Earlier this spring, Alex Balazs organized an internal hackathon at Business Software Provider Intuit, where the Chief Technology Officer challenged his global team of engineers to embrace artificial intelligence in their work as much as possible.
To give them the tools to do this, a week before the biennial meeting, Balazs approved the use of “every AI coding tool that we are aware of” good, including Qodo, Windsurf and Cursor. He soon saw that thousands of Intuit’s engineers used these tools, higher than hundreds earlier.
“Our statistics show us that engineers love it,” says Balazs. Intuit follows how often the AI coding tools are used, the stability of that use in time and productivity. Early indications are that efficiency buyers can be up to 40%.
On the way to go all-in on AI Codes Assistant Tools, Balaz’s is seduced by a new concept called ‘Vibe Coding’. It is a term that was conceived earlier this year by OpenAI-MEDE founder Andrej Karpathy, where users describe the goals of a project in simple, natural language for an AI-driven coding platform such as Windsurf, which then writes the code.
Balazs recently experimented with atmospheric coding when using windsurf to create a tool with which he could import data from a competitor directly into the Quickbooks software of Intuit. “I noticed that I made considerably more progress with my rusty coding skills than probably otherwise,” says Balazs.
The extent to which the hype ‘atmospheric coding’ ultimately turns out to be a real power that the workflows absorb throughout the business world is still to be seen. But the largest AI hyperscalers in the industry pay very good attention. OpenAi is reportedly in conversation to acquire the startup windsurf of the software for $ 3 billion. Apple and Anthropic have reportedly joined forces to develop their own AI-driven atmospheric coding platform. And more and more code is already written by AI: at Microsoft, no less than 30% today.
Better reasoning models, which can understand the nuances of natural language and which can translate into code, are a key factor that propagates the trend. “The last winter is where the coding wave of the atmosphere really started,” says Alex Albert, head of developer relationships at AI Startup Anthropic. Newer reasoning models, such as Claude 3.7 Sonnet, developers can follow a more hands-off, management approach.
Walter Sun, a worldwide head of AI at the German business software company SAP, says that the conversation around atmospheric coding reminds him of the earliest days of his career. His older colleagues would tell him that he was not a real developer because Sun did not do punch card programming, the use of physical cards to instruct early computers. “I believe that coding atmosphere is only a further evolution of productivity for development,” says Sun.
He says that the principles of atmospheric coding, together with the broader use of AI-compatible coding tools, is an encouraging trend for developers who make generative AI tools to eliminate everyday tasks for the wider workforce. But they must also think of ways to use AI to make their work more efficient. “I think it’s great that developers drink their own champagne,” Sun adds.
In March, SAP rolled out an AI-estimated tool for the developers of the company, intended to accelerate workflows by performing code competition and to offer code explanations to older forms of code that newer developers may not be familiar with. This tool embraces the principles of atmospheric coding, says Sun, and internal analysis shows that code 20% is produced faster than the speed before this tool was rolled out.
Albert says that Anthropic currently has sales conversations with Enterprise customers to explain what an atmospheric coding is. Anthropic has done this type of education around AI agents and explained what these systems are, when and when they should not be used, and the frameworks needed to implement them. “When a company says:” I want to lean in atmospheric coding, “it is more about how we can give our software more power and more control over these AIs and enable them to build much more, much faster,” says Albert.
Todd Olson, co -founder and CEO of Pendo, says that atmosphere coding helped him to inspire to cod again. “It’s harder for me to lead the company if I have no experience with what these solutions do,” says Olson. “This allows me to set a strategy much, much easier.”
About a third of the 900 employees of the software company are engineers and coding the atmosphere penetrates the wider workforce, usually useful in the earliest stages of software development, including prototyping. With a more iterative process, Olson expects Pendo and others to be able to speed up the software development cycle quickly. “Historically it was difficult and slow to get software out,” says Olson. “We are all going to test more.”
Technologists say that coding atmosphere will not replace all the work of the engineer. As an example, all AI-generated code that Pendo can use during the Software Development Lifecycle is not immediately in production. There is still a staff assessment process to ensure that the code that has been generated does not have security patches and attaches itself to the coding standards of Pendo.
That said, AI coding tools are already changing workflows. The primary role of a software engineer is more tilting to edit code, instead of writing it.
“Technology has been happening literally for thousands of years,” says Balazs. And every time it occurs, he argues, employees must make a choice to adapt or not. “All best engineers will use these tools,” he added. “So if you want to be one of the best engineers, use these tools.”
John needs
Send thoughts or suggestions to CIO -intelligence here.
This story was originally visible on Fortune.com
