Every few years, a new technology arrives that promises to change everything. Digital. Cloud. Blockchain. And now, generative AI.
Each time, the pattern repeats.
Headlines scream inevitability. Boards feel FOMO. Pilots proliferate. ROI disappoints. And suddenly the narrative flips from “this will change everything” to “this doesn’t work.”
AI is simply the latest chapter in a very familiar story.
The Problem Isn’t AI. It’s How We Approach Transformation.
Recent headlines claim that “95% of enterprise AI projects fail.” That number is provocative—and deeply misleading.
The MIT study behind it defined “success” as delivering measurable financial ROI within six months of a pilot. By that standard, most transformative IT initiatives in history would be labeled failures. Email. The internet. ERP. Cloud. None reliably delivered material ROI in their first six months.
The real issue isn’t whether AI “works.”
It’s whether organizations are approaching AI with critical thinking, scientific rigor, and historical memory—or whether they’re building strategy off headlines and hype.
AI is big, complicated, and nuanced. You cannot run your AI strategy on memes.
The Four Pillars of Real Transformation
Successful AI transformation rests on four pillars, not three:
- People – Skills, subject-matter expertise, judgment
- Process – How work actually gets done
- Technology – Code, platforms, models, infrastructure
- Culture – How change is accepted, resisted, or amplified
Culture matters independently. Sometimes changing culture changes people. Sometimes changing people reshapes culture. Ignoring culture guarantees failure.
Organizations that focus only on technology—without addressing process, people, and culture—don’t transform. They automate dysfunction.
Start With the Workflow, Not the Model
AI success flows from deep understanding of workflows, especially where those workflows bottleneck or fail today.
My guiding principle is simple:
Use code where code works best.
Use people where people work best.
Use AI where AI works best.
This sounds obvious—and yet it’s routinely ignored.
Too many AI initiatives try to replicate existing human workflows with AI agents, rather than reimagining the workflow itself. That’s like using a jet engine to power a horse carriage.
If a task is deterministic and repetitive, code beats AI.
If it requires judgment, empathy, or creativity, people beat AI.
If it involves pattern recognition, scale, or probabilistic inference, AI shines.
Real value comes when you redesign the workflow around these truths—not when you bolt AI onto a broken process.
We Learned This Lesson With the Cloud. Then Forgot It.
The cloud offers a perfect analogy.
Companies that “didn’t understand the cloud” simply lifted legacy systems and dropped them into AWS or Azure. Costs went up. Complexity increased. ROI disappointed.
Companies that did understand the cloud fell into two successful categories:
- Cloud-native organizations, architected from the ground up for cloud
- Cloud-enabled organizations, which selectively modernized workflows while maintaining legacy cores
Both groups understood the same thing: the cloud isn’t magic. Value comes from re-architecting workflows and operating models—not from relocation alone.
AI is no different.
Simply “moving work to AI” without redesigning how that work happens will produce the same disappointing results we saw with cloud lift-and-shift projects.
The Headline Trap—and the Discipline to Avoid It
We’ve seen this movie before:
- Every company needed an “e-” or “i-” prefix in the 1990s
- Every company needed a cloud strategy in the 2010s
- Every company flirted with blockchain in the late 2010s
- Now every company feels pressure to “do GenAI”
Each wave creates winners—but also massive waste—because too many organizations react emotionally instead of analytically.
The antidote is boring, disciplined, and effective:
- Ignore the headline
- Read the source
- Examine the methodology
- Apply critical thinking
- Go deeper before you go bigger
That’s how you go farther, faster, with fewer missteps and more ROI.
AI as Expansion, Not Replacement
The most powerful AI strategies don’t just reduce costs or replace labor. They expand the pie.
AI enables new workflows, new service levels, new products, and new business models—if you’re willing to rethink how work is organized.
This is augmentation, not substitution.
Humans plus AI, properly orchestrated, can deliver outcomes neither could achieve alone. But that requires intentional design, patience, and leadership—not panic-driven pilots.
The Bottom Line
AI is not a failure.
AI hype is.
Organizations that treat AI as a shortcut will be disappointed. Organizations that treat AI as a serious transformation discipline—grounded in people, process, technology, and culture—will win.
We’ve been here before.
We already know the playbook.
The only real question is whether we’ll remember it this time.
-Marc d. Paradis
About the Author: Marc d. Paradis’ professional journey is a fusion of academic rigor with real-world impact. He began his career over 30 years ago as an academic molecular neurobiologist, instilling in him a deep respect for critical thinking and the scientific method.
Transitioning into industry, he held leadership roles that bridged data and healthcare: as Vice President of Data Strategy at Northwell Health, Marc leveraged one of the world’s most diverse clinical data sets to drive patient-centered innovation via a $100M partnership with Aegis Ventures, launching multiple AI-centered startups; and as Vice President & Dean of Data Science University at Optum, he spearheaded the training of thousands of professionals in practical, product-centric AI, data-driven decision making, and ethical data practices. In each role, he fostered cultures of curiosity, critical thinking, and collaboration – precursors to the Constructive Inquiry ethos.
About SIYOM Consulting: Founded by Marc d. Paradis, SIYOM Consulting is a boutique advisory specializing in Data and AI Strategy for Healthcare and Life Sciences.
We help health-system executives, pharma innovators and investors identify, evaluate and execute on high-value data and AI opportunities.