What is Artificial Intelligence?

An AI presentation from SAS 2. Understand Context Learn Patterns Recognize Objects Artificial Intelligence is the science of training systems to emulate human tasks through Learning and Automation 3. Evolution of Artificial Intelligence Neural Networks 1950s-1970s Machine Learning Deep Learning and Cognitive Systems 1980s-2010s Present Day 4. Today, AI can be both Threat and Opportunity Gain Competitive Edge Find Growth Trends Customer Centricity New Capabilities Efficiency in Process Process Elimination Workforce Transformation Reduced Time to Value Reduced Cost Improved Margin THREAT OPPORTUNITY RATIONAL Lose Competitive Edge Miss Market Trends Reduced Customer Engagement IRRATIONAL Massive Job Loss Robots Replace Humans We Lose Control 5. Rogers Telecom 53% fewer customer complaints SciSports Pro Sports 200,000 players analyzed to find the next star Honda Manufacturing 60 secs to identify suspicious claim WildTrack Data for Good 90% accuracy for ID of wildlife using tracks SunTrust Financial Services 90% improvement in response rate How we're working with customers today to make AI an Opportunity 6. How does artificial intelligence work? 7. US GOOD AT COMMON SENSE INTUITION CREATIVITY EMPATHY VERSATILITY MACHINES GOOD AT LARGE DATA SETS COMPLEX CALCULATIONS LEARNING AUTOMATION What are we and machines good at? 8. AI enhances our capability and gives organizations competitive advantage What are we and machines good at? Machine Learning (Deep Learning) More compute power DEEP LEARNING Larger data sets More complex models Better algorithms 11. Natural Language Natural Language Processing Natural Language Understanding Natural Language Interaction Natural Language Generation 12. Predict Target Value Time Series Variability Automate Process Account for Possible Actions Assimilate Trade-Offs Define Critical Constraints Forecasting Optimization Increase Robustness Analyze Complexity Handle Uncertainty SCALE SOLUTIONS Forecasting and Optimization 13. Energy Forecasting Use short and long- term variability to improve accuracy Deep Learning 16.