Applications of AI in Business

Introduction

  • A recap of AI fundamentals for MBA and executive audiences
  • Framing the managerial questions that will carry through the rest of the course
Knowledge Check

From a strategic perspective, why does AI matter to a firm's executive leadership?

Week 1

The Strategic Case for AI

"The executive question is not whether AI exists, but where it creates measurable business advantage."

  • Changes how firms make decisions, scale expertise, and redesign workflows.
  • Shows up through revenue growth, cost reduction, risk control, and faster learning cycles.
Knowledge Check

How is the most business value derived from AI systems?

Foundations

What AI Is, and Is Not

What It Is

Systems that perform tasks associated with perception, prediction, language, recommendation, or generation.

What It Is Not

Magic, autonomous wisdom, or a substitute for managerial accountability.

Key Takeaway

Most business value comes from pairing statistical systems with clear process design and human judgment.
Concept Map

The AI Family Tree

Artificial Intelligence

Artificial Intelligence is the broad umbrella that includes machine learning, deep learning, and generative AI.

Machine Learning

Machine learning focuses on learning patterns from data rather than following only hard-coded rules.

Deep Learning

Deep learning is a more specialized approach that becomes especially useful with complex data such as images, speech, and language.

Generative AI

Generative AI produces new text, images, code, or designs rather than only classifying or predicting.
Knowledge Check

Which business task involves AI assigning items to specific categories, such as sentiment or risk level?

Core Tasks

Three Business Tasks AI Does Best

Prediction

Prediction: estimating what is likely to happen next, such as churn, demand, or fraud.

Classification

Classification: assigning items to categories, such as sentiment, risk level, or document type.

Generation

Generation: creating drafts, options, summaries, or designs that accelerate human work.
Value Creation

The Business Value Chain of AI

1
The Catalyst
Improves a business decision, customer interaction, or operational workflow.
2
The Advantage
Comes from embedding technology into repeatable, proprietary processes.
3
The Outcome
Connects a model output to a clear action, owner, and measurable result.
Data

Data: The Fuel, the Constraint, the Differentiator

  • Depends entirely on access to relevant, timely, and trustworthy data.
  • Poor data quality can render a sophisticated model less useful than a simple heuristic.

Key Takeaway

The Differentiator: Firms with better data access, cleaner processes, and stronger governance often outperform those with more ambitious models.
Mechanics

How AI Systems Learn from Patterns

1. Training

Trained on historical examples to detect statistical regularities.

2. Application

Learned patterns are applied to new, unseen cases. Performance relies on condition similarity.

3. Limitation

Models do not understand context; they only optimize against the patterns they have seen.
Leadership

Where Human Judgment Still Matters

"The strongest operating model is not human versus machine, but human judgment amplified by machine speed."

The Executive Role

  • Define the goal and acceptable tradeoffs.
  • Assess the consequences of failure.
  • Govern decisions affecting reputation, fairness, safety, or capital allocation.
Discussion

Discussion: Human Judgment

Group Discussion

Group Discussion

A highly accurate AI model flags a long-time, high-value vendor for contract termination due to sudden risk indicators. Do you automate the termination to prevent immediate loss, or require human review despite the delay?
Reality Check

The Limits of AI in Real Organizations

Brittleness

AI systems can falter when customer behavior, market conditions, or internal processes unexpectedly change.

Opacity

Models can make confident but incorrect predictions for reasons that are entirely opaque to non-technical operators.

Key Takeaway

The Implementation Gap: Many failures stem from workflow misalignment, weak adoption, or unclear ownership rather than the algorithm itself.
Governance

The Tradeoff: Accuracy, Explainability, and Control

High
Accuracy
Deep Learning / Black Box
"High-performing models are not always the easiest to explain, audit, or govern."
High
Explainability
Decision Trees / Linear Models

Key Takeaway

The right balance depends heavily on regulation, customer trust, operational risk, and the reversibility of mistakes.
Discussion

Discussion: Accuracy vs Control

Group Discussion

Group Discussion

Your bank has two loan-approval models. Model A is a "black box" deep learning model that increases profit by 15%. Model B is a simple decision tree that increases profit by 5% but provides exact reasons for denial. Which do you deploy?
Looking Ahead

Bridge to Deep Learning

Richer Patterns

Deep learning extends our fundamentals into settings with images, language, and highly complex unstructured data.

Increased Complexity

Deeper models come with steep tradeoffs: higher compute costs, severe opacity, and challenging governance demands.
Before discussing advanced architectures, executives must clearly define the business problem the model is meant to solve.
Looking Ahead

Bridge to EDII

"Ask not only whether a model works, but for whom it works and who may be harmed."

  • AI systems actively reflect the assumptions, exclusions, and historical biases of their creators and data.
  • Equity, diversity, inclusion, and indigeneity are central governance requirements, not ethical add-ons.
Discussion

Discussion: EDII in Action

Group Discussion

Group Discussion

A new AI screening tool cuts hiring time by 40% and improves candidate quality metrics, but an audit reveals it systematically rejects resumes with large gaps in employment, disproportionately affecting mothers returning to the workforce. Do you pull the tool, or try to patch it while it runs?
Looking Ahead

Bridge to Sustainability

The Optimization Benefit

AI drives massive efficiencies by optimizing logistics, demand forecasting, energy usage, and resource allocation across the supply chain.

The Environmental Footprint

Training and deploying AI systems consume immense compute, infrastructure, and energy that carry significant environmental costs.

Key Takeaway

The managerial challenge is to rigorously evaluate the net impact: balancing the sustainability benefits AI creates against the footprint it imposes.
Looking Ahead

Bridge to Product Development

Scale
Ideation & Prototyping
Accelerating the lifecycle from generation through testing.
  • The core logic remains unchanged: value stems from pairing data, models, and human expertise around clear business objectives.
  • Our goal is establishing a managerial lens to judge where AI augmentation improves innovation versus where human design leadership is strictly required.
Knowledge Check

What is the recommended starting point for a strategic AI initiative?

Action

An Executive Adoption Playbook

1. Identify

Start with a specific business problem, not a vague ambition to "do AI."

2. Select

Choose use cases with accessible data, clear owners, and a measurable business outcome.

3. Execute

Pilot narrowly, learn quickly, and scale only when governance is securely in place.
Opportunity Map

A Functional Map of AI Opportunities

Marketing

Marketing can use AI for segmentation, content support, pricing insight, and demand forecasting.

Operations

Operations can use AI for scheduling, quality monitoring, inventory planning, and exception detection.

Finance and Risk

Finance and risk teams can use AI for anomaly detection, forecasting, compliance monitoring, and scenario analysis.

HR and Service

HR and service teams can use AI for talent support, knowledge retrieval, service triage, and workflow automation.
Checklist

Executive Questions Before You Scale

Key Takeaway

1. Accountability: What specific decision or workflow is being improved, and who owns the ultimate result?

Key Takeaway

2. Foundation: What data, governance process, and change management effort are required to succeed?

Key Takeaway

3. Risk: What would failure look like across operational, ethical, legal, and reputational dimensions?
Conclusion

From Curiosity to Competitive Advantage

"The goal of this course is to build managerial fluency, not technical theater."

  • AI becomes strategic when tied to core business priorities, operating discipline, and responsible leadership.
  • This foundation allows us to explore advanced methods and sector applications with sharp, critical judgment in the coming weeks.