All Literacy Courses — AESOP AI Academy

자격	기초	Core Course Credits （3년마다 갱신）	Active Course Credits （매년 갱신）
Level I 인증 AI 전문가	입문+	6	—
Level II 인증 고급 AI 전문가	기초+	12	6
Level III 인증 전문 AI 전문가	심화	24	12

✅ QA Courses 🎓 Accreditation

📡 AI Models

🔭 AI Progress

🎨 Art

💼 Business

⚙️ Development

📚 Core Courses

시작하기

과정 선택

위 드롭다운으로 카테고리별 과정을 탐색하세요

모듈 탐색

각 과정에는 수업, 퀴즈, 실습이 포함된 스토리 기반 모듈이 있습니다

학습 시작

진행 중인 모듈을 클릭하여 시작하세요 — 진도가 자동으로 기록됩니다

Course modules

Defining AI Governance

What it is, what it isn't, and why the question matters more than the answer

The World's First AI Law

How Europe decided to regulate AI — and what that means for everyone

The US Approach

Voluntary frameworks, principles-based guidance, and the theory behind American AI policy

China's AI Governance Framework

State-centric regulation, targeted rules, and a different theory of what AI governance means

Corporate AI Governance Structures

How organizations govern AI internally — and the gap between policy and practice

How AI Auditing Works

Technical audits, process audits, and outcome audits — and why each alone is insufficient

What Corporate AI Policies Contain

The anatomy of a corporate AI policy — and why structure matters for reading critically

Policy Drafting Fundamentals

How to write AI governance policy that creates real obligations — not just intentions

Enter Course →

Course modules

AI Is Already Everywhere

Mapping the systems shaping your day before you notice them

What AI Actually Does to Jobs

Task automation, labor market power, and what the evidence says

Disinformation at Scale

How AI changes the production, spread, and detection of false information

Facial Recognition and Biometric Surveillance

How AI identifies people at scale — and what that changes about being in public

The Digital Divide and AI Access

Who gets AI's benefits — and who gets left out

The Digital Divide and AI Access

Who gets the benefits of AI — and who gets left behind

The Energy Cost of AI

What training and running AI actually costs the planet — and why the numbers are contested

Forecasting AI — What We Know and Don't Know

How to think about AI's future without falling for hype or dismissal

Enter Course →

Course modules

How Generative AI Works

The models behind text, image, audio, and video generation

AI & Writing

Using AI for drafting, editing, and storytelling — and its limits

AI & Visual Art

Tools, techniques, and the authorship question

AI & Music

From composition to production — how AI is changing music

The Consent & Copyright Problem

Training data, style theft, and the legal gray zone

Protecting Your Creative Voice

How to use AI without losing what makes your work yours

AI as a Creative Partner

Workflow design for human-AI creative collaboration

The Future of Creative Work

Economic and cultural implications of generative AI

In Development

Course modules

When Machines Decide

What it means for an AI system to make a decision — and who is responsible

Surveillance at Scale

How AI-powered monitoring transforms the relationship between institutions and individuals

The Consent You Never Gave

Data collection, inference, and the gap between what you share and what companies know

The Professional's Dilemma

What engineers, designers, and researchers owe — to users, to the public, and to themselves

AI and Democracy

How AI systems interact with elections, political discourse, and the conditions for democratic self-governance

AI in Healthcare

Clinical decision support, diagnostic AI, and the ethics of machines in medicine

Autonomous Systems and Moral Agency

What changes ethically when AI acts in the world — not just recommends, but decides and executes

AI and Human Identity

What AI systems that simulate persons, relationships, and emotional connection change about what it means to be human

Enter Course →

Course modules

Thinking Like a Builder

Mental models for designing AI-powered systems

Prompt Engineering

Writing effective prompts for any task or model

Working with APIs

How to connect AI tools to your own projects

AI Tool Integration

Combining multiple AI tools into a working workflow

Designing AI Products

UX, safety, and product design for AI-powered experiences

Evaluating AI Output

How to test, validate, and improve what your AI produces

Responsible Building

Ethics, safety, and accountability for AI builders

Your First AI Project

A guided capstone to launch something real

Enter Course →

Course modules

The AI Career Landscape

Every major AI career path mapped out

Technical Roles

ML engineering, data science, AI research — what each requires

Policy & Governance Roles

AI policy analyst, ethicist, compliance — growing fast

Creative & Design Roles

AI-adjacent creative careers and how to build them

Reading AI Research

How to find, read, and evaluate AI papers

Your Skill Roadmap

A personalized plan to reach your AI career goal

In Development

Course modules

AI Diagnostics

How AI reads scans, detects disease, and assists clinicians

Drug Discovery & AI

Accelerating research from years to months

AI in Mental Health

Companion tools, therapy bots, and the risks of over-reliance

Health Data & Privacy

HIPAA, patient consent, and the value of health data

AI & Medical Equity

Who benefits from medical AI — and who is systematically excluded

Regulation & Safety in Medical AI

FDA pathways, clinical validation, and real-world deployment

Enter Course →

Course modules

AI in the Classroom

How AI tools are already changing teaching and learning

Personalized Learning

The promise and the risks of AI-driven adaptive learning

Academic Integrity in the AI Era

Plagiarism, assessment design, and what fairness means now

AI & the Achievement Gap

Does AI narrow or widen educational inequity?

Designing AI-Resilient Curriculum

Building courses and assessments that work with AI

The Future of Credentials

Degrees, certificates, and competency in an AI world

Enter Course →

Course modules

AI for Climate Science

Modeling, prediction, and monitoring at planetary scale

AI in Energy Systems

Smart grids, demand forecasting, and renewable optimization

AI for Environmental Monitoring

Satellites, sensors, and real-time ecosystem tracking

The Carbon Cost of AI

Training emissions, data center energy, and the trade-off

AI & Climate Justice

Who bears the environmental cost of AI infrastructure

Building for Sustainability

How AI developers can reduce their environmental footprint

In Development

Course modules

Algorithmic Trading

How AI moves markets and what can go wrong

AI in Lending & Credit

Automated decisions, bias, and the fair lending question

Fraud Detection & Cybersecurity

AI on the defense — and on the attack

AI & Financial Risk

Systemic risk, model risk, and what regulators are watching

Central Banks & AI

How monetary policy intersects with machine learning

The Future of Money

CBDC, DeFi, and AI's role in reshaping finance

In Development

Course modules

How AI Shapes Attention

Recommendation systems, engagement loops, and cognitive load

Trust & Anthropomorphism

Why we trust AI — and when that trust is misplaced

AI & Addiction

Persuasive design, dopamine loops, and digital wellbeing

AI & Identity

How AI tools are changing how we see ourselves

Emotional AI

Affective computing, chatbot companions, and emotional manipulation

Designing for Wellbeing

Building AI products that support — not exploit — human psychology

In Development

Course modules

What Is AI Safety?

The field, its history, and why it matters now

Alignment Fundamentals

RLHF, constitutional AI, and the core alignment techniques

Failure Modes & Edge Cases

How aligned AI can still behave in dangerous ways

Interpretability

Opening the black box — what mechanistic interpretability reveals

Existential Risk

The long-term safety arguments and how to evaluate them

AI Safety Research Landscape

Key organizations, open problems, and how to contribute

In Development

Course modules

Python & AI Tooling

The essential stack for applied AI development

Working with Foundation Models

APIs, prompting at scale, and model selection

Fine-Tuning

When to fine-tune, how to do it, and what it costs

RAG & Knowledge Systems

Retrieval-augmented generation for real-world applications

Evaluation & Testing

How to measure whether your AI system actually works

Deployment & Monitoring

Shipping AI to production and keeping it healthy

AI System Design

Architecture patterns for scalable, safe AI applications

Capstone Project

Build and deploy a complete AI-powered application

In Development

Course modules

AI Strategy for Leaders

How to think about AI as a strategic asset or risk

Building an AI Team

Hiring, structuring, and managing AI capability in your org

Procurement & Vendor Evaluation

How to evaluate and buy AI products responsibly

Change Management for AI

Getting your organization ready for AI-driven change

AI Risk for Executives

Reputational, legal, and operational risks of AI deployment

Board Governance & AI

Fiduciary duty, oversight, and board-level AI literacy

Enter Course →

Course modules

AI in Journalism

From automated reporting to AI-assisted investigation

Deepfakes & Synthetic Media

Detection, impact, and the future of visual trust

AI & Misinformation

How AI creates, spreads, and can combat false information

Content Moderation at Scale

How platforms use AI to police speech — and where it fails

AI & Advertising

Programmatic ad systems, targeting, and manipulation

The Future of Media

Business models, audience trust, and AI's role in it all

In Development

Course modules

AI in Defense Systems

Autonomous weapons, decision support, and international law

Intelligence & Surveillance

How AI is transforming intelligence collection and analysis

Cyber Operations & AI

AI-enabled attacks, defense, and the evolving threat landscape

AI Geopolitics

The US-China AI race and what it means for the international order

Arms Control for AI

Can AI weapons be regulated? What treaties already exist?

Dual-Use Research

When civilian AI becomes a military capability

In Development

Course modules

What Is Intelligence?

Definitions, theories, and what AI reveals about the question

The Consciousness Problem

Why we can't agree on what consciousness is — and why it matters for AI

The Chinese Room Revisited

Searle, Turing, and what modern AI tells us about each

Moral Status & AI Rights

When — if ever — might AI systems deserve moral consideration?

AI & Human Identity

What it means to be human in a world of intelligent machines

The Long View

Where AI fits in the arc of intelligence on Earth

In Development

Course modules

Reading the Trajectory

How to assess AI progress without hype or dismissal

Transformative AI Scenarios

What different futures look like and how likely each is

Economic Transformation

AI and the long-run future of work, wealth, and human purpose

Governing Transformative AI

What institutions, laws, and norms we need — and don't have

The Human Flourishing Question

What does a good future with AI actually look like?

Your Role in It

How to contribute — as a learner, builder, citizen, or leader

In Development

Course modules

The Model Landscape

Who built what, and why the differences between models actually matter

How GPT-4o Works

OpenAI's architecture, training approach, and what makes it distinctive

Understanding Claude

Anthropic's constitutional AI, safety focus, and where Claude excels

Gemini and Google's Approach

Multimodal-first design and how Google's AI strategy differs

Benchmark Reality Check

What benchmarks measure, what they miss, and how to read them critically

Cost, Speed, and Context

Practical tradeoffs — tokens, latency, pricing, and context window limits

Choosing the Right Model

A framework for model selection across use cases and constraints

The Frontier Is Moving

How to track model releases and evaluate new models as they launch

Enter Course →

Course modules

What Multimodal Means

Why combining modalities is harder than it sounds — and why it matters

Vision-Language Models

How models like GPT-4V and Gemini understand and describe images

Image Generation Deep Dive

Diffusion models, latent space, and how images are generated from text

Audio and Speech AI

Transcription, voice synthesis, and real-time audio understanding

Video Understanding

How AI processes and generates video — current capabilities and limits

Cross-Modal Reasoning

Tasks that require connecting information across different input types

Building Multimodal Applications

Practical patterns for integrating multimodal models into products

What's Coming in Multimodal AI

The next capabilities — and the hard problems still unsolved

In Development

Course modules

The RL Framework

Agents, environments, rewards, and policies — the core vocabulary

How Reward Signals Work

Designing rewards that actually produce the behavior you want

From Games to Real Problems

How RL moved from Atari and chess to robotics and language models

RLHF Explained

How human feedback is used to align language model outputs

Policy Gradient Methods

The math and intuition behind modern RL training approaches

Exploration vs. Exploitation

The fundamental tradeoff at the heart of every RL system

When RL Goes Wrong

Reward hacking, misalignment, and famous RL failures

Decision Models in Production

Real-world applications — recommendation systems, logistics, and more

In Development

Course modules

Why Run Models Locally

Privacy, cost, latency, and the cases where local beats cloud

The Open-Source Model Ecosystem

Llama, Mistral, Phi, and the landscape of locally-runnable models

Hardware Requirements

What GPU, RAM, and storage you actually need for different model sizes

Quantization Explained

How 4-bit and 8-bit quantization shrinks models without killing quality

Setting Up Ollama

The fastest path from zero to running a local model

LM Studio and llama.cpp

Alternative runtimes and when each makes sense

Prompt Formatting for Local Models

Why chat templates matter and how to get consistent outputs

Building Applications on Local Models

API compatibility, speed tuning, and practical production patterns

In Development

Course modules

The Transformer Architecture

Attention mechanisms, tokens, and what's actually happening inside an LLM

Pre-Training at Scale

How models learn from the internet — and what that means for their behavior

Tokens, Context, and Memory

Why LLMs don't 'remember' and what context windows actually do

Temperature and Sampling

How randomness is controlled to produce different output behaviors

Emergent Capabilities

Why large models can do things small models can't — and the debate around why

Fine-Tuning vs. Prompting

When you need to train and when a good prompt is enough

Hallucination and Confabulation

Why models make things up and what can be done about it

The Limits of Scale

What bigger models can't fix — and where the field is going

In Development

Course modules

How Diffusion Models Work

Noise, denoising, and the math behind image generation

Text-to-Image Pipelines

How a prompt becomes an image — step by step

The Role of CLIP and Vision Encoders

How models understand what text means visually

Stable Diffusion Architecture

UNet, VAE, and the components that make open-source generation possible

LoRA and Model Customization

How to train lightweight adapters for consistent styles and subjects

ControlNet and Guided Generation

Depth maps, poses, edges — controlling generation with structure

Prompt Engineering for Images

What actually works — syntax, weights, negative prompts, and style tokens

Evaluating and Selecting Generated Images

How to assess quality, consistency, and fitness for use

In Development

Course modules

Why Benchmarks Exist

The history and purpose of AI evaluation — from ImageNet to MMLU

Reading Benchmark Scores

What leaderboards show, what they hide, and how to read them honestly

Common Benchmarks Explained

MMLU, HumanEval, HellaSwag — what each tests and why it matters

Benchmark Contamination

How training on test data distorts scores — and how to detect it

Task-Specific Evaluation

Designing evals for your actual use case, not the general case

Human Evaluation Methods

When to use humans and how to do it without introducing bias

Building an Eval Pipeline

Automating evaluation across model versions and prompt changes

The Limits of Evaluation

What no benchmark measures — and what that means for trust

In Development

Course modules

What Alignment Means

Defining the problem — and why 'just make it do what we want' isn't simple

Instrumental Convergence

Why many different AI goals lead to similar dangerous sub-goals

The Specification Problem

Why it's nearly impossible to fully specify what we actually want

RLHF and Its Limits

How current alignment techniques work and where they break down

Deceptive Alignment

The concern that systems might appear aligned while pursuing other goals

Scalable Oversight

How to supervise AI systems smarter than us — proposed approaches

The Debate in the Field

Safety researchers, doomers, and optimists — mapping the disagreements

What You Can Do

How individuals, companies, and governments are engaging with alignment

In Development

Course modules

The State of Voice AI

From Siri to GPT-4o voice — how the technology evolved

Automatic Speech Recognition

How transcription works and why accuracy varies by speaker and context

Text-to-Speech Synthesis

Neural TTS, voice cloning, and what makes voices sound natural

Real-Time Conversation Models

How low-latency voice models handle interruption and turn-taking

Voice Interfaces in Products

Design principles for conversational UX — what works and what doesn't

Speaker Identification and Diarization

Who said what — identifying speakers in multi-party audio

Emotional and Prosodic AI

How models are learning to interpret and express tone and emotion

Where Voice AI Is Headed

Multimodal conversation, always-on assistants, and the ambient computing future

In Development

Course modules

Why Hardware Determines AI Capability

The link between compute, model size, and what AI can do

NVIDIA's Dominance

How CUDA lock-in and GPU architecture made NVIDIA the AI hardware king

Google's TPU Strategy

Custom silicon, distributed training, and why Google built its own chips

The New Entrants

AMD, Intel, Cerebras, Groq — challengers and what they're trying to do differently

The Memory Bandwidth Bottleneck

Why moving data is often harder than computing it

Inference vs. Training Hardware

Different workloads demand different silicon — what that means in practice

The Export Control Dimension

How US chip restrictions are shaping the global AI race

What Comes After GPUs

Neuromorphic chips, optical computing, and long-run hardware possibilities

In Development

Course modules

What Synthetic Data Is

Generated vs. real data — definitions, use cases, and tradeoffs

Why Real Data Is Running Out

Token scarcity, copyright, and the limits of internet-scale training

How Models Generate Training Data

Self-play, distillation, and prompting models to produce their own examples

Quality Control for Synthetic Data

Filtering, verification, and avoiding model collapse

Constitutional AI and Self-Critique

How Anthropic uses AI feedback to improve AI behavior

Model Distillation

Training smaller models on larger model outputs — efficiency gains and limits

The Recursive Improvement Question

Can AI improve itself indefinitely — and what are the risks?

Where Synthetic Data Is Going

Current research frontiers and what the next generation of training looks like

In Development

Course modules

What a Context Window Is

Tokens, attention, and why longer context is technically hard

The History of Context Expansion

From GPT-3's 4K to Gemini's 1M — how context grew and why

Attention Complexity

Why quadratic attention makes long contexts expensive to compute

Efficient Attention Methods

Sliding window, sparse attention, and the architectures solving long context

What You Can Do With Long Context

Full codebases, books, legal documents — real use cases unlocked

Lost in the Middle

Why models don't use long context equally — attention distribution problems

RAG vs. Long Context

When to retrieve and when to just put it all in the prompt

Where Context Length Is Going

Theoretical limits, practical limits, and what the next generation enables

In Development

Course modules

What Reasoning Means in AI

Defining reasoning — and why it's different from retrieval or generation

Chain-of-Thought Prompting

How asking models to show their work dramatically improves accuracy

The o1 Architecture

What OpenAI changed to produce reasoning models — and what we know

Test-Time Compute

Trading inference cost for accuracy — the tradeoff that reasoning models exploit

When Reasoning Models Beat Standard Models

Task types where extended thinking actually helps

Math and Code as Reasoning Benchmarks

Why formal domains reveal reasoning capability most clearly

Reasoning Failures

Where reasoning models still get it wrong — and why

The Road to Reliable Reasoning

Open research problems and where the field is heading

In Development

Course modules

AI as a Scientific Tool

How AI fits into the scientific method — and where it fits best

AlphaFold and Protein Structure

The breakthrough that changed biology and what came after it

Drug Discovery and Molecular Design

How AI is compressing the timeline from target to candidate

Climate and Earth System Modeling

AI's role in simulating complex systems at planetary scale

Materials Science and Discovery

Finding new materials by searching chemical space with AI

AI in Astronomy and Physics

Pattern recognition at cosmic scale — what AI finds that humans miss

Reproducibility and Scientific Integrity

AI-specific challenges for peer review and scientific validation

The Future of AI-Augmented Research

What science looks like when every researcher has an AI co-pilot

In Development

Course modules

What Makes Something an Agent

Perception, reasoning, action, and memory — the agent loop

Browser and Computer-Use Agents

Agents that operate software — current capabilities and limits

Customer Service Agents

The most deployed category — what works and what still breaks

Coding Agents

Devin, GitHub Copilot Workspace, and the state of autonomous coding

Research Agents

Deep research tools — how they work and where they fall short

Multi-Agent Systems

Orchestration, delegation, and what happens when agents talk to agents

Failure Modes and Safety

How agents go wrong and what guardrails are being built

The Frontier of Agent Capability

What the next 12 months of agent development looks like

In Development

Course modules

The History of Multimodal AI

From image classification to GPT-4V — the major milestones

CLIP and Contrastive Learning

The idea that connected text and image understanding

Dall·E, Stable Diffusion, Midjourney

Three approaches to image generation and what made each matter

Whisper and Audio Understanding

OpenAI's speech model and the impact of open-source audio AI

GPT-4V and Gemini Vision

The moment frontier models could see — capabilities and limits

Sora and Video Generation

What video generation changes and what it still can't do

Real-Time Multimodal Interaction

Live video understanding, voice, and the ambient AI interface

Where Multimodal Goes Next

Open problems — 3D understanding, tactile sensing, embodied AI

In Development

Course modules

How to Read AI Progress

Signal vs. noise — what announcements, papers, and demos actually mean

The Research Pipeline

From paper to product — how long things actually take to arrive

Capabilities on the Near Horizon

What's in labs right now that hasn't shipped yet

The Agentic Transition

How AI is moving from tools to actors — and what that changes

Hardware and Infrastructure Bets

The infrastructure investments that tell you where the field is going

Geopolitical AI Dynamics

US, China, EU — how policy and competition are shaping the frontier

How to Stay Current

Resources, communities, and practices for tracking AI without drowning in it

Your Role in What Comes Next

How to engage with AI progress as a builder, critic, or citizen

In Development

Course modules

AI's Role in Game History

From rule-based NPCs to neural networks — how game AI evolved

Procedural Content Generation

Algorithms and AI that create levels, maps, and worlds

NPC Behavior and Decision-Making

Behavior trees, finite state machines, and ML-driven characters

Dialogue Systems

Branching narrative, dynamic conversation, and LLM-powered NPCs

AI Game Design Tools Today

What's available now — and what designers are actually using

Player Modeling

Using AI to understand player behavior and personalize experience

Ethical Questions in AI Game Design

Addiction mechanics, manipulation, and the designer's responsibility

The Changing Role of the Game Designer

What AI changes about the craft — and what it can't replace

Enter Course →

Course modules

Generative Level Design

AI systems that build playable spaces from constraints

Dynamic Difficulty Adjustment

Adapting challenge in real time to player skill and frustration

Emergent Narrative

Systems that generate story from player action — not author intention

AI Dungeon Masters

LLM-powered game masters — what they can do and what they can't

Asset Generation for Games

Textures, sprites, audio — AI tools accelerating game production

AI Playtesting

Using AI agents to find bugs, balance issues, and progression blockers

Player Sentiment Analysis

Mining reviews and telemetry with AI to guide design decisions

Case Studies in AI Game Design

Real games using AI meaningfully — what worked and what didn't

In Development

Course modules

Fully Generative Game Worlds

What it means to generate everything — story, space, characters, and rules

AI Companions and Relationship Systems

Long-term, memory-driven AI characters and what they demand

Real-Time Neural Rendering

How AI is changing how game worlds look and perform

The Economic Impact on Game Studios

Job changes, cost structures, and what AI means for indie vs. AAA

Multiplayer and Competitive AI

Bots that play fairly, anti-cheat systems, and matchmaking AI

AI-Generated Games at Runtime

The vision of games that write themselves as you play

Player Rights and AI Content

Who owns AI-generated in-game content — emerging legal questions

Designing for an AI-Native Future

Principles for game designers working with and against AI systems

In Development

Course modules

The AI Video Landscape

Text-to-video, image-to-video, and the tools reshaping production

How Video Generation Models Work

Diffusion, temporal consistency, and the hard problem of motion

Sora, Runway, and Kling

Comparing the leading generative video tools — capabilities and limits

AI-Assisted Scriptwriting

Using LLMs for story development, dialogue, and script drafting

AI in Editing and Post-Production

Automated cuts, upscaling, color grading, and VFX with AI

Voice and Audio for AI Video

Synthesis, dubbing, sound design — the audio side of AI production

Legal and Ethical Dimensions

Building an AI Video Workflow

Practical integration — what to use, when to use it, what to avoid

In Development

Course modules

How Prompts Become Images

The pipeline from text to image — what the model actually processes

Core Composition Vocabulary

Framing, lighting, subject, style — the building blocks of a good prompt

Style and Reference Techniques

Artist references, era tags, and how to guide aesthetic direction

Negative Prompting

What to exclude, why it matters, and common negative prompt patterns

Prompt Weighting and Syntax

Using parentheses, brackets, and numeric weights to control emphasis

Iterative Refinement

Treating prompt craft as a design process — systematic iteration strategies

LoRA and Embedding Prompts

Activating trained concepts, styles, and characters in your prompts

Building a Prompt Library

Organizing, versioning, and sharing prompts that reliably work

In Development

Course modules

AI in the Design Process

Where AI fits in architectural workflow — ideation, documentation, analysis

Generative Design Tools

Parametric AI tools architects are actually using — Midjourney, Stable Diffusion, Rhino AI

Floor Plan and Space Generation

AI systems that generate functional layouts from programmatic constraints

Structural Optimization

How AI finds efficient structural forms that humans wouldn't design

Environmental Simulation

AI-driven energy modeling, daylight analysis, and climate-responsive design

Urban Planning and City-Scale AI

Traffic modeling, zoning optimization, and AI in urban policy

Computational Heritage and Preservation

Using AI to document, analyze, and restore historic buildings

The Future of the Architect

How the profession is changing — and what AI can't replace in design

In Development

Course modules

How AI Composes Music

The models and methods behind AI-generated melody, harmony, and rhythm

Generative Music Tools

Suno, Udio, MusicGen — comparing the leading AI music platforms

Prompt Engineering for Music

Describing what you want — genre, mood, instrumentation, and structure

AI in Music Production

MIDI generation, stem separation, mastering assistance, and sample creation

Training on Human Music

Collaboration Between Human and AI

Workflows where AI augments rather than replaces the composer

Generative Music for Games and Film

Adaptive soundtracks, procedural audio, and real-time composition

Authorship and Ownership

Who owns AI music — current law, industry practice, and open questions

In Development

Course modules

AI's Place in the Design Workflow

Where AI helps, where it hurts, and what designers actually use it for

Generative Image Tools for Designers

Firefly, Midjourney, DALL·E — comparing outputs for design use cases

Logo and Brand Identity Generation

AI in brand exploration — speed, iteration, and the limits of generation

Layout and Composition Assistance

AI tools for page layout, spacing, and visual hierarchy

Typography and AI

Font recommendation, variable fonts, and AI-generated letterforms

Brand Consistency with AI

Training models on brand assets and maintaining visual identity

Presentation and Infographic Design

AI-powered slide design, data visualization, and document creation

The Changing Design Profession

What AI changes about the designer's role — and what it doesn't

In Development

Course modules

What Voice Means in Writing

Defining the thing AI threatens to flatten — and how to protect it

AI as Drafting Partner

Using LLMs for first drafts, brainstorming, and overcoming blocks

Preserving Voice When Using AI

Prompting techniques that pull AI toward your style, not away from it

AI-Assisted Editing

Grammar, clarity, flow — where AI editing helps and where it damages writing

Research and Fact-Finding with AI

Using AI for research without inheriting its hallucinations

Genre-Specific AI Writing

Fiction, nonfiction, poetry, journalism — how AI tools differ by form

Disclosure and Authorship Ethics

When and how to disclose AI use — industry norms and personal ethics

Developing Your AI-Augmented Practice

Building a writing workflow that uses AI without depending on it

In Development

Course modules

AI in the Camera

Computational photography — how your phone already uses AI before you edit

AI Photo Editing Tools

Lightroom AI, Luminar, Photoshop generative fill — what they actually do

Image Upscaling and Restoration

Super-resolution, noise reduction, and restoring damaged photographs

Generative Fill and Inpainting

Adding, removing, and replacing content in photographs with AI

AI-Powered Photo Selection

Culling thousands of images — AI tools for editing down a shoot

Style Transfer for Photography

Applying aesthetic styles to photographs — techniques and limits

The Ethics of AI Photography

Disclosure, manipulation, deepfakes, and photojournalism standards

The Future of the Photographer

What AI changes about the craft — and where human vision still leads

In Development

Course modules

AI on Stage

Current uses of AI in theater and performance — and the debates they spark

Generative Text for Theatre

LLMs in playwriting, dialogue generation, and dramaturgical research

AI in Choreography

Motion capture, movement analysis, and AI-assisted dance creation

Synthetic Actors and Digital Doubles

De-aging, resurrection, and virtual performers — where the industry stands

Immersive and Interactive Performance

AI-driven audience interaction and responsive performance environments

AI for Music and Sound in Live Performance

Real-time composition, adaptive scoring, and live AI instruments

Accessibility and AI in Performance

Sign language interpretation, audio description, and real-time captioning

The Human Question in Performing Arts

What live performance means when performers and writers can be simulated

In Development

Course modules

Narrative Structure and AI

Teaching AI what makes a story work — structure, tension, and payoff

Character Development with LLMs

Building characters AI can portray consistently across a long narrative

World-Building Assistance

Using AI to develop consistent history, geography, and culture

Interactive and Branching Narrative

AI-powered choice-based storytelling — technical and creative patterns

AI as Story Editor

Using AI critique to find structure problems, pacing issues, and inconsistencies

Collaborative Fiction and Co-Writing

Workflows for human-AI collaborative story development

AI-Native Story Forms

New forms of narrative only possible with AI — generative fiction, living stories

Audience and Reader Considerations

How AI storytelling changes the reader relationship and disclosure ethics

In Development

Course modules

The AI-Augmented Solo Founder

What's possible now — and the mindset shift required to use it

Operations with AI

Project management, scheduling, email, and admin — eliminating the busywork

Market Research and Competitor Analysis

AI-powered research workflows that used to require a team

Content and Marketing at Scale

Using AI to produce consistent, on-brand content without a marketing department

AI-Assisted Product Development

Prototyping, spec writing, and feedback analysis with AI

Sales and Customer Communication

AI tools for outreach, follow-up, and customer support

Financial Management with AI

Forecasting, reporting, and scenario planning with AI assistance

Building Your AI Toolkit

Selecting, integrating, and iterating your personal AI stack

Enter Course →

Course modules

The AI Marketing Landscape

What's changed, what's hype, and what's actually driving results

AI-Generated Content at Scale

Blog posts, ads, social copy — production systems that maintain quality

Audience Intelligence and Segmentation

Using AI to understand who your customers are and what they want

Personalization at Scale

Dynamic content, email personalization, and AI-driven customer journeys

AI in Paid Advertising

Smart bidding, creative testing, and AI campaign optimization

SEO and Content Strategy with AI

Keyword research, content briefs, and AI-assisted SEO workflows

Analytics and Attribution

AI tools for measurement — what's working and what the numbers actually mean

Ethics and Transparency in AI Marketing

Disclosure, manipulation concerns, and building trust with AI-assisted marketing

In Development

Course modules

The AI Risk Landscape

Mapping the risks — what to worry about and in what order

Operational Risk

Model failures, hallucinations, and what happens when AI gets it wrong in production

Reputational Risk

Bias, controversy, and the AI incidents that have damaged brands

Legal and Regulatory Exposure

Liability, IP ownership, privacy, and the emerging legal framework for AI

Vendor and Supply Chain Risk

Depending on third-party AI — lock-in, outages, and policy changes

Data Risk and Privacy

What AI systems do with your data — and your customers' data

Building an AI Risk Framework

Practical governance structures that work at different organizational scales

Enter Course →

Course modules

The AI Investment Landscape

Where money is flowing, who's funding what, and why the AI bubble is different

What Investors Look for in AI Companies

Defensibility, data moats, team, and the product-market fit question

Crafting the AI Pitch Narrative

How to explain what your AI does without losing the room

Technical Due Diligence Preparation

The questions sophisticated investors ask — and how to answer them

Valuation in AI Businesses

Revenue multiples, ARR, and how investors value AI-native companies

Competitive Differentiation

How to answer 'why won't OpenAI just do this?' convincingly

Term Sheet Dynamics for AI Deals

IP ownership, data rights, and AI-specific terms to negotiate

From Pitch to Close

Managing the fundraising process — timelines, updates, and decision-making

In Development

Course modules

What the Evidence Shows

Labor market data, productivity research, and what's actually happening to jobs

Task Automation vs. Job Automation

Why the right unit of analysis is tasks, not jobs — and what that means

High-Risk Occupations

Which roles face the most disruption — and the timeline for that disruption

Human-AI Collaboration Models

Centaur models, AI augmentation, and the jobs that grow with AI

Reskilling and Workforce Transition

What effective reskilling looks like — and what doesn't work

Organizational Change Management

Leading teams through AI adoption — communication, trust, and resistance

AI and Labor Relations

Union responses, worker rights, and the politics of AI in the workplace

Building an AI-Ready Organization

Culture, structure, and practices for organizations that adapt well

Enter Course →

Course modules

AI in the Product Lifecycle

Where AI fits at each stage — and where it currently breaks down

AI-Assisted User Research

Interview analysis, survey synthesis, and pattern detection in qualitative data

Rapid Prototyping with AI

Going from idea to testable prototype faster with generative tools

Spec and PRD Writing with AI

Using LLMs to accelerate documentation without losing precision

AI in Development Workflow

Copilot-assisted coding, code review, and the developer experience shift

AI-Powered Testing

Automated test generation, edge case discovery, and regression coverage

Product Analytics with AI

Behavioral analysis, anomaly detection, and AI-assisted decision-making

Building AI Features Users Trust

Explainability, error states, and designing for AI uncertainty

In Development

Course modules

What AI-First Actually Means

The difference between using AI tools and building AI into your business model

Identifying AI-Native Opportunities

Finding problems where AI creates structural advantage, not just efficiency

Data Strategy and Moats

Building proprietary data assets that compound your AI advantage

The AI-First Tech Stack

Infrastructure decisions — models, infrastructure, and build vs. buy tradeoffs

Team and Culture for AI-First Orgs

Hiring, roles, and the cultural norms that make AI-first work

Speed as a Strategic Asset

How AI-first businesses move faster — and how incumbents respond

Ethical and Governance Foundations

Building responsibility into the architecture before you need it

Scaling an AI-First Business

What breaks at scale — and how to build systems that hold

In Development

Course modules

AI in Financial Analysis

Automating reports, identifying patterns, and AI-assisted financial modeling

Forecasting and Scenario Planning

AI tools for revenue forecasting, demand planning, and risk modeling

Accounts Payable and Receivable Automation

Document processing, matching, and exception handling with AI

AI in Supply Chain and Logistics

Demand sensing, inventory optimization, and routing intelligence

Expense Management and Fraud Detection

AI systems for identifying anomalies and controlling spend

AI in HR Operations

Recruiting, workforce planning, and employee experience with AI

Operational Dashboards and AI Reporting

Building executive-level AI-powered reporting systems

Risk and Compliance in AI Finance Systems

Audit trails, explainability, and regulatory considerations

In Development

Course modules

The State of AI Customer Service

What's deployed, what customers think, and where the bar is today

Chatbot Architecture and Design

Intent recognition, dialog management, and building bots that don't frustrate

LLM-Powered Support Agents

Moving beyond scripted bots to generative AI in customer-facing roles

Agent Assist Tools

Real-time AI assistance for human agents — how it improves speed and quality

Intelligent Ticket Routing

Classification, prioritization, and AI-driven support workflow

Voice AI in Customer Service

IVR evolution, voice bots, and real-time transcription for support

Measuring AI Customer Service Performance

CSAT, resolution rate, and the metrics that matter for AI-powered support

The Human-AI Support Team Model

Designing the blend — when to hand off, escalate, and intervene

In Development

Course modules

The AI Vendor Landscape

Categories of AI vendors, what they sell, and how to orient in the market

Defining Requirements Before You Shop

How to specify what you need before vendors tell you what they have

Technical Evaluation Criteria

Performance, reliability, latency, scalability — what to measure and how

Security and Data Handling Assessment

What to audit in vendor data practices and where the risks are

Contract and IP Negotiation

Data rights, model ownership, SLAs, and AI-specific contract terms

Proof of Concept Design

Running a rigorous PoC that actually predicts production performance

Vendor Risk Monitoring

Ongoing governance — how to track vendor health and catch problems early

Build vs. Buy Decision Framework

When to procure, when to build, and how to make the case either way

In Development

Course modules

What Is an AI Agent

Agents vs. tools vs. assistants — precise definitions and why they matter

The Agent Loop

Perception, reasoning, action, observation — the core cycle of agent operation

Survey of Agent Use Cases

Research, coding, customer service, data analysis — mapping the landscape

When Agency Adds Value

The decision framework for whether an agent is actually the right solution

Agent Architectures Overview

ReAct, Plan-and-Execute, and other patterns — a comparative map

Multi-Agent Systems Introduction

Orchestrators, sub-agents, and the cases that require multiple agents

Evaluating Agent Performance

Task completion, cost, reliability — how to measure agent quality

Risks and Failure Modes

What goes wrong with agents — and the categories of failure to design against

Enter Course →

Course modules

Memory Types for Agents

Short-term, episodic, semantic, and procedural memory — what each does

Implementing Working Memory

Context management, summarization, and keeping agents on track

Long-Term Memory with Vector Databases

Embeddings, retrieval, and giving agents persistent knowledge

Knowledge Graphs for Agents

Structured knowledge and when graphs outperform vector search

Agent Planning Strategies

Task decomposition, subgoal generation, and backtracking

Reasoning Under Uncertainty

How agents handle ambiguity, missing information, and conflicting signals

Skill Libraries and Agent Capabilities

Building reusable skill modules that agents can compose

Benchmarking Agent Cognition

How to measure reasoning and planning quality — practical eval methods

Enter Course →

Course modules

Tool Use Fundamentals

Function calling, tool schemas, and how agents decide which tool to use

Web Search and Browsing Tools

Search APIs, browser automation, and web-capable agent design

Code Execution Environments

Sandboxed code runners — security, capabilities, and integration patterns

File System and Document Tools

Reading, writing, and manipulating files safely within an agent

API Integration Patterns

Connecting agents to third-party services — auth, rate limits, and error handling

MCP Integration

Model Context Protocol — architecture, setup, and building MCP-compatible agents

Tool Selection and Orchestration

How to design tool selection logic that doesn't get expensive or stuck

Security and Safety for Tool-Using Agents

Injection attacks, privilege limits, and safe tool design

Enter Course →

Course modules

OpenClaw Architecture Overview

Design decisions, component structure, and the philosophy behind the framework

Core Agent Loop Implementation

Building the perception-reasoning-action cycle from scratch

Tool Registry and Plugin System

Dynamic tool loading, schema validation, and plugin architecture

Memory System Implementation

In-memory and persistent memory — the OpenClaw approach

Orchestration and Sub-Agent Management

How OpenClaw handles multi-agent coordination

Observability and Logging

Tracing agent runs, debugging failures, and monitoring in production

Testing Your Agent System

Unit tests, integration tests, and end-to-end agent evaluation

Deploying OpenClaw to Production

Containerization, scaling, and operating an agent system in the real world

Enter Course →

Course modules

Profiling Agent Performance

Finding bottlenecks — where time and money actually go in an agent run

Reducing Latency

Parallelization, caching, and prompt engineering for faster agent response

Token Optimization Strategies

Shrinking context, summarizing history, and reducing unnecessary LLM calls

Cost Modeling for Agent Systems

Estimating, tracking, and controlling the cost of agent operations

Caching for Agents

What to cache, where to cache it, and invalidation strategies

Graceful Degradation

Designing agents that fail safely — fallbacks, retries, and circuit breakers

Reliability Engineering for Agents

SLA design, uptime monitoring, and incident response for agent systems

Continuous Improvement Pipelines

Feedback loops, eval-driven iteration, and keeping agents getting better

Enter Course →

Course modules

AI Application Architecture Patterns

Gateway, pipeline, mesh — the major patterns for AI app structure

Model Routing and Selection

Dynamic model selection — routing by task, cost, and quality requirements

Latency Management

Streaming, async patterns, and designing for perceived vs. actual response time

Caching Strategies for AI

Semantic caching, prompt caching, and when caching helps vs. hurts

Context Management at Scale

Handling long contexts, conversation history, and multi-user state

Multi-Model Pipelines

Chaining models — when to split tasks and how to manage handoffs

Infrastructure and Deployment

Containerization, orchestration, and scaling AI workloads

Reliability and Fallback Design

Graceful degradation, fallback models, and building systems that hold

In Development

Course modules

Prompt Engineering as Engineering

Treating prompts as code — version control, testing, and iteration

System Prompt Design

Persona, constraints, output format — the anatomy of a good system prompt

Few-Shot Examples

When to use examples, how many, and how to select the right ones

Chain-of-Thought Techniques

Getting models to reason step by step — and when it helps

Structured Output Prompting

JSON, XML, and format enforcement — reliable structured generation

Prompt Injection Defense

How injection attacks work and how to design prompts that resist them

Evaluating and Testing Prompts

Building test suites for prompts — coverage, regression, and red-teaming

Prompt Management in Production

Storage, versioning, A/B testing, and prompt deployment workflows

In Development

Course modules

AI Observability Fundamentals

What to measure in an AI system and why standard APM isn't enough

Logging for AI Applications

What to log, at what granularity, and how to do it without killing performance

Tracing LLM Calls

End-to-end tracing from user request to model response — tools and patterns

Monitoring Model Quality Over Time

Drift detection, output quality monitoring, and knowing when to retrain

Alerting and Anomaly Detection

Setting up alerts that catch real problems without alert fatigue

Cost Monitoring and Budgeting

Tracking token spend, setting limits, and building cost dashboards

Incident Response for AI Systems

On-call runbooks, postmortems, and recovering from AI outages

Continuous Evaluation in Production

Running evals continuously — not just at deployment

In Development

Course modules

Why RAG Exists

The problem RAG solves — grounding, currency, and context limits

Embeddings Deep Dive

How text becomes vectors — models, dimensions, and similarity search

Chunking Strategies

Fixed-size, semantic, and recursive chunking — what works for different content

Vector Databases

Pinecone, Weaviate, Chroma, pgvector — comparing the options and tradeoffs

Hybrid Search

Combining dense and sparse retrieval for better coverage and precision

Retrieval Quality and Evaluation

Measuring retrieval — recall, precision, and end-to-end quality

RAG Pipeline Design

Query rewriting, re-ranking, and the full retrieval-to-generation pipeline

Production RAG Architecture

Scale, latency, cost, and operating a RAG system in production

In Development

Course modules

The Messages API

Request structure, roles, content types, and the basics of the API

System Prompts and Context

How to structure system prompts and manage conversation context at scale

Tool Use and Function Calling

Defining tools, handling tool calls, and building reliable tool-use workflows

Vision and Document Input

Passing images and documents to Claude — formats, limits, and use cases

Streaming Responses

Server-sent events, incremental processing, and streaming UX patterns

Rate Limits and Error Handling

Understanding limits, implementing retries, and building resilient clients

Cost Optimization

Prompt caching, token efficiency, and controlling spend on Claude

Production Patterns

Batching, async processing, and architectural patterns for Claude at scale

In Development

Course modules

Chat Completions API

Request structure, roles, parameters, and the core API surface

Function Calling

Defining functions, handling calls, and building reliable function-calling workflows

The Assistants API

Threads, runs, and stateful conversations — when to use Assistants vs. completions

File Handling and Retrieval

Uploading files, vector stores, and the Assistants file search tool

Vision with GPT-4o

Image input formats, multi-image prompts, and vision use cases

Streaming and Real-Time Output

SSE streaming, delta processing, and building responsive UX

Rate Limits, Quotas, and Error Handling

Tier limits, backoff strategies, and production resilience

Cost Management and Optimization

Prompt caching, token tracking, and building cost-efficient OpenAI applications

In Development

Course modules

When Fine-Tuning Beats Prompting

The cases where training beats prompting — and the ones where it doesn't

Full Fine-Tuning vs. PEFT

The spectrum from full retraining to lightweight adapters — tradeoffs explained

LoRA Deep Dive

Low-rank adaptation — the math, the intuition, and the practical parameters

QLoRA and Memory Efficiency

4-bit quantization plus LoRA — fine-tuning large models on consumer hardware

Dataset Preparation

Collecting, cleaning, and formatting training data for instruction fine-tuning

Training Configuration

Learning rates, batch size, epochs, and the knobs that matter most

Evaluation and Avoiding Catastrophic Forgetting

Measuring fine-tuned model quality and preserving base capabilities

Deploying Fine-Tuned Models

Serving your model — inference infrastructure and integration patterns

In Development

Course modules

The AI Security Threat Model

Mapping the attack surface — what adversaries want and how they get it

Prompt Injection Attacks

Direct and indirect injection — how they work and defense strategies

Jailbreaking and Policy Bypass

Techniques for bypassing safety training and how to harden against them

Data Exfiltration via LLMs

How attackers extract sensitive data through AI systems — and how to prevent it

Model Inversion and Extraction

Reconstructing training data or model weights — attack and defense

RAG System Security

Poisoning retrieval, manipulating context, and securing the knowledge pipeline

Red-Teaming Methodology

How to structure an AI red-team exercise — scope, techniques, and reporting

Building a Secure AI System

Defense in depth — layered security for production AI applications

Coming Soon

Course modules

Why AI Testing Is Different

The non-determinism problem — why standard software testing doesn't fully transfer

Designing Eval Suites

What to test, how many examples, and how to ensure coverage

LLM-as-Judge Evaluation

Using models to evaluate models — strengths, weaknesses, and calibration

Human Evaluation Design

When to use humans, how to minimize bias, and annotation best practices

Regression Testing for AI

Catching capability regressions across model updates and prompt changes

Red-Teaming and Adversarial Testing

Finding failure modes before users do — systematic red-team approaches

Benchmark Selection and Interpretation

Choosing the right benchmarks for your system and reading results honestly

Continuous Evaluation in CI/CD

Running evals in your deployment pipeline — tooling and thresholds

In Development

Proposed modules

What Is Algorithmic Bias

Real World Bias Examples

Detecting Bias in Systems

Building Fairer AI Solutions

In Development

Proposed modules

How AI Uses Personal Data

Privacy Risks and Threats

Protecting Your Digital Footprint

AI Governance and Regulation

In Development

Proposed modules

Jobs AI Will Transform

Skills for AI Era

Human AI Collaboration Strategies

Preparing for Career Changes

In Development

Proposed modules

Assistive Technology Applications

Inclusive AI Design

Breaking Down Barriers

Universal Access Principles

In Development

Proposed modules

Self-Driving Vehicle Technology

Autonomous Drones and Robots

Safety and Reliability

Human Oversight Requirements

In Development

Proposed modules

Data Quality and Preparation

Understanding Training Data

Data Sources and Collection

Privacy and Data Rights

In Development

모든 과정

AI 기초

AI 전문 자격증