Winning With AI: The Next Phase of Work

Key Concepts

• Mission-Critical IT: Infrastructure that is essential for business operations (e.g., airline reservation systems, banking mainframes, hospital records).
• Agentic AI: AI systems capable of performing tasks, making decisions, and chaining together technologies to achieve outcomes with minimal human intervention.
• Proof of Outcome: A shift from "Proof of Concept" (which implies potential failure) to a focus on delivering measurable business results.
• Uncertainty as a Skill: The ability to question assumptions, remain curious, and navigate incomplete information, which is increasingly vital as AI provides "certain" but potentially flawed answers.
• Structured vs. Unstructured Data: The shift toward using raw, unstructured data to avoid the "thumb on the scale" bias inherent in pre-structured datasets.
• Private Cloud: A growing trend where enterprises maintain cloud-like infrastructure on-premises for security, regulatory, or performance reasons.

---

1. The Role of Kyndryl and Mission-Critical IT

Jamie Rutledge describes Kyndryl as a "four-year-old startup" managing the world's most vital infrastructure. The company operates on the principle that mission-critical IT should be "invisible"—it must function perfectly to keep airlines flying, banks processing payments, and hospitals managing patient data.

• Complexity as an Asset: Kyndryl thrives on complexity. While simple tasks are easily commoditized, complex, legacy-heavy environments (like those running 50 million lines of COBOL code) require specialized management.
• The "Race Car" Analogy: Kyndryl’s value proposition is the ability to "change the engine and tires while the race car is moving at full speed," now accelerated by AI.

2. AI Strategy and Implementation

Rutledge emphasizes that the "tech works"—the challenge is no longer the capability of AI, but the how of implementation at scale.

• Customer Zero: Kyndryl uses its own internal tools as a testing ground for AI before deploying them to clients.
• From "What/Why" to "How": Clients have moved past asking what AI can do; they now demand actionable strategies for rapid implementation and speed.
• Productivity Gains: Rutledge promises clients a 30–40% reduction in costs over three years through AI, but notes that the goal is not just cost-cutting—it is freeing up human talent to tackle massive backlogs of innovation projects.

3. The Human Element: Skills and Culture

A central theme of the discussion is the evolution of the IT professional in an era of rapid change.

• The Value of Uncertainty: As one rises in an organization, the need to be an "expert" (which implies certainty) becomes a liability. Rutledge argues that the most valuable leaders are those who are comfortable with uncertainty, as this fosters curiosity and prevents the rigidity that stifles innovation.
• The "Death of the Individual Contributor": With AI agents handling routine tasks, the human role shifts to managing, instructing, and auditing the work of agents.
• Continuous Learning: Rutledge highlights the necessity of "unlearning" outdated methods. He shares an anecdote about hiring in India, where he realized he had to change the teachers because the existing training curriculum was 20 years out of date.

4. Risk, Guardrails, and Data

The transition to AI introduces new threat vectors, particularly when agents can chain together disparate technologies to find vulnerabilities.

• The Guardrail Problem: There is currently no standardized way to discover and govern all AI agents within an enterprise. Advanced models (like the hypothetical "Mythos") may ignore traditional guardrails, creating significant security risks.
• Data Integrity: Rutledge warns against the common mistake of using only structured data. By forcing data into a structured format, humans often "put their thumb on the scale" to support a pre-existing hypothesis. He advocates for using unstructured data to uncover genuine anomalies and insights.

5. Synthesis and Conclusion

Winning with AI is defined by growth and margin expansion. Rutledge posits that we are currently in a period of "slow" change compared to what is coming. Companies that successfully crack the code on adopting AI at scale will "lap the field," creating a competitive advantage that will be impossible for others to overcome.

Key Takeaway: The future of business is not about replacing humans with AI, but about cultivating a workforce that is curious, comfortable with uncertainty, and capable of managing AI agents to solve complex, mission-critical problems. As Rutledge notes, "The technology is no longer the gating factor; it’s how comfortable you can get with the pace and scale of AI."