MCLI Crayon Edition: AI — Two Roads Diverged

MCLI Crayon Edition: AI — Two Roads Diverged

By Gene Fliman, Practice Director at MCLI

Everyone’s talking about “AI transformation.” But here’s the truth: there are only two roads ahead.

Road 1: Rehash the Past

This is what most executives are doing. They treat AI like another IT upgrade — the same way they treated ERP in the 2000s or CRM in the 2010s. Bolt AI onto broken processes. Automate bad workflows. Speed up garbage in, garbage out.

It feels safe. It looks like progress. But all it really does is institutionalize yesterday’s mediocrity.

We’ve seen this movie before:

• ERP systems that were supposed to reinvent finance ended up customized to death so they could preserve outdated approval chains.

• CRM rollouts that promised to transform customer relationships turned into bloated databases nobody wanted to touch.

• $2B power plants like Ivanpah sold as “transformations” collapsed under weak math and a refusal to face market realities.
  

By the time leadership changes come, it’s too late. The consultants have been paid, the execs exit with bonuses, and the rest of the org eats the loss. That’s not transformation. That’s adding another tombstone to an already crowded graveyard of failed initiatives.

Road 2: Break From the Pack

The minority path is harder — and it’s where the upside is.

This road starts by asking: What if AI isn’t just another tool for efficiency, but a chance to reinvent what we deliver, how we deliver, and who we deliver to?

• Instead of automating call centers, ask: Should customers even need to call us at all?

• Instead of bolting AI onto supply chain workflows, ask: Could we use it to design entirely new demand models?

• Instead of chasing chatbots, ask: What experiences could we deliver if we gave customers a truly intelligent front door?
  

That’s not an upgrade. That’s business reinvention. And it’s the only way AI becomes more than another failed capital project.

The Lesson We Refuse to Learn

Every time a high-profile project collapses, the industry calls it a lesson. But here’s the uncomfortable truth: the archive of lessons is already massive. ERP fiascos. Nuclear reactors. Energy grids. Solar plants. Banks that spent billions on tech only to deliver the same service slower and more expensively.

If you’re still adding to the archive in 2025, that’s not bad luck — it’s negligence. The patterns are clear. The fixes are known. Failure at this point is willful.

The Point

AI is not an upgrade. It’s a fork in the road.

Executives who use it to cement old processes will look busy for a few years — until competitors who reimagined their business models leave them behind.

Executives who treat it as an opportunity to reshape their future won’t just save money. They’ll own markets.

AI won’t make your business better if your business was broken to begin with. Reinvent first. Automate second.

--

• CIO.com. (2023). 18 famous ERP disasters, dustups and disappointments. Link

• Panorama Consulting. (2024). The Hidden Costs of ERP Failure. Link

• Davenport, C. (2016). Ivanpah, the Complex Solar Plant, Is Struggling to Live Up to Its Promise. The New York Times. Link

• Waldman, S. (2025). Older Ivanpah Solar Plant in California Will Close Units as Tech Shifts. Engineering News-Record. Link

• Associated Press. (2025). Ivanpah Solar Plant to Close Two Units as Cheaper Tech Prevails. Link

• Taleb, N.N. (2016). Big is Fragile: An Attempt at Theorizing Scale. arXiv preprint — argues large projects amplify fragility and collapse faster under deviations.

• Forbes. (2025). Gleeson, B. AI and the Future of Leadership in Tech and on Wall Street. Link — discusses risks of AI adoption without organizational change.

• McKinsey & Company. (2023). The State of AI in 2023. Link — highlights both opportunities and the failures of firms treating AI as “just another tool.”