AI use cases
3056 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.
-
Use Google Maps data for AI-era applications and self-driving cars
— The mapping database is described as a strategically valuable data asset.
-
Assist with content writing
— Handle structure, scaffolding, basic definitions, research support, and sentence flow while human expertise supplies differentiation.
-
Draft content outlines from research notes
— Create a search-aligned structure before writing.
-
Validate and cluster keywords using search data
— Classify search volume, difficulty, and intent, then organize related keywords into content groups.
-
-
Identify content gaps from a website sitemap or competitor sitemap
— Compare existing coverage with topics a similar site should cover and find missing content opportunities.
-
Software development
— AI was described as enabling software to be built more quickly on top of directory traffic.
-
Data enrichment
— Read listing URLs and reviews to identify attributes such as shade and benches at dog parks.
-
Data parsing and filtering
— Help remove irrelevant records from scraped directory data, while preserving high-quality listings through manual review.
-
Sentiment analysis
— Analyze Reddit discussions and comments to help validate whether a directory problem is worth solving.
-
Generate or research ideas for items to rent
— Kyler used ChatGPT and internet searches while experimenting with potential rental categories.
-
Generating content-production assets
— Produced scripts, thumbnails, and teleprompters in Sandy Lee's content platform.
-
Building websites and apps
— Allowed a non-engineer to create a website, a GEO audit app, and other software products.
-
Learning unfamiliar subjects before calls and content production
— Enabled Sandy Lee to understand SEO, automation, and other topics immediately before pitching or teaching them.
-
Developing client models, workflows, and automation concepts
— Helped Sandy Lee propose and deliver SEO and AI-related client work despite lacking prior formal experience.
-
Ad budget optimization
— Mike uploaded a screenshot of Meta Ads Manager to Claude for suggestions about reallocating ad budgets.
-
Ad copy generation
— Mike used Claude to generate headlines and copy for concrete coatings ads.
-
Business research and startup planning
— Mike used AI while determining what it would take to start a concrete coatings business.
-
Automate customer-service email responses
— Draft replies using information from previous emails, frequently asked questions, return policies, and shipping policies.
-
Analyze customer reviews
— Process scraped Amazon reviews and produce summaries of compliments, complaints, recurring phrases, and possible product improvements.
-
Generate and modify website code
— Enable a nontechnical founder to create a simple product interface around an existing API or scraper.
-
Analyze customer sessions alongside the live codebase.
— Distinguish product bugs from user-comprehension gaps and create a faster feedback loop.
-
-
-
Analyze nuanced cohorts, user logs, forecasts, sensitivity models, and regression models.
— Help product managers understand product behavior and make faster decisions with less dependence on data scientists.
-
Generate dummy financial data
— Test the financial application without using real bank statements.
-
Improve prompts by asking an AI system to rewrite them
— Make a basic app-building prompt more detailed and effective for a specified AI coding environment.
-
Build a financial application from a natural-language prompt
— Create an application that accepts bank statements and generates profit-and-loss statements and balance sheets.
-
Create and train voice agents
— Answer questions, handle appointments, follow up with leads, and qualify prospects for a human sales call.
-
Generate and improve workflow instructions
— Build an email-follow-up workflow that checks whether a message was opened and triggers a voice-agent call after a delay.
-
Create lead magnets and business-plan PDFs from a single prompt
— Produce downloadable material for audience members in exchange for their email addresses.
-
Learn about credit card points
— Build enough knowledge to understand the subject
-
Request just-in-time consent
— Obtain human approval before accessing sensitive financial or personal information or taking an action without an applicable policy.
-
Reason about prompts and execute actions across an agentic system
— Allow an agent to autonomously determine how to perform requested work while operating within identity, governance, and consent controls.
-
Suggest explanations for the rise in cancer incidence among women under 49
— The interviewer says he uploaded a graph to AI and asked it about the trend; the AI returned breast cancer incidence, obesity trends and delayed childbearing as possible explanations.
-
Brand-context retention
— Use persistent project memory and local knowledge architecture to produce more consistent brand messaging and outputs.
-
Workflow execution
— Coordinate file retrieval, video generation, permissions, scheduling, and social media publishing.
-
Caption generation
— Create a social media caption describing the property shown in the generated reel.
-
Image-to-video generation
— Convert a product or property photo into cinematic video content for social media.
-
Answer image questions after connecting a frozen Kimi vision encoder to GLM-5.2
— Add vision capability without changing the underlying GLM-5.2 language-model weights
-
Constrain model output with a state machine or grammar
— Ensure structured outputs such as tool calls follow a required format
-
Generate draft tokens for speculative decoding
— Allow the larger model to verify multiple proposed tokens in one forward pass and increase decode speed
-
-
Integrating AI into businesses
— Helping companies use AI as part of their operations and services.
-
Creating social-media systems
— Supporting the production and scheduling of consistent posts.
-
Building software with natural-language instructions
— Allowing a non-coder to create and launch a software product quickly.
-
Managing non-human identities
— Provide AI agents with appropriate identities and privilege levels at machine speed.
-
Generating deepfakes and malware
— Enable impersonation, fraud, and malicious code generation for attacks.
-
Threat detection and response
— Reduce breach costs and shorten the time attackers remain active.
-
Discovering security vulnerabilities
— Identify gaps before attackers exploit them and shorten the discovery-to-exploitation window.
-
Generate app changes from natural-language requests
— Fix or modify app functionality and design without manually writing code.
-
Generate a knowledge base or script for an AI voice agent
— Provide background information, caller-handling instructions, general rules, and script instructions for the agent.
-
Generate detailed prompts for building web applications
— Translate a nontechnical founder's requirements into prompts optimized for a no-code app builder.
-
Conversation intelligence
— Convert recorded conversations into searchable information and extract useful details.
-
Robot model training
— Use teleoperation, human sensor data, egocentric video, general video, simulation, and real-world practice to improve robot control and generalization.
-
Cloud reasoning and historical analysis
— Provide contextual understanding and analyze historical downtime and operational data after real-time functions are handled on the robot.
-
Robot autonomy
— Handle obstacle avoidance, mission execution, data-quality checks, and increasingly general physical tasks.
-
Industrial inspection
— Analyze thermal, acoustic, visual, and gas-sensor data to detect leaks, overheating, equipment conditions, and other anomalies.
-
-
Support vulnerability tracking and patch-management workflows
— Help security teams handle growing vulnerability volumes and prioritize remediation.