Chapter 06 — Artificial Intelligence in Business#

Business Context#

The Monday Morning That Changed Everything#

Maya Garcia arrived at her marketing internship on a Monday morning ready for spreadsheets and social media posts. Her supervisor James asked for help explaining why online sales jumped 40% the previous quarter. “The IT team keeps talking about our new ‘AI-powered recommendation system,’ but I need someone to help me figure out what that actually means for our business strategy.”

Maya was a business major, not a computer science student — but she was already using AI before arriving at the office. Her GPS route? AI analyzing real-time traffic. Her credit card fraud alert? AI protecting her account. Her commute playlist? AI predicting what songs she’d enjoy.

Understanding the AI Landscape#

Maya discovered that Artificial Intelligence (AI) — computer systems capable of performing tasks that require human-like intelligence, such as learning and decision-making — was everywhere in business, often invisible but always working.

Machine Learning (ML) — a subset of AI where algorithms learn patterns from data to improve over time — powered their recommendation system. It wasn’t programmed with rigid rules; it observed patterns: customers who bought running shoes often browsed athletic wear; people purchasing baby products frequently returned within months.

Their analysts used to manually update product recommendation lists for hours each week. Now the system processed millions of customer interactions automatically — far more accurate than any human team could achieve.

Generative AI and Human-in-the-Loop#

The content team experimented with Generative AI — AI that produces new content (text, images, code) based on patterns learned from data. Content director Sarah was skeptical at first: “A computer can’t write with emotion.” But the tools generated three different marketing angles for a sustainable activewear line within seconds — leaving writers to refine voice and add the human touch.

The key principle: Human-in-the-Loop — AI system design that incorporates human judgment and oversight at critical decision points. The AI generates options; humans make the final call.

Operations and Predictive Analytics#

Operations manager David showed Maya how AI optimized inventory. UPS saves 100 million miles annually with AI-powered routing. Their company used similar principles: the AI analyzed historical sales, seasonal patterns, and local events. During a music festival, it automatically increased inventory for relevant products at nearby stores.

This was Predictive Analytics — using AI/ML to forecast trends and outcomes from historical data — reducing excess inventory by 25% while decreasing stockouts by 30%.

Chatbots and Natural Language Processing#

Customer service manager Tom showed Maya how their chatbot — an AI conversational agent used for customer service and support — handled 60% of inquiries without human intervention. It used Natural Language Processing (NLP) — AI technology that interprets and generates human language.

When a customer asked “Do you have this jacket in blue?”, the chatbot understood intent, checked inventory across all stores, and offered three options: online order with free shipping, in-store pickup, or similar blue jackets currently in stock. It learned from every interaction, routing unclear cases to humans — which became training data to improve the system.

Computer Vision in the Warehouse#

Computer Vision — AI that enables machines to interpret and analyze visual data such as images and video — reduced fulfillment errors by 40%. Cameras detected damaged products, identified when shelves needed restocking, and monitored for safety hazards.

The Dark Side: Bias and Governance#

Amazon built a recruiting AI trained on historical data that learned to penalize resumes mentioning “women’s” — reflecting existing industry bias. Apple Card’s algorithm was investigated for gender bias in credit limits.

This is why organizations need AI Governance — policies and frameworks guiding responsible AI adoption. AI systems reflect the biases in their training data. Before deploying systems affecting customers or employees, the company conducted bias audits, involved diverse design teams, and maintained human oversight for high-stakes decisions.

The Recommendation Engine That Changed Everything#

Their system analyzed behavior across multiple touchpoints: browsing, purchase history, wishlists, email click-throughs, and dwell time on products. Amazon generates 35% of sales from recommendations. This company went from 8% to 22% of sales attributed to AI recommendations in six months — transformational, not incremental.

A human salesperson can provide personalized service to perhaps 20 customers per day. The AI provided personalized experiences to thousands of customers simultaneously, 24/7.