The AI reskilling paradox: Slow down to get there faster 

The four zones of AI readiness

Every organization sits somewhere on a simple grid: infrastructure readiness (data, tooling, skills) multiplied by urgency (how hard leadership is pushing). There are four zones:

• Stagnation (low urgency/low readiness): low innovation and little progress.
• Complacency (low urgency/high readiness): resources available but no drive; potential sits idle.
• Frustration (high urgency/low readiness): lots of action without infrastructure; spinning wheels.
• Innovation (high urgency/high readiness): upskilling success and balanced momentum.

Let’s look in more detail and chart the moves that pull you toward innovation:

Stagnation Zone: drifting toward irrelevance

In the Stagnation Zone, urgency is minimal and readiness is lacking. These are the companies that “wait and see” while the world changes. They dabble with pilots (if any) and lack data governance, cloud foundations, or skills. By the time they “see” a clear impact, hungrier rivals have lapped them. As Cisco’s Jeetu Patel warns, “Eventually there will be only two kinds of companies: those that are AI companies, and those that are irrelevant.”

Escape move: inject urgency and pick a first real use case. Make a small, visible bet that proves momentum is possible. Then use it to unlock budget for plumbing (data standards, shared access, basic governance). The goal isn’t a headline; it’s a working pattern that you can repeat.

Complacency Zone: big resources, but little fire

The Complacency Zone is, in some ways, the most tempting and insidious of all. Here, a company has built high readiness. Maybe they spent years on digital transformation, possess modern data centers, employ talented engineers, and have even installed AI tools. But there is low urgency to put them to bold use.

Escape move: manufacture urgency without theatrics. Set bold, public targets (e.g., “20% of customer contacts assisted by AI within 12 months”). Run challenge programs or hackathons to channel energy into real business problems. DHL does this through an internal startup lab that gives intrapreneurs the time, money, and skills to build. Pair the fire with two-way doors (reversible experiments): make it easy to try and safe to roll back. The more reversible the decision, the faster it should move.

Frustration Zone: when hype runs ahead of infrastructure

This is the 2025 default: high urgency, low readiness. Leaders decree, “AI everywhere!” Hackathons sprout. Consultants swarm. Then reality: data is siloed or dirty, teams aren’t trained, risk is unclear. Enthusiasm turns to whiplash. One telling stat: while nearly everyone invests in AI, close to 80% admit major gaps in data preparation.

Escape move: slow down to speed up. Redirect urgency into building capacity:

• Fix the plumbing. Standardize definitions, clean key tables, and move from 50 Excel islands to a simple, searchable cloud backbone. No data, no sandbox, and no upskilling.
• Create safe sandboxes. Put real tools next to real (but low-stakes) data. Make it easy to spin up a pilot, and just as easy to shut it down. DHL’s pilot funnel is a great model: prove value with a team, then widen the aperture.
• Reskill by doing. Lectures don’t make muscles. Give teams two-week sprints to solve a live problem with AI. Celebrate prototypes, not PowerPoints.
• Govern with doors. Label decisions as two-way (reversible: try fast) or one-way (hard to undo: review deeply). Don’t let one-way rigor slow every two-way idea.

The Frustration Zone ultimately is a transitory phase; or it should be. No company can sustain frenetic AI hype without results for long. You either regress (give up on AI initiatives, sliding toward stagnation), or you level up by building the muscle memory and infrastructure that convert high urgency into high readiness. The companies that successfully navigate through frustration emerge into the promised land: the Innovation Zone.

Innovation Zone: where transformation happens

Here, urgency and readiness meet. Executives don’t just talk; they practice, and they fund the boring stuff (data, access, MLOps). Middle managers are linchpins, translating strategy into sprints and fighting for data access. Frontlines are empowered to propose and run micro-sandboxes that solve real work problems. Governance is clear enough to go fast and strong enough to keep you safe.

Here are two telltale behaviors:

1. Scale conditions, not just solutions. When a plant manager halves defects with an AI assist, give them a trophy, then scale the conditions that made it work: shared data, the sandbox, and the budget line. That’s how wins become a flywheel.

Take DHL again as an example of an innovation mindset: they run structured pilots for genAI, but they also tie every experiment to business needs (better customer proposals, streamlined data handling, etc.). And Shopify, the e-commerce company, recently required all employees to start using AI tools – a bold move signaling that AI isn’t the domain of a few specialists but part of everyone’s job.

2. Add the missing role: the scout. Someone must constantly tap early alpha/beta models, bring them into the sandbox, and brief leadership on what’s real vs. hype. As Coca-Cola’s Thakar describes, part of his job is reaching out to AI innovators for early access, so Coke knows what’s coming and where to experiment next. The company formed a global partnership with OpenAI and Bain to explore the application of AI across its operations.

In the Innovation Zone, we also see a clear delineation of decision types. Leadership “think in doors”:

• Two-way doors: reversible experiments – change a price band in one region, try an LLM to triage emails, pilot a pattern change on packaging. Move fast. If it flops, walk back.
• One-way doors: irreversible bets – AI in safety-critical manufacturing, new product formulations, public-facing automation in regulated markets. Move deliberately and test in the sandbox until you’re confident.

This shared language enables everyone to move faster with fewer meetings, as the risk is explicit.

Everyone has a role

As the AI buzz continues to reverberate through every industry, it’s easy to get swept up in either panic or paralysis. But the stories above carry a clear message: the winners won’t be the loudest, they’ll be the most prepared to act decisively.

• C-suite and AI wizards: set the ambition. Fund the plumbing. Appoint your Scout. Make two-way vs. one-way doors the default governance language. Model the behavior by using AI yourself.
• Middle managers: you’re the force multiplier. Fight for data access, create team sandboxes, and pick crisp, reversible experiments tied to your KPIs. Coach the habits of using AI, and don’t just talk about it.
• Frontline and individual contributors: if there’s no official sandbox, build a micro one. Start with public or non-sensitive data, propose a tiny pilot, and document results. Tinker, share, and repeat. If your environment is truly stifling, the skills you build now will be valued elsewhere.

When the hype dust settles, the organizations that practice, not just preach, will still be standing, stronger. Real transformation happens when infrastructure meets urgency. Get that right, and you’ll find yourself in the only place that compounds results.

Jialu Shan, Lawrence Tempel, and Alexandre Sonderegger contributed to this article.