The AI-powered products we’re seeing today are pretty mind blowing. But there are still some practical use cases we’re still waiting on.
In this post, Nathan and Matt share three AI business ideas that you should start today. Well, not all three, just one — unless you want to. Who are we to put a limit on your brilliance?
1. Video Footage Fetching
The Problem
As a video content creator, you probably have hundreds of hours of footage on your devices. One of the biggest challenges with having so much footage is sorting through them to find the most relevant content.
Say you’re a vlogger and you have hours of footage from your vacation last week, which includes a visit to the Eiffel Tower, a family stroll on the beach, and a stop at a donut shop. Finding that footage can feel like a needle in a haystack and content creators often have to dedicate hours just to organize their content.
The Solution
Imagine an AI tool that scans, tags, and organizes your video content and can pull the most relevant video footage based on a query.
Google’s Google Photos app and Apple’s Photos app offer this functionality at a basic level but the features are limited.
This AI tool would be able to:
- Tag and identify recurring people, animals, places and things.
- Identify landmarks, locations, and spaces (like parks, beaches, etc)
- Use audio to fetch videos that mention the query.
- Provide time stamps for when the query is featured or mentioned in the video(s).
This would drastically reduce the time it takes to find footage and allow for better video organization.
Captions.ai recently hinted at having something like this in the works. Until then, there is a need in the market for a product like this.
2. Email Filtering by Intent
The Problem
This idea comes from Siqi Chen, angel investor and founder of the finance startup Runway. As you can imagine, he gets flooded with emails and he uses Sanebox to prioritize and filter them.
And it’s pretty good. In addition to basic spam filters, it can snooze important emails, like if it’s Friday at 5 p.m. or you’re on vacation, until you’re ready to address them. It sends you helpful follow-up reminders and provides a daily digest of unimportant emails. .
The best part is it’s embedded within the Gmail app, so you don’t have to use an external app.
However, it’s missing a level of flexibility and functionality that most providers don’t offer. Most email providers allow you to sort and filter emails based on domain, specific words, and general tone.
The problem is they often take a while to learn your labeling habits and can take constant prompting.
The Solution
What we need in the market – and what Siqi has attempted to cobble together on his own – is an AI-powered email filtering system that quickly and accurately detects an email’s context and can categorize it to the appropriate label based on the prompt you’ve created.
This tool’s key advantage would be its ability to understand the intent of each label/category, categorize based on the email itself and not just the domain or sending, and fine-tune itself based on insights.
For instance, say I receive an email from a fellow parent about my children. That email would ideally go in the “Kids” category. However, in current email providers, it would likely go uncategorized (or miscategorized) as the sender wouldn’t already be assigned to that category.
This tool would be able to understand the context of the email and accurately label it, without input by the user.
3. Daily Digest based on Current Content Consumption
The Problem
If you’re anything like us, your media consumption is incredibly high. From newsletters to podcasts to video channels – it’s easy to get overwhelmed.
There are tools like Jellypod that create automated podcasts based on the newsletters you subscribe to. But what if you wanted to extend it to other mediums?
The Solution
Imagine an AI-powered daily digest that summarizes all of the media you’ve received that day and presents the big ideas with proper sourcing and linking to dig in further.
Each newsletter offers something unique, which is why we’re subscribed in the first place, but it can be hard to keep up. Plus, not every edition will grab your attention.
This tool would make consumption easier without drowning out the quality.
There you have it! Three problems that you could solve by leveraging LLMs. Don't say we've never done anything for you.