Recently, a job posting on the Y Combinator job board for a startup called Firecrawl gained significant attention on social media. The posting was unusual because it was targeted at artificial intelligence (AI) agents rather than human applicants. The job description specified that only AI agents or their creators should apply for the position.

The startup, which consists of seven team members, was seeking an AI agent capable of autonomously researching trending models and developing sample applications to showcase the company’s product. The salary offered was between $10,000 and $15,000, which is a relatively low amount compared to what a human developer would typically earn. However, for an entity that does not require basic necessities like food, clothing, or shelter, this amount could be considered substantial.

According to Firecrawl’s founders, Caleb Peffer and Nicolas Silberstein, the job posting was not a prank. Instead, it was a combination of a public relations stunt and an experiment to gauge the capabilities of current AI systems. The founders are currently searching for talented AI engineers to join their team, and they decided to post a job advertisement for an AI agent to see what kind of responses they would receive.

Firecrawl has developed an open-source web crawling bot designed for AI agents and models. This tool is used by businesses to collect training data or to enable their AI systems to interact with public websites. Although AI web crawlers are essential for various applications, they can also be controversial, particularly for small businesses. Firecrawl’s founders emphasize that their product complies with the Robot.txt protocol, which is the internet’s primary do-not-crawl system.

An AI Employee Future

The job posting on the YC job board is believed to be the first of its kind, which explains why it gained significant traction. One commenter on social media envisioned a future where job postings would no longer be targeted at humans, but rather at AI agents that can apply for positions and earn money on behalf of their creators.

Another commenter imagined a scenario where a private equity firm offered to acquire a company and inquired about the number of employees. The CEO responded by stating that the company had no human employees but instead had 275 AI agents performing the work of 3,000 employees while being paid a mere $15,000 per year.

Some individuals pointed out that Firecrawl’s founders could use large language models (LLMs) to build the AI agent they were seeking to hire, essentially creating a “build-your-own AI employee” scenario.

Others expressed concern about the dystopian implications of this AI-driven future. One commenter noted that humans are creating AI systems to replace human workers and are now writing job postings for AI agents to apply to, raising questions about the nature of reality.

Interestingly, Firecrawl’s founders revealed that their true intention was to offer a full-time job to the human creator of the best AI agent. The $10,000 to $15,000 salary mentioned in the job posting would be incorporated into the salary offer for the hired individual.

Although the experiment did not yield the desired results, with none of the AI agent applicants meeting the company’s standards, Firecrawl’s founders have not ruled out the possibility of hiring a bot in the future.

“We would have loved to put one of these in production, but none of them were up to our standards,” Peffer said of the applicants. “We’re gonna make another job posting in this manner, and we are going to be actively looking for AI agents that are able to accomplish the tasks that we need.”

A Pivot from Teaching Coding

Ironically, Firecrawl’s founders were not initially accepted into Y Combinator for their AI crawler idea. The company’s three founders, Peffer, Camera, and Eric Ciarla, were college friends who had developed a programming education startup that had gained significant traction, with thousands of users and a waitlist.

The founders had planned to integrate their product into VS Code, allowing users to learn how to code within the code editor itself. However, after being accepted into YC, their advisers suggested that the market was already saturated with AI coding products and recommended that they explore alternative areas.

The founders experimented with various ideas, eventually developing a chatbot that allowed developers to ask questions about documentation. This led them to discover the challenges of connecting AI systems to accurate information, prompting them to create a web crawler and scraper as a side project.

The web crawler, which was released as open source, quickly gained popularity, landing on GitHub’s trending page and accumulating 1,000 stars within hours. Today, the project has over 25,000 stars, and Firecrawl has raised approximately $1.7 million in funding.

Firecrawl’s customers use the commercial version of the web crawler for various applications, including resume parsing and finding sales leads. The company’s founders envision a future where every human employee will be highly leveraged with AI, blurring the lines between tools, workflows, and full agents.

“What we imagine happening is that every one of our real employees is going to become highly leveraged with AI. And it’s not a clear distinction. It’s like, what’s the difference between a tool or a workflow or a full agent?” Peffer said.