Bricks and Mortar
AI Success Requires Human Hands
TL;DR: AI won't build the future. We will. But we need more than powerful algorithms or language models to do it right. We need leaders and architects with the know-how to build with AI. To do so, we need more investment in people, not just tech.
The future is not something that happens to us; it's something we build. When constructing with AI, we need to see it not as an endpoint but as a powerful set of tools in the hands of human builders.
As tech limitations become clear, expectations of fully automating complex tasks—from strategic planning and creative content generation to intricate data analysis—are unrealistic without humans in the loop. With that, the euphoric narrative is hitting a wall.
Skeptical voices are getting louder. Gartner predicts that by 2025, 30% of AI projects will be abandoned. Goldman Sachs adds to the pessimism, noting that generative AI tools are cumbersome and “not yet smart enough to make employees smarter.” Media reports paint OpenAI as an internal mess (despite a $6.6 billion funding round announced yesterday). Stock analysts recently downgraded Microsoft. The list could go on, but the meta-story is now tainted with caution.
Critics may be overlooking the bigger picture. While the hype is exaggerated, so is growing pessimism. The truth lies in the middle. We must balance AI’s growing power with people’s ability to produce with it.
It’s Time to Build. But How?
Manifestos from VCs with heavy bets in AI proclaim that it's Time to Build and evangelize a brand of Techno-Optimism. But what does that look like in practice? How do we build the future they envision?
Start by likening it to constructing a building. A skyscraper's beauty, strength, and function rely on solid architectural design. And it’s not just made of flashy glass and steel—it needs a solid foundation and essential elements like bricks and mortar to form durable, enduring structures.
Large language models are like building blocks. Microsoft Co-pilot is an example of how these models are now embedded in Word, Excel, and PowerPoint. Across Fortune 500 companies, generative AI now supports everyday tasks such as writing, editing, and planning. But general-purpose assistance isn’t enough. Other components—task-specific agents, small language models, and APIs—combine to make AI something to build with.
These AIs are only as useful as the “mortar,” the human bond that binds them into something meaningful. What gets built requires human intuition, practical experience, and commercial-grade know-how. Without human intelligence, AI is just a stack of potential with no direction, like a heap of bricks waiting for a blueprint.
Put another way, just as architects consider how a building fits its environment, leaders must consider how AI fits into their growth strategy and business functions.
Planning to do so gets complicated if you’re swayed by bubble-bursting speculation. Cynics suggest that AI is overhyped, overvalued, and overcapitalized. But zoom out and look at AI power curves. This perspective seems misguided. The actual value derived from these systems—the real story—is that people are underprepared, undervalued, and undercapitalized to use them effectively. If the AI bubble pops, don't blame the tech. Look to the human element.
Seeing Beyond the Hype
Writing for Harvard Business Review, Eric Siegel cautions against getting distracted by narratives and hype cycles. Blindly buying into AI without a clear plan is fraught, but so is sitting back and waiting for the hype to subside. Unlike the deployment of the Internet or social media, which took years to take hold, the rapid evolution of AI models leaves little room for complacency. Delays lead to being left behind as competitors discover new avenues for growth, efficiency, and accelerated decision-making.
Remember, we're in the early stages of AI deployment. AI in its various forms is still in its "DSL phase." Like dial-up modems, which gave way to broadband and laid the foundation for the Internet, bellwethers like NVIDIA, OpenAI, Microsoft, and Meta have built essential AI infrastructure for us to build on today.
Like how Cisco and Oracle shaped the early web, AI developers will create possibilities we can’t fully imagine. Think personalized education that adapts to each student, creative design tools that reshape entire industries, or simulated scenario planning that informs big decisions.
To reach potential destinations like these, it helps to investigate what may impede progress. Gartner's Hype Cycle suggests we've passed the peak of inflated expectations and are heading into disillusionment. If this is true, we must confront the human barriers that stand in its way.
Blocks to AI Adoption
Let's explore critical, human-centered considerations that might flatten the curve. Beyond the research cited by consultants, I’ve experienced each of these firsthand working with clients and teams.
Leadership Intent: While 94% of business leaders agree that AI is critical to success in the next five years, only 20% report that their organizations are "AI-ready,” according to a survey by Deloitte. This gap between acknowledgment and action highlights the need for a more substantial executive commitment.
Structure: McKinsey revealed that organizations successfully scaling AI are 3.5 times more likely to have C-suite executives working together on a transformation agenda. A collaborative approach driven by executive leaders is necessary to break down departmental barriers that lead to silo-busting pilots.
Procurement: According to EY, the number of companies investing $10 million or more in AI is set to nearly double next year to 30%, up from 16% currently investing at that level. However, despite the forecasted investment, the survey found that leaders ignore department-level needs that AI investments require to pay off.
Silver Bullet Mentality: Research published in MIT Sloan Management Review indicates that companies benefiting from AI are likelier to use multiple, specialized AI applications than a single, general-purpose system. This underscores the importance of adding task-specific AI tools over one-size-fits-all platforms.
Talent Networks: PwC’s Global Jobs Barometer found postings for AI jobs are growing 3.5x faster than all jobs, with AI skills commanding a 25% wage premium. This suggests a growing need for specialized talent, new roles, and partnerships to embed deep AI expertise in the business.
Staff Development: The World Economic Forum's Future of Jobs Report predicted that by 2025, while 85 million jobs may be displaced by automation, 97 million new roles are projected to emerge, “reflecting a shift in the division of labor between humans, machines, and algorithms.” This highlights the need for ongoing, collaborative development programs to integrate AI tools into workflows.
Employee Caution. According to a Thompson Reuters Institute study, 53 percent of professional services employees said they were hesitant, concerned, or fearful of using generative AI. Of those hesitant, 20% said they felt that way due to accuracy concerns, 18% said they were skeptical GenAI could deliver promised results, and 16% were concerned about over-reliance. 45% said they had no plans to use it.
Controlled Pilots: According to a report by BCG, companies that run multiple AI pilots and scale the successful ones are 3.5 times more likely to see financial benefits from AI. A pilot strategy allows organizations to test AI applications in controlled environments and identify use cases that inform full-scale implementation.
Strategic Evolution: MIT Sloan Management Review also found that companies with flexible, iterative approaches to AI implementation were 1.5 times more likely to report value from AI than those with rigid, top-down plans. This emphasizes the importance of collective learning from all parts of the organization to inform an agile AI strategy.
ROI: According to Gartner, 49% of survey participants identified the ability to accurately prove the value of AI projects as the most significant barrier to AI adoption. This highlights the need for alternative ways to assess the impact of AI investments, like return on the insight that makes businesses perform better.
Human-Centered Investment
Yuval Noah Harari, the renowned intellectual and AI critic, suggests that for every dollar and minute we invest in building AI, we should put an equal amount into developing ourselves. BCG takes the human thesis even further by taking a 10-20-70 approach that emphasizes algorithms (10%), tech and data (20%), and people and processes (70%).
A rebalanced approach is necessary to develop indispensable tools and successfully integrate them into the workplace. AI requires architects with business acumen to deploy them for practical use. Most importantly, AI needs human ingenuity and creativity to reduce workplace burnout, automate grunt work, and create the capacity to solve complex problems.
The future isn’t about AI replacing us; it’s about AI augmenting what we can build and make work better. As former IBM CEO Ginni Rometty said: “Some call this artificial intelligence, but the reality is this technology will enhance us. It will augment our intelligence.”
All this is a lot to take in. Still, if we’re overwhelmed by the present, we risk missing out on activating vast collective intelligence just a few prompts away. Creating space for organizational development is essential to building the future with AI. Without a practiced understanding of what it can do, you may find yourself sitting on the sidelines, blinded by the hype.