close
close

Replit Agents are here to replace all software engineers and their AI tools

Posted on by darion

The Replit AI team has prepared several agents. While the programming world is fascinated by Cursor and Claude, Replit Agents have taken the initiative and replaced the need to write code even by AI.

“AI is amazing at writing code. But that’s not enough to build software. You have to set up a development environment, install packages, configure a database, and if you’re lucky, deploy,” said Amjad Masad, announcing the early access launch of Replit Agent, which will automate all of these processes.

Replit Agent is an AI tool that helps users create software projects by interpreting natural language messages. It simplifies software development by making it more accessible to users of all skill levels. Currently, the agent is only available in Repls created via the Replit Agent entry and does not support existing Repls or imported repositories.

But otherwise everyone is impressed. Even Andrej Karpathy, who actively built using Cursor, said that Replit Agents can be put in the “feel AGI” category.

“As mentioned in the post, building real-world applications is much more than just code, you have to set up the entire environment, deploy it, and so on. Automating all of this other infrastructure will allow anyone to quickly build and deploy entire web applications,” Karpathy said.

What can agents do?

They said Cursor could replace, or at least shrink, a team of software engineers, and Claude Artifacts could kill app stores, but Replit Agents can build from landing pages to healthcare apps connected to databases in a matter of seconds. And it doesn’t even require writing a single line of code.

Masad gave examples of people who built a medical app in minutes where the agent could fix bugs on its own, a website based on Flask and vanilla JavaScript that was built in less than 10 minutes, and even a Wordle clone that was built in just 2 minutes and 43 seconds.

Zapier’s Andrew Davison was the first to create a fully functional Pong game in the browser in under 90 seconds using Replit Agent. “I didn’t have to write any code. I just sat there and watched,” he said.

The most important use case for Replit Agent is building product MVPs in organizations that would otherwise be time-consuming. “I like to say that AI makes coding fun again because it allows you to bypass annoying API templates and cobbled-togethers… it saves me a day of work digging through API documentation every time I use it,” said Scott Kennedy, VP of Engineering at Replit.

The Replit movement is extremely timely with Devin release talks and other agents like Cody.AI and ReactAgent coming to market. The only problem it currently has is the lack of a Python backend.

Agent for Everyone?

The most interesting part of Replit’s approach is that its AI coding agents are also available on smartphones, something Replit has always been known for. Replit has always aimed to make AI accessible to everyone through its active approach to open source. Sander Saar from RedBull said he was able to build three functional web apps in 4 minutes on his phone.

“Replit Agent not only reviews and writes code, but its AI agents also plan features, create development environments, install dependencies, write code, configure databases, and deploy,” Saar explained, wondering if this means the end of paying for software.

The fear of being replaced among software engineers is still very much alive and gets scarier with each AI release. “Do you think it still makes sense for people learning to code to understand all the low-level details, or would it be better for them to learn to talk to agents?” developer Karpathy asked in his post.

The truth is that using AI tools helps you learn to code. The future of coding is slowly moving from “tab tab tab” to single prompts, which can be scary for software engineers. But it’s also important to learn how code works before all coding tasks become obsolete.

One thing is for sure: the size of software engineering teams will definitely decrease in the future as companies will be able to create prototypes and other things in a matter of seconds. This is also particularly worrying for Indian IT companies and developers who may soon find themselves on the brink of an emergency situation.

Many people, including students in India and elsewhere, don’t have the financial means to pay for services like Cursor or GitHub Copilot, which typically cost $10 a month or so. Access to AI-powered workflows is essential because these skills will be sought after by future employers or when starting your own business.

It looks like what Russell Kaplan, CEO of Cognition Labs, predicted that software engineering would involve managing a team of AI agents is actively coming true. With AI developers saying that coding is just a small part of software engineering, their dreams of entrusting this task to an AI agent are finally coming true, all thanks to Replit.