Cheryl Johnson is the CTPO of Betterworks, shaping product strategy that redefines performance management practices for global enterprises.
The first phase of AI adoption was about exploring the art of the possible to land on useful capabilities. With all of the experimentation, a lot of fruit died on the vine, and honest post-mortems tell us that business impact was minimal.
This is, of course, typical with any new technology until organizations find their stride with it.
The next phase is no less exciting: allocating and optimizing tech capabilities and human capital to derive business value and, ideally, create a competitive advantage.
It’s a fine but critical line that requires human judgment as much as anything else. The reality is that as AI moves from experimentation into production, every AI decision becomes a business decision. Every prompt fed to an LLM, every AI-powered workflow and every agent unleashed on a business process carries with it a cost.
It’s incumbent upon organizations to weigh their capabilities against cost, speed, human expertise and customer value. Those trade-offs will reshape product strategy and engineering priorities, and create a winding path that determines how organizations create business value.
Oftentimes, this unfurls in unpredictable, jagged ways. While I’d argue that engineering is where this shift materializes first, it won’t stop there.
As organizations optimize for value instead of capability, they’ll redefine what great engineering looks like, the skills they reward and ultimately how they create competitive advantage for years to come.
Organizations must optimize intelligence, not just build it.
The conversation organizational leaders must have is deciding where human judgment creates more value than AI, and where AI creates more value than human effort.
These are not easy discussions for any department or business line. They’re certainly not easy for IT, no matter its skill set in supporting disparate domains.
Consider that, until recently, engineering organizations optimized hybrid infrastructure environments, carefully calibrating factors such as networking latency, scalability and reliability, among other tasks.
Engineering is increasingly taking on not only how to select and optimize models but also how to analyze and account for prompt efficiency, orchestration, caching and the budget for inference costs. And that’s not even including the human-versus-AI trade-offs, the thousands of micro-decisions around where the human is in the loop, on the loop or operating the loop.
It’s already happening. This changes how organizations allocate capital, prioritize innovation and ultimately build competitive advantage.
As AI makes code abundant, judgment becomes scarce.
The subheading above sounds worse than it is at first blush. Although far from perfect, AI systems are getting much better at producing clean, reliable code. As exciting as it may be to generate a lot of code dynamically, too much may lead to a deficit in judgment.
This is because there is a tendency for some people to fall prey to cognitive offloading, in which humans lean heavily on AI to do their critical thinking. On a long enough timeline, this increases cognitive debt, making judgment the scarcest resource in engineering.
So what happens when AI assumes the heavy lifting of coding? Where do we shift judgment to generate value to the business?
In my view, the value comes from focusing on architectural thinking, systems design and decisions around product development. What to build, how to build it and in what order, as well as how to govern what gets built so that it not only protects organizational data but preserves compliance policies.
It bears repeating: The value will also come from evaluating trade-offs between not only architectures and products, but also where and when to keep humans in the loop, as well as where to allocate them.
Organizations will compete differently.
How will this change the downstream dynamics with regard to competition? It’s a great question and one I’ve been thinking a lot about.
Engineering teams are the first to experience this shift because they’re closest to AI deployment. But the same economic pressures exerted on resources, from technology to human capital, will eventually influence product strategy, hiring, budgeting and organizational design across the business.
Because every AI decision now has both a technical cost and a business cost, engineering organizations will evolve around a different set of priorities. These include business outcomes, governance, reliability and trustworthiness (this is AI-specific) and productivity (through automation AI enables).
We will see an increase in new functions such as AI platform engineers, AI security engineers and AI operations specialists.
This may seem like the next wave of tech-influence roles in the mold of the site reliability engineer, the cloud deployment head or the DevOps lead. This is the next wave, but grounded in AI demand. And every engineer will develop skills in balancing technical excellence with business efficiency because the business demands it.
Where traditional measures focused on output, future measures must focus on customer value created per AI investment, as well as efficient use of compute, business impact fueled by AI and human judgment where and when it matters most. This is how organizations will compete.
Every AI strategy eventually becomes a workforce strategy.
But what about the rest of the business? Let’s game this out.
We already established that engineering success in this AI era hinges on AI literacy, systems thinking, architectural judgment and optimization skills. But with AI becoming so pervasive across businesses, the technology strategy must become inseparable from the people strategy.
Organizations must rethink hiring, internal mobility, performance expectations, learning and development and workforce planning. This includes knowing where to lean into AI for automation and intelligent operations, when to exercise human judgment and the wisdom to allocate both.
The ability to discern the right combination of humans plus machines to build solutions that solve real pain points for real people will ultimately be a winning strategy. This will create a competitive advantage for your business.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
