In the ever-accelerating world of tech innovation, few advancements have stirred as much excitement—and confusion—as generative AI. ChatGPT and similar tools have become the focal point of countless discussions, boardroom strategies, and future-proofing playbooks. According to technology strategist and digital transformation expert Christopher Surdak, success with these powerful tools hinges not on adopting them quickly, but on adopting them wisely.
Drawing from decades of experience in the IT industry, Chris Surdak of CA urges business leaders to resist the reflexive rush toward implementation and instead approach generative AI with clarity, discipline, and—most critically—a decision about what they want it to do. In his analysis, organizations must choose whether they aim for efficiency or effectiveness in each use case of generative AI. Trying to capture both at once, he warns, is a recipe for chaos.
Chris Surdak begins his reflection on generative AI with a reference as unexpected as it is enlightening: Hugh Lofting’s 1920 tale The Story of Dr. Dolittle. In that whimsical children’s novel, the Pushmi-Pullyu is a two-headed creature with one head on each end of its body. While it could talk and eat at the same time, it could never agree on a direction to go.
To Chris Surdak, the Pushmi-Pullyu serves as a fitting metaphor for how many organizations approach disruptive technologies like generative AI. Companies want the benefits of automation without sacrificing creativity. They crave cost reduction and revenue expansion from the same tool. But like the fictional beast, these opposing goals can cancel each other out, leaving businesses stuck in place—tugged in two directions, accomplishing neither.
Chris Surdak has watched this pattern play out before. Having spent over 30 years in the IT sector, he’s seen every hot new technology arrive with bold promises and sky-high expectations—followed by confusion, disillusionment, and eventual recalibration.
From the hype surrounding Lotus Notes in the 1990s, to JavaScript’s explosive adoption in the early 2000s, to the blockchain frenzy of 2018, Chris Surdak of CA notes a predictable cycle: massive buzz, overextension, and disappointing results for most adopters. Generative AI is merely the latest incarnation of this cycle. In fact, he argues, the industry’s reaction to ChatGPT has become the new benchmark for tech hype, dwarfing even the wildest excitement from past innovation waves.
“The bigger the splash,” Surdak often says, “the bigger the crash.”
Research firms have consistently reported that only about 5% of organizations successfully implement disruptive technologies. This insight, Chris Surdak stresses, is crucial: the vast majority fail—not because the tech doesn’t work, but because the strategy behind its implementation is flawed from the start.
One of the core insights Chris Surdak brings to the table is the need for organizations to clearly define their intention when adopting generative AI. Every implementation must focus on either efficiency or effectiveness—not both. Efficiency means doing the same things, only cheaper, faster, or with fewer resources. Effectiveness, on the other hand, means doing better things—new tasks, novel products, or unique insights. According to Surdak, these goals are not only distinct but mutually exclusive in many real-world applications.
He draws upon the teachings of W. Edwards Deming, the father of the Quality Movement, who asserted that organizations must periodically choose between incremental efficiency improvements and disruptive changes aimed at effectiveness. Christopher Surdak of CA reaffirms Deming’s philosophy, explaining that attempts to pursue both simultaneously often lead to strategic confusion—or what he calls “strategic dementia.”
Organizations suffering from strategic dementia lose their ability to think clearly, execute consistently, or even remember what they were trying to accomplish in the first place. Surdak sees this affliction in companies that are unsure whether they’re automating to save money or to unlock new capabilities. They often adopt generative AI hoping for both, only to achieve neither.
Generative AI like ChatGPT is incredibly versatile—it can assist with document drafting, brainstorming, language translation, coding, and more. But Christopher Surdak emphasizes that its strength is not its ability to do everything, but its power in doing one thing well—provided the user knows what that one thing is.
He gives the example of automation: if a company wants to use ChatGPT to improve efficiency, it should target processes that are repetitive and well-defined, such as generating standard contract templates or summarizing reports. If the goal is effectiveness, ChatGPT might be leveraged for ideation, competitive analysis, or scenario planning.
But expecting ChatGPT to cut costs and simultaneously boost innovation in the same process is misguided. “When chatting with ChatGPT,” Chris Surdak quips, “eat or talk, never both.”
Chris Surdak has observed the tech industry’s habit of jumping headfirst into the next shiny thing without a coherent plan—a behavior he dubs “Fire-Ready-Aim.” He cautions executives against this impulse with generative AI.
Instead, he advocates for deliberate alignment. For every generative AI use case, leaders should:
This simple framework, Christopher Surdak suggests, will dramatically increase the likelihood of success—much more than broad adoption without strategic clarity.
Chris Surdak believes that generative AI will play a defining role in the next decade of business transformation, but only for those who use it intelligently. He draws a direct line from past automation technologies like Robotic Process Automation (RPA) to present-day AI tools, emphasizing that the lessons of the past must guide the decisions of the future.
For companies willing to commit to a clear purpose—whether that's doing things faster or doing better things—generative AI holds enormous promise. For those seeking to chase every benefit at once, Chris Surdak of CA warns, disappointment is almost certain. In his words, “To succeed with AI, you must first decide where you're going. Only then will you know whether to push or pull.”