AI use cases
3459 ways people actually use AI, compiled from the videos in this
library. Each names the job and why it is worth doing, and links to the
moment in the source video.
-
-
Prepare, schedule, and assign complex maintenance work orders
— Reduce manual preparation and give the maintenance manager a ready-to-approve work order
-
Operate a coding assistant
— Search a repository, open files, run tests, and help debug problems.
-
Build and orchestrate AI agents
— Structure agent instructions, tool use, reasoning loops, memory, workflows, and multi-agent coordination.
-
Connect to external tools and data
— Use APIs, databases, files, web resources, and other external sources to answer questions or perform tasks.
-
Reasoning, communication, and execution across connected systems
— AI agents use an LLM and agentic components to respond to users, interact with an MCP server, and connect to processes or data.
-
Support enterprise workflows
— Help organizations apply agents to changed, better-documented workflows and improve access to information.
-
Automate social engineering
— Match tone and generate deepfakes, making deceptive communications more convincing.
-
Verify patches and find vulnerabilities
— Accelerate defensive testing and identify bugs before they reach production or shortly afterward.
-
Discover and exploit software vulnerabilities
— Scan files for vulnerabilities, chain multiple exploit steps, and conduct longer penetration tests.
-
Generate and execute agent skills
— Use natural language for authoring while compiling deployment logic into structured Python programs with deterministic behavior and selective language-model calls.
-
-
Provide an investment thesis and conditions for selling.
— Make portfolio decisions more systematic and explain the rationale behind each trade.
-
-
Generate stock, ETF, options, and cryptocurrency trade recommendations.
— Manage five separate real-money portfolios in a performance competition.
-
Generate artwork
— Potentially reduce artwork production costs, although Print Our Pet's artwork was still being created by human artists.
-
Generate images for blog posts
— Create visual content for blog publishing.
-
Automate customer-facing business services
— Support potential upsells such as AI voice reception, lead generation, customer service, booking, follow-up, and email marketing.
-
Integrate website functions with business systems
— Connect the website to a CRM, calendar, booking schedule, and related workflows.
-
Build and customize a local-business website
— Produce a handyman website by adapting a reference site, changing the business name and location, and including local cities and suburbs.
-
Generate a detailed website-building prompt
— Create a prompt for an AI website tool that includes CRM, calendar, workflows, SEO, metadata, design, and local service areas.
-
Synthetic data generation
— Create training data for predictive models when real data is scarce or sensitive.
-
Summarization
— Distill long documents into their most important content.
-
Conversational assistance
— Handle customer queries through conversational systems.
-
Code assistance
— Write, explain, and debug code.
-
Content generation
— Create emails, marketing copy, product descriptions, images, and other new content.
-
Forecasting measurable outcomes
— Predict fraud, sales, customer cancellation, inventory demand, equipment failure, or credit default.
-
Scientific assistance
— Brainstorm ideas, iterate on experiments, and improve paper write-ups.
-
Customer support
— Provide live transcription, surface relevant knowledge articles, generate smart replies, and summarize calls.
-
Research and report writing
— Search for relevant references, outline them, summarize their content, and synthesize a comprehensive report.
-
Code generation and testing
— Generate code, run unit tests, and use test results to determine whether the output is correct and guide correction.
-
Generate and select answers
— Repeatedly sample a fixed language model and use a verifier or selection mechanism to identify correct outputs.
-
Debugging
— AI-assisted debugging traces execution paths and surfaces state changes during a run, reducing the need to step through code line by line.
-
Continuous code analysis
— AI-assisted analysis identifies issues during development before they reach testing.
-
Drafting implementations from plain-language instructions
— AI provides a starting implementation based on intended behavior, such as sending a welcome email when a new customer signs up.
-
Generating code suggestions
— AI uses surrounding code and existing code-base patterns to produce suggestions that fit established conventions.
-
Refactoring code
— AI helps update and simplify code within the IDE, including targeting duplication and logic that can be simplified.
-
Generate responses from user prompts.
— Return outputs shaped by framing, constraints, context, prior information, and expressed intent.
-
Predict token sequences and statistical language patterns.
— Produce outputs that appear to demonstrate understanding without inherently grounding symbols in lived reality.
-
-
Generate a web page
— Create a usable web page quickly.
-
Analyze financial data
— Gain insight into how the business is operating.
-
Analyze Facebook advertising data
— Evaluate advertising efficiency.
-
Analyze strings of emails received
— Process thoughts and extract business insights from customer or business communication.
-
Content distribution
— Automatically distribute generated content to connected social platforms.
-
Video format adaptation
— Analyze a viral video's visual style, text overlays, pacing, and subject matter to produce a version in the user's style.
-
Short-form video generation
— Create faceless TikTok videos based on a brief and selected creative direction.
-
Marketing production
— Generate advertising materials, a website package, a brand bible, and campaign content.
-
Brand development
— Create brand identity, positioning, messaging, colors, typography, and marketing assets.
-
Trend research
— Find viral content, formats, hooks, hashtags, creators, and ad-spend insights in a selected niche.
-
-
Compare old and new services
— Detect output differences and provide context about where systems diverged before customers or regulators are affected.
-
Generate test coverage
— Capture real behavior and edge cases before code is changed.
-
Identify safe boundaries for change
— Determine what can be separated, what depends on it, and how a new service should communicate with the existing system.
-
Analyze applications and build working models
— Map call paths, data flows, hidden dependencies, and undocumented processing sequences.
-
Navigate and understand codebases
— Reduce the time developers spend investigating legacy systems and uncover lost system understanding.
-
Skill execution
— Turn natural-language procedures into repeatable agent workflows.
-
Coding and knowledge work
— Increase individual output across coding, design, product management, growth, and other work.
-
Research synthesis
— Read large bodies of material, cross-reference them with the user's interests, and produce usable summaries.
-
Meeting preparation
— Generate preparation documents containing prior conversations, changes, and suggested questions.