Skip to main content

Classifications

Classification based on how AI learns and improves over time.

  • Reactive AI
    AI that can only react to situations based on predefined rules. It does not learn from past experiences. IBM’s Deep Blue (beat world chess champion Garry Kasparov), spam filters, rule-based fraud detection.
  • Limited Memory AI
    AI that learns from past data and improves over time but does not retain long-term knowledge. Self-driving cars (Tesla Autopilot), AI-powered chatbots, machine learning recommendation systems.
  • Theory of Mind AI
    Hypothetical AI that can understand human emotions, beliefs, and social interactions. Still under research; could be used for AI assistants with emotional intelligence.
  • Self-Aware AI
    AI with consciousness and self-awareness, capable of forming thoughts and emotions. Does not exist yet; purely theoretical.