How To Get AI Startup Ideas

Key Concepts

AI Startup School, AI startup ideas, founder market fit, pivoting, expertise, internships, bleeding edge, problem discovery, undercover research, AI agents, B2B SaaS, product-market fit, competitive spaces, technical strength, morale reserves.

AI Startup School Announcement

YC is hosting its first AI Startup School in San Francisco on June 16th and 17th. The conference features AI experts and founders like Elon Musk, Satya Nadella, Sam Altman, Andrej Karpathy, Andrew Ng, and Fei-Fei Li. It's a free conference for computer science grad students, undergrads, and new grads in AI and AI research, with travel covered. Application link is in the description, but space is limited.

The Importance of Persistence in AI Startups

Even if you haven't found the right idea yet, don't give up and go back to Google or college. There's a high probability that your lucky break is just around the corner. This is the most exciting thing you could be doing with your time.

Finding Startup Ideas: Beyond Hackathons and Bandwagons

The default bad startup idea is lazy, like a hackathon idea or jumping on a bandwagon. The best startup ideas are often hard to ship the first version of. Founders should be pushed towards harder, more ambitious ideas.

Two Avenues for Idea Generation: Introspection and Exploration

To find a good startup idea, you need to get out of the house. You have to either aggressively introspect and look within into your history and what you are uniquely great at, or you need to aggressively get out of the house into other places, being industry, being government, being other places that serve Humanity in some way.

Looking Within: Leveraging Unique Experiences and Expertise

Salient: AI Voice Agent for Loan Processing

The founder of Salient, an AI voice agent for auto debt collection, learned about the problem while working at Tesla Finance Ops. He saw the manual process of getting lease payments back and thought of building an AI agent for it. They now service large banks.

Diode Computer: AI Circuit Board Co-pilot

The founders of Diode Computer, building an AI circuit board co-pilot, had a unique intersection of skills as both electrical and software engineers. They saw the gap between software and hardware and the frustrations of hardware engineers, leading them to build a solution.

The PhD Analogy: Reaching the Edge of Knowledge

Starting a successful startup requires you to be similar to a PhD or postdoc researcher in some sense where you have to go all the way to the edge of what human beings know and understand. Instead of publishing research, you're creating a product or service that people really want.

Founder Market Fit: The N=1 Moment

When founders get into this, they have such a unique fit with founder market fit. They're the best in the world literally. There's no one like like them that had that work experience and that happened to want to do a startup that happened to now be really interested in AI. There's this moment in time that is only n equals one that's only them that can do it which is cool.

Spur: AI QA Agent

Spur is building an AI QA agent. One of the founders worked at Figma, which has a complex front end that's hard to test. She realized engineers spent a lot of time testing and maintaining tests, and AI could automate it.

Data Curve (formerly Uncle GPT): AI for Product Managers

Data Curve, founded by 19-year-olds, initially started as a hackathon idea called Uncle GPT, a ChatGPT wrapper. It didn't work, but the founder had interned at Cohere, working on data tools for large language models. She went back to her old boss, identified a need, and built a successful product for product managers.

The Allergic Reaction to Expertise

Founders often have an allergic reaction to what they were doing by definicious because they were experts and worked and grinded out years and years. They're like, "Oh, I don't want to do another decade on this thing that I put all this time." It's like this is so so boring boring and they want to chase some shiny object that they don't know anything about sort of the grass is greener.

The Value of Internships

A huge percentage of YC's billion-dollar companies can be traced back to an internship that one of the founders had. Do internships at really cool companies on the bleeding edge of something.

Being Picky About Workplaces

The founder of Data Curve worked at Cohere, which is at the bleeding edge. The founder of Cohere was one of the authors of the "Attention is All You Need" paper. David AI's founders worked at Scale, which provides data sets for the AI boom.

Looking Outward: Identifying Unmet Needs and Opportunities

Can of Soup: AI Instagram

The founder of Can of Soup, Gabriel, an early engineer at Substack, lost confidence in his initial idea. Chris, the CEO of Substack, advised him to "work on something that captures the human imagination." This led to Can of Soup, an AI Instagram-like social network.

Happenstance: Intelligent Search

The founder of Happenstance, an Apple AI researcher, realized that LinkedIn's search was outdated. Happenstance uses LLMs and vector search to provide intelligent search capabilities.

Thinking Bigger: Overcoming Blinders

Founders should think bigger and consider how the world could change if they 10x their idea. Read or reread Paul Graham's essay called "How to Get Startup Ideas" which is really kind of how definitive work on the on the topic.

Easy Dubs: Universal Translator

Easy Dubs is building the universal translator from Star Trek, providing real-time simultaneous translation.

ESS Health: AI-Powered Back Office for Dentists

One of the founders of ESS Health went to work with his dentist mother for a day and realized the admin work could be automated by an LLM. They built an LLM-powered back office for dentists.

Leveraging Family Connections

Many of YC's billion-dollar startups originated from family connections. Access is all you need.

The Value of AI Agents

AI agents are going to be so much bigger than SaaS. The value of software has increased significantly with AI, as it can replace humans.

Going On-Site: Undercover Research

Anytime you can find anyone working in any industry who will let you do that is freaking gold. You're going to discover something cool.

Happy Robot: AI Agents for Trucking Logistics

The founders of Happy Robot, building AI agents for trucking logistics, are Spanish PhD students. They landed on the idea because they are personable and friendly.

Undercover Jobs: Gaining Insider Knowledge

One of the co-founders of an AI billing company got a job as a medical biller to understand how it works from the inside. He built his own software to automate the work.

Open Source and Legal Automation

Open source is an important piece of this because the reason why this person was able to do it legally was he was building his own software to automate the work all locally on his own computer.

Able Police: AI for Police Paperwork

The founder of Able Police, which works with police departments, discovered that police officers are drowned in paperwork. He went on ride-alongs and investigated the root cause.

Finding Jobs for Automation

Search for jobs on Indeed.com with keywords like "remote analyst" or "remote clerk."

Sweet Spot: AI for Government Contracting

The founders of Sweet Spot had a friend whose job was to refresh a government website for contract bids. They built an AI platform for government contracting.

Outsourcing as a Signal

Any category of jobs that is being outsourced to a low-wage country is like a strong signal that there is a startup to be built right now in the in the current Era.

Lilac Labs: Automating Drive-Thru Orders

Lilac Labs is automating the person at the drive-thru who takes orders, as this job has been outsourced to BPOs in low-wage countries.

Automata: Better RPA

Automata is building a better RPA product that actually works without the need of expensive consultants.

Living at the Edge

PG has this line in this essay uh live at the edge and notice What's missing. You're in a way better position to have a great AI startup idea if you're constantly trying the latest stuff and like actually personally consuming and developing on the the very L stuff.

Pre DB: Pushing the Limits of PG Vector

The founders of Pre DB had friends who wanted a real-time sync between their Postgres database and Pinecone. They realized they could replicate Pinecone's functionality in Postgres using PG Vector.

Redacto: Perfect Chunks and Data from PDFs

The founders of Redacto found an interesting problem for rag applications: the need to extract perfect chunks.

Juice Boox/People GBT: LLM-Powered People Search

The founders of Juice Boox, initially a freelancer marketplace, realized the need for LLM-powered people search for recruiters and built People GBT.

The Disconnect Between Founders and Investors

Investors don't get it because they're trying to do it the way like a Founder would trying to be an ex- influencer trying to just reading feeds from like literally their toilet like and posting right.

Overcoming Fear of Competition

Founders psych themselves out because spaces seem too competitive.

Gig ML: AI for Customer Support

The founders of Gig ML, initially working on edtech, pivoted to fine-tuning as a service. They focused on customer support and secured a deal with Zepto.

The Importance of Technical Strength

So many of the things just actually don't work very well doing like AI that can really replace your customer support team of humans is a hard technical problem and so although lots of people are pitching that they have it very very few people can actually deliver to the level that customers want.

The Changing Landscape of Seed Funding

AI moves so quickly that every few months there's just a new set of possible ideas that I generated.

Morale Reserves in AI Startups

It's so exciting to work on a startup and work on AI right now that the teams just have um morale Reserves for longer.

Conclusion

To find a billion-dollar business using AI, you must either look very deeply within yourself to find that you're already on the edge because of something you've done, or you need to radically get out of the house, go into other people's real businesses and the real problems that Humanity faces, and then get first principal's understanding of what's going on out there.