Overview of 1041: What are normies using AI for?
Wes Bos and Scott Tolinski explore how regular people—not just developers—are actually using AI in day-to-day life. The big theme: AI is most useful for tedious, messy, highly specific busywork like troubleshooting devices, managing home systems, parsing forms, and organizing data, while it’s much less valuable when it’s used to generate generic text, marketing slop, or fake personalization. They also discuss where AI is already making life easier for hobbyists and self-hosters, and where it still feels overhyped or risky.
Main AI Use Cases They Discuss
Household and personal admin
- Troubleshooting printers, Wi‑Fi devices, and other annoying consumer tech
- Fixing broken or obscure hardware like:
- car parts and VIN lookup
- a skylight calendar / Android tablet with broken Wi‑Fi
- Filling out PDFs, school forms, passport paperwork, and other bureaucratic documents
- Pulling VINs, tracking down reference info, and handling repetitive copy/paste tasks
Home automation and self-hosting
- Managing Home Assistant integrations when they break or expire
- Updating NAS apps and Docker containers
- Organizing files for Plex/Jellyfin
- Checking and repairing home network or server issues through SSH/Tailscale
- Keeping dashboards and YAML/TOML-style config manageable
Calendar and inbox management
- Scanning emails from schools, teachers, and kids’ activities
- Extracting important dates, reminders, and attachments
- Adding events to calendars and inviting family members
- Helping with travel logistics and remembering trip details
Creative and technical workflows
- Video editing assistance:
- selecting clips based on transcript or audio
- generating a rough first-pass timeline
- automating repetitive edit tasks
- 3D printing / CAD help:
- designing brackets, cases, and stands
- fixing Fusion 360 / Blender workflow issues
- improving airflow and structural fit
- Generating or adjusting graphics/infographics for videos and media
Marketplace and resale tasks
- Creating Marketplace listings
- Estimating price points
- Writing plain descriptions from photos
- Researching item details to reduce listing friction
What They Think AI Is Good At
Best at “busywork”
They repeatedly emphasize that AI shines when it:
- handles repetitive admin
- helps with technical troubleshooting
- works through messy data
- saves time on “small but annoying” tasks
Great for non-experts
A big takeaway is that AI lowers the barrier for:
- hobbyists
- home lab users
- self-hosters
- people who are only moderately technical but still want to do technical things
What They Think Won’t Hold Up
Generic text generation
They’re skeptical of:
- long, formal AI-written emails
- mass marketing outreach
- fake personalization
- AI content that exists only to produce more words
AI-generated real estate staging
They push back on AI-staged property photos because:
- the proportions often look fake
- it can be misleading
- buyers usually want the real space, not a stylized mockup
Auto-generated job applications
They note that AI-written job applications are becoming easy to spot:
- everyone is using similar phrasing
- answers sound generic
- automated “personalization” often feels hollow
- referrals and real human connections matter even more now
Risks and Cautions
Prompt injection / calendar access
Scott mentions being cautious about giving AI access to calendars because:
- other people can inject content into shared calendars
- it can create security and trust issues
Hallucination is still a concern
They acknowledge AI can still make things up, but say it’s much better than before for:
- extracting data from emails
- finding known identifiers like VINs
- handling clearly bounded tasks
Cost vs. value
They question whether these use cases are enough to justify huge AI valuations, since many of the tasks are:
- relatively small
- cheap to automate
- not necessarily revenue-generating at scale
Notable Takeaways
- AI is most valuable when it reduces friction in boring, specific tasks.
- “Normies” are using AI less as a chatbot and more as a practical assistant.
- The strongest use cases are often invisible: forms, calendars, troubleshooting, file management, and home systems.
- Generic AI slop is increasingly easy to ignore and may stop working as a strategy.
- Human taste, context, and referrals still matter a lot—especially for jobs and travel planning.
Sick Picks
Wes’s pick
- Silicone chair “socks” with felt bottoms for hardwood floors
- He likes them because they:
- stay on better than sticky pads
- slide smoothly
- protect floors without constant replacement
Scott’s pick
- A backpack from Nomadix
- Highlights:
- lightweight
- lots of compartments
- good for travel
- included a useful fanny pack and microfiber accessories
Final Thought
The episode’s core message is that AI is becoming genuinely useful when it acts like a tireless assistant for tedious real-world problems—not when it’s used to generate polished but empty content. For regular people, the killer app seems to be “make annoying stuff disappear.”