this new technology. But for a lucky few, perhaps a lucky many, gen AI will not only generate savings and new revenue streams within existing business models but will also help create new business models and market opportunities from scratch.

Consider the case of Amazon Web Services. Amazon.com, you may remember, started life as an online bookseller. As the business expanded into general e-commerce, Amazon built a sturdy IT infrastructure to handle its increasingly vast online operations. Rather than hoard this technology, Amazon in 2006 launched Amazon Web Services (AWS), a standalone cloud-computing service that countless companies, from startups to Fortune 500 companies, now license to support their own operations.

That AWS today accounts for “only” 16% of Amazon’s total revenue is a testament to the persistent health of its e-commerce business. AWS’s own revenues, however—a cool $80 billion in 2022—make it a behemoth by any objective measure.

While Amazon built and launched AWS without the help of gen AI, the pattern it represents—that of a business model born, incubated and launched within the structure of an existing and very different business model—will be much more common in the age of generative AI. The unlocking of value, at every level of a business from the micro to the macro, will create a fertile ground for innovation and make such “happy accidents” all but inevitable.

Turning the key

Start small with gen AI. Before any wholesale commitment can take place, there is groundwork to be laid: reskilling workers, restructuring management and, perhaps most important, “replumbing” data pipelines and IT architectures for gen AI integration.

Outside consultants can bring tremendous value here, as they can help with the subsequent work of building and/or fine-tuning a large language model. Consultants may also be able to immediately identify the first-order inefficiencies and potential savings that existing gen AI tools can unlock without major capital investment.

By steering a course toward this North Star, you’ll find the road will be much clearer, less daunting and radically shorter to realizing significant gains in value from generative AI.

To learn more, visit the Generative AI section of our website or contact us.