Skip to content

Latest commit

 

History

History
135 lines (107 loc) · 4.31 KB

Artificial Intelligence.md

File metadata and controls

135 lines (107 loc) · 4.31 KB
Raw

#AI

Artificial intelligence is system of simulation of human intelligence by a machine.

AI is of two types:

  • General AI (general reasoning abilities in a wide range of fields)
  • Narrow AI (specialized in solving specific tasks)

Machine Learning is a subset of AI where machines learn from data, recognize patterns and make decisions without being programmed explicitly.

Deep learning is a subset of ML which uses artificial Neural Networks to analyse large data.

Artificial Neural Networks (ANN) mimics the structure and function of biological human brain to solve tasks.

Natural Language Processing (NLP) generates and understands human language.

Prompt Engineering

Prompt engineering is the art and science of crafting effective inputs (prompts) to get the best possible outputs from AI models. It's similar to learning how to ask questions in a way that gets you the most helpful answers.

Common Techniques

  • Chain-of-thought prompting: Asking the AI to explain its reasoning step by step
  • Few-shot learning: Providing examples of desired inputs and outputs
  • Role prompting: Giving the AI a specific perspective to work from
  • Using guardrails: Setting clear boundaries and constraints

Best Practices

  • Be explicit about what you want
  • Provide context and background
  • Iterate and refine based on responses
  • Test different approaches to see what works best

Claude AI's Definition

Artificial Intelligence (AI) is the field of computer science focused on creating systems that can perform tasks typically requiring human intelligence. These systems aim to mimic human cognitive functions like learning, problem-solving, pattern recognition, and decision-making.

Key Subdomains of AI

1. Machine Learning (ML)

  • Core concept: Systems that learn and improve from experience without explicit programming
  • Types:
    • Supervised Learning: Learning from labelled data
    • Unsupervised Learning: Finding patterns in unlabelled data
    • Reinforcement Learning: Learning through trial and error with rewards/penalties

2. Deep Learning

  • A subset of ML using neural networks with multiple layers
  • Specializes in:
    • Image/video processing
    • Natural language processing
    • Speech recognition
    • Complex pattern recognition

3. Natural Language Processing (NLP)

  • Focuses on interaction between computers and human language
  • Applications:
    • Machine translation
    • Text summarization
    • Sentiment analysis
    • Question answering
    • Text generation

4. Computer Vision

  • Enables machines to understand and process visual information
  • Applications:
    • Object detection
    • Face recognition
    • Image segmentation
    • Scene understanding
    • Visual search

5. Robotics

  • Combines AI with physical machines
  • Includes:
    • Motion planning
    • Perception
    • Manipulation
    • Navigation
    • Human-robot interaction

Types of AI

ANI (Artificial Narrow Intelligence):

  • Also called "Weak AI"
  • Current state of AI technology
  • Designed for specific tasks or narrow domains
  • Examples:
    • Chess engines (like Deep Blue)
    • Image recognition systems
    • Virtual assistants (Siri, Alexa)
    • Recommendation algorithms
    • Language models (like myself)
  • Characteristics:
    • Excels in specific domains
    • Cannot transfer learning to new domains
    • No true understanding or consciousness
    • Limited to programmed or trained functions

AGI (Artificial General Intelligence):

  • Also called "Strong AI" or "Human-level AI"
  • Currently theoretical/not yet achieved
  • Matches or exceeds human-level intelligence
  • Capabilities would include:
    • Self-awareness
    • Common sense reasoning
    • Transfer learning across domains
    • Abstract thinking
    • Consciousness (debated)
    • General problem solving
    • Emotional intelligence
    • Creative thinking
    • Learning from limited examples

Key Differences:

  1. Scope
  • ANI: Single domain or task-specific
  • AGI: Multiple domains, like human intelligence
  1. Learning Ability
  • ANI: Learns within predefined parameters
  • AGI: Can learn and adapt like humans
  1. Problem Solving
  • ANI: Solves specific types of problems
  • AGI: Can solve any problem a human can
  1. Consciousness
  • ANI: No self-awareness
  • AGI: Potentially self-aware (debated)
  1. Flexibility
  • ANI: Limited to trained scenarios
  • AGI: Can handle novel situations