🧠 A Simple Framework for Understanding AI’s Capabilities

Design: Canva relies on GenAI to generate layout options for text box positions with optimal font combinations and complimentary color palettes. For example, Canva’s Magic Design creates custom, on-brand, and attention-grabbing logos in seconds. Magic Design speeds up the creative process and inspires new artistic ideas but can’t replace a designer’s eye.

Finance: Bloomberg leverages GenAI to provide real-time financial news. For example, BloombergGPT analyzes a company’s earnings announcements paired with industry-specific macroeconomic trends to highlight complex stock market behaviors. BloombergGPT helps financial analysts make quicker, more informed financial decisions but can’t replace a financial analyst’s human judgment around unpredictable events like a political crisis.

Healthcare: Rad AI uses GenAI and CNNs to empower physicians by generating automated radiology reports. For example, Omni Unchanged extracts stable findings from a patient’s prior radiology checkups and inserts them into a radiologist’s existing preferred report template. Omni Unchanged speeds up radiology reports but can’t replace a radiologist’s expertise in making a final diagnosis.

Design: Editing platforms like Adobe Photoshop use CNNs to automate complex image editing workflows. For example, suppose a designer uploads a portrait and wants to make it appear happier. In that case, Adobe’s Neural Filters adjust the portrait’s mouth curvature, eye openness, and cheek elevation without requiring inputs from the designer for each facial adjustment.

Finance: Cybersecurity platforms like Darktrace monitor financial transaction data for unusual occurrences. For example, if a customer’s financial account shows activity in two countries within minutes, Darktrace automatically freezes the customer’s financial account and alerts the Fraud Team.

Healthcare: Administrative platforms like Nuance automatically translate medical records into billing codes based on the International Classification of Diseases (ICD). For example, if a doctor diagnoses Type 2 Diabetes and prescribes Metformin to treat high blood sugar levels, Nuance’s Clintegrity recognizes the keywords and assigns the ICD code E11.9.

Banking: J.P. Morgan’s LOXM autonomously executes security trades (i.e., stocks, bonds, or options) as fast as possible at the optimal market price. In financial trading, the Bid/Ask Spread (i.e., the difference between a buyer’s preferred prices and a seller’s preferred prices), Order Volume, and Market Volatility determine the Price and Speed of each financial transaction. Leveraging RL, LOXM trains on billions of historical financial transactions. When deployed, it dynamically adjusts Limit Orders (i.e., a bid to buy a security below a certain price or a bid to sell a security above a certain price) to execute as many financial transactions at preferred prices as possible.

Self-Driving Cars: Waymo’s Jaguar I-PACE SUVs rapidly adapt to changes in the driving environment. For example, when a scooter rider fell in front of a moving Waymo, the RADAR, LIDAR, and 360-Degree Cameras on the Jaguar I-PACE SUV generated a 3D Map of the current driving environment, identifying the fallen scooter rider and their impact trajectory. Using PPAs, it adjusted the steering wheel, altered the current speed, and deployed braking methods to safely swerve around the scooter rider without hitting them.

Customer Service: Intercom’s Fin is an AI Agent that resolves over 50% of customer questions without human intervention. For example, imagine if a customer asks: “Why is my payment method failing?” Fin first searches internal knowledge bases like Help Center articles and FAQ sections to identify potential causes. Then, it suggests corrective steps like: “Check your credit card’s expiration date.”