Why Uncertainty Changes How IT Must Reason

2026-04-21 · Source: Featured Blogs - Forrester · Field: Technology & Digital — Software Development & Engineering, Data Science & Analytics · Depth: Intermediate, medium

Summary

Murray Cantor, in conversation with Charles Betz, advocates for reframing technical debt as economic liability and embracing uncertainty as a core feature of IT investment. He argues that this perspective inexorably leads to Bayesian thinking, where initial estimates serve as priors updated with new information. Cantor highlights that computational feasibility, such as running Bayesian analysis and Monte Carlo simulations on a MacBook Air, has removed the historical barrier to adopting these methods. He emphasizes that sparse data is not a showstopper, suggesting Bayesian nets can achieve results with significantly fewer observations than frequentist approaches. The discussion draws an analogy between IT portfolio management and meteorology, explaining how ensemble methods are used to generate probability distributions for hurricane forecasts, a principle applicable to managing IT uncertainty. Cantor also critiques the agile movement for failing to formalize uncertainty, advocating for "better estimates" through probabilistic reasoning rather than "no estimates." He concludes that the necessary math, data from observability platforms, and computational tools are now readily available, making probabilistic reasoning a practical necessity for high-stakes IT decisions.

Key takeaway

For VPs of Engineering or CTOs managing large IT portfolios, adopting probabilistic reasoning is crucial. Your organization should move beyond point estimates and embrace Bayesian models to quantify and manage uncertainty in IT investments. This shift, supported by modern observability data and computational tools, enables more adaptive decision-making and better risk management, ultimately improving outcomes when plans inevitably diverge from reality.

Key insights

Embrace Bayesian thinking to manage IT investment uncertainty, leveraging available data and computational tools.

Principles

• Think like an investor.
• Embrace uncertainty.
• Take a systems perspective.

Method

Utilize Bayesian thinking to update initial estimates (priors) with new data. Employ Bayesian nets for sparse data and ensemble methods to model uncertainty, generating probability distributions for outcomes.

In practice

• Use Monte Carlo simulations for arithmetic.
• Explore AgenaRisk for Bayesian net modeling.
• Measure uncertainty using probability theory.

Topics

• Bayesian Thinking
• Uncertainty Management
• IT Investment
• Probabilistic Reasoning
• Technical Debt

Best for: VP of Engineering/Data, CTO, Consultant

Related on AIssential

• Teaching LLMs to reason like Bayesians — Training LLMs to mimic optimal Bayesian models significantly enhances their probabilistic reasoning and generalization capabilities.
• Concerning Uncertainty -- A Systematic Survey of Uncertainty-Aware XAI — UAXAI integrates uncertainty into AI explanations to improve transparency, reliability, and human trust in high-stakes decision-making.
• 80% of Data Science Probability Interview Questions Come From This 20% — Probability in Data Science interviews is about decision-making under uncertainty, not just formulaic calculation.
• Baseten CEO Tuhin Srivastava on Custom Models, and Building the Inference Cloud — The AI inference market is experiencing explosive growth, driven by custom models and a severe, persistent compute supply crunch.
• A Gentle Introduction to Stochastic Programming — Stochastic programming integrates uncertainty into linear programs, offering robust, probabilistic, or multi-stage recourse solutions.
• Navigating Common Pitfalls in Data Science: Lessons from Pierpaolo Hipolito - ML 183 — Understanding data paradoxes and causal reasoning is crucial for building accurate and interpretable machine learning models.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Featured Blogs - Forrester.