Embedded Files
Modern AI Categories
Modern AI Categories
A. By Application / Purpose
A. By Application / Purpose
Generative AI
Generative AI
Creates new content (text, images, music, code).
Examples: ChatGPT, Midjourney, DALL•E, Stable Diffusion.
Predictive AI
Predictive AI
Analyzes data to forecast outcomes.
Examples: Stock prediction models, weather forecasting, demand planning tools.
Conversational AI
Conversational AI
Engages in natural language dialogues.
Examples: Customer service chatbots, virtual assistants like Alexa, Siri.
Recommendation AI
Recommendation AI
Suggests products, movies, music, etc. based on user behavior.
Examples: Netflix recommendations, Amazon product suggestions.
Computer Vision AI
Computer Vision AI
Interprets images and videos.
Examples: Facial recognition, object detection, medical imaging analysis.
Autonomous AI
Autonomous AI
Acts independently in the physical world.
Examples: Self-driving cars, delivery robots, drones.
Expert Systems
Expert Systems
Uses rule-based logic for decision-making in specific fields.
Examples: Medical diagnosis systems, legal research AI.
Robotics AI
Robotics AI
Combines physical robots with AI to perform complex tasks. Examples: Factory automation robots, surgical robots.
B. By Technology Approach
B. By Technology Approach
The following are the main technology approaches that power modern AI systems:
- Machine Learning (ML)
- Deep Learning
- Natural Language Processing (NLP)
- Computer Vision
- Reinforcement Learning
- Knowledge-Based Systems
- Multi-Agent Systems
Generative AI
Generative AI
Generative AI refers to any artificial Intelligence intended to create something new and original, unlike cataloging or analyzing pre-existing data.
The mechanism functions by learning patterns from huge sets of data and then leveraging that knowledge to generate outputs that imitate work done by humans.
Key Points
Key Points
Purpose: Generating new content data, be it text, images, music, code, video, or 3D models.
How it Works: By uses of advanced machine learning models-often deep learning and neural networks-Generative Adversarial Networks, or GANs, and transformer models.
Output: Has the potential of being highly realistic, creative, and directly relevant.
Examples
Examples
Text: ChatGPT generating essays, stories, or answers.
Images: Midjourney, DALL·E creating illustrations from text prompts.
Music: AI tools composing original songs.
Code: GitHub Copilot writing programming code.
Video: Runway generating video clips from descriptions.
Applications
Applications
Creative industries, art, design, writing
Marketing and advertising
Game development
Education, content creation
Product, prototyping
Predictive AI
Predictive AI
Predictive AI is a kind of artificial intelligence that looks at past and current data to predict future outcomes or trends.
It doesn’t generate new content like Generative AI. Instead, it focuses on probability and pattern recognition to make informed predictions.
Key Points
Key Points
Purpose: Anticipate what is likely to happen in the future.
How it Works: Uses statistical algorithms, machine learning models, and sometimes deep learning to find patterns in large datasets.
Output: Predictions, probability scores, and trend forecasts.
Examples
Examples
Business: Predicting sales trends or customer churn.
Weather: Forecasting rain, storms, or temperature changes.
Healthcare: Predicting disease risk based on patient data.
Finance: Stock price forecasting, fraud detection.
Logistics: Demand forecasting for supply chains.
Applications
Applications
Risk assessment and fraud prevention
Market trend analysis
Predictive maintenance in manufacturing
Personalized recommendations (when combined with recommendation engines)
Early warning systems in disaster management
Conversational AI
Conversational AI
Conversational AI is a kind of artificial intelligence that interacts with people through natural, human-like conversations. This can happen through text, voice, or both.
It brings together Natural Language Processing (NLP), Natural Language Understanding (NLU), and sometimes speech recognition and synthesis to understand what users say and respond in a suitable way.
Key Points
Key Points
Purpose: Enable smooth, context-aware communication between humans and machines.
How it Works:
Input Understanding, AI processes user queries (text or voice).
Intent Recognition, identifies what the user wants.
Response Generation, produces a relevant, often natural-sounding reply.
Examples
Examples
Customer Support: Chatbots on websites or apps.
Virtual Assistants: Alexa, Siri, Google Assistant.
Help Desks: AI-driven live chat agents.
Language Learning: Duolingo’s AI tutor.
Applications
Applications
24/7 customer service
Voice-controlled devices
Interactive education tools
Booking systems (flights, hotels, services)
Personal productivity assistants
Recommendation AI
Recommendation AI
Recommendation AI refers to artificial intelligence (AI) technology that uses a user's behavior, preferences, and patterns to recommend products, services, or content that are relevant to them.
The aim of recommendation AI is to add personalization to the user experience by predicting what a person is most likely to be interested in next.
Key Points
Key Points
Objective: Assist users in locating sample items that they would want or require.
Mechanism:
Data Collection - Collects user data - past purchases, clicking, rating, watching histories, etc.
Pattern Finding - Looks for patterns between users or items.
Recommendation - Recommends items by relevancy.
Methods include:
Collaborative Filtering (which finds user patterns from other similar users),
Content-Based Filtering (matching item features to user profile preferences),
Hybrid Systems (mixing several approaches).
Examples
Examples
Shopping: Amazon’s “Customers who bought this also bought…”
Streaming: Netflix or Spotify suggesting shows or songs.
Social Media: YouTube recommending videos.
News: Google News tailoring headlines for you.
Applications
Applications
E-commerce personalization
Media and entertainment
Travel and hotel suggestions
Online education content recommendations
Targeted advertising
Computer Vision AI
Computer Vision AI
In essence, Computer Vision AI means that machines have the ability to see, interpret, and understand visual information from the world - search images, videos, and live camera feeds.
More specifically, Computer Vision AI attempts to replicate the same capability of seeing that humans possess, as well as potentially altering, and/or improving these capabilities through computer vision AI algorithms, and machine learning models.
Key Points
Key Points
Purpose: Extract meaningful insights from visual data.
How it Works:
Image Acquisition – Captures images or videos.
Preprocessing – Enhances and cleans the data.
Feature Extraction – Identifies shapes, colors, objects, or patterns.
Interpretation – Classifies or analyzes what’s in the image.
Often powered by Convolutional Neural Networks (CNNs) and deep learning.
Examples
Examples
Facial recognition systems
Object detection in autonomous cars
Medical image analysis (X-rays, MRIs)
Quality control in manufacturing
Wildlife monitoring through camera traps
Applications
Applications
Security & surveillance
Healthcare diagnostics
Retail checkout automation
Augmented reality (AR)
Industrial inspection
Autonomous AI
Autonomous AI
Autonomous AI is a type of artificial intelligence that can perform tasks and make decisions without the need for human input on an ongoing basis, which is often in dynamic and real-world environments.
It integrates perception, decision-making, and action by being free to function autonomously and change with the situation.
Key Points
Key Points
Purpose: Allow machines or systems to act on their own to achieve specific goals.
How it Works:
Perception – Collects information from sensors (cameras, lidar, GPS, etc.).
Decision-Making – Uses AI models to analyze data and choose actions.
Action Execution – Controls motors, systems, or processes to carry out decisions.
Feedback Loop – Learns and adjusts based on outcomes.
Often integrates Computer Vision, Predictive AI, and Reinforcement Learning.
Examples
Examples
Self-driving cars
Delivery drones
Autonomous underwater vehicles
Industrial robots in smart factories
Mars rovers
Applications
Applications
Transportation & logistics
Agriculture (autonomous tractors, harvesters)
Military & defense operations
Search and rescue missions
Space exploration
Expert Systems
Expert Systems
An Expert System is a type of artificial intelligence that attempts to emulate the decision-making ability of a human expert in a particular area of knowledge. Expert Systems contain a knowledge base (factual and rule-based information about the subject), and an inference engine (the logic to apply the rules) and can be used to solve problems, recommend actions, or diagnose problems.
Key Points
Key Points
Purpose: Provide expert-level advice or decisions in a narrow field.
How it Works:
Knowledge Base – Stores facts, data, and rules from experts.
Inference Engine – Applies logical reasoning to draw conclusions.
User Interface – Lets humans input questions and receive advice.
Often rule-based and deterministic, but can integrate machine learning for improvement.
Examples
Examples
Medical Diagnosis: MYCIN (early AI for bacterial infection diagnosis).
Technical Support: Troubleshooting systems for electronics.
Business: Loan approval or risk assessment tools.
Legal: Law research assistants.
Applications
Applications
Healthcare decision support
Engineering problem-solving
Customer service troubleshooting
Agriculture crop management advice
Financial risk analysis
Robotics AI
Robotics AI
Robotics AI is the application of artificial intelligence to robotics so that robots perceive, decide, and act in a way that gives them abilities to execute tasks autonomously or semi-autonomously.
Robotics AI brings together mechanical engineering (the physical robot) and artificial intelligence technologies (e.g. natural language processing, computer vision, and machine learning) to enable the robot to exist in a dynamic environment to accommodate new/unexpected change to the robot's environment.
Key Points
Key Points
Purpose: Give robots the ability to interact intelligently with their environment.
How it Works:
Perception – Uses sensors and cameras to gather data about surroundings.
Processing & Decision-Making – AI algorithms interpret the data and choose actions.
Action Execution – Physical actuators and motors perform the chosen actions.
Learning & Adaptation – Improves performance over time through feedback.
Examples
Examples
Factory robots assembling cars.
Surgical robots assisting in medical procedures.
Warehouse robots sorting and moving goods.
Service robots in hotels or restaurants.
Humanoid robots like Boston Dynamics’ Atlas or Honda’s ASIMO.
Applications
Applications
Manufacturing automation
Healthcare surgeries and patient care
Logistics and warehousing
Space exploration
Disaster recovery operations
Google Sites
Report abuse