"US AI Engineer Uses AI Assistance Tools During Coding Development... Clear Effect on Work Time Reduction"

99% of US AI Developers Use AI Assistants
63% of Developers Save Up to 3 Hours a Day
Over 25% of Google’s New Code Written by AI
IBM, Google, OpenAI Compete in Coding AI
"Entry-Level Developers May Lose Their Place"

It has been revealed that 99% of artificial intelligence (AI) developers use AI assistants (support tools) for programming code writing. Major technology companies are expanding AI services specialized for developers, and there are even forecasts that some entry-level developer positions may be replaced.

According to the IT industry on the 27th, a survey conducted by IBM and Morning Consult at the end of last year targeting 1,063 enterprise AI developers in the United States showed that more than 99% of developers use AI assistants that help with coding during the development process.

During the development process, AI support tools are used to convert the developer’s input into programming code or to verify (debug) errors in the written code. They also perform simple repetitive tasks required in programming or document the code content.

Developers using AI show high satisfaction. According to the survey, 22% of developers using AI assistants reported saving more than three hours a day. The proportion of those who said they reduced their working hours by about one to two hours a day was also 41%. This indicates that AI tools have an effect on increasing development productivity.

The trend of developers utilizing AI tools is confirmed in other surveys as well. According to a report by Google Cloud DevOps Research Team (DORA) in October last year, more than 75% of developers used AI in their daily tasks. They use AI for tasks such as code writing (74.9%) and information summarization (71.2%). In this survey, more than one-third of developers also responded that they experienced productivity improvements. More than 25% of Google’s new code is generated by AI and then reviewed and approved by engineers.

In line with this trend, major companies are also expanding AI services for developers. IBM introduced the AI coding tool 'IBM Watsonx Code Assistant,' which offers features such as code generation and explanation, documentation, transformation, and unit test automation. This tool supports real-time code writing based on IBM’s large language model (LLM) 'Granite.' Developers can write and modify code and even detect errors as if conversing with a chatbot.

Google Cloud also released a public preview version of the personal AI assistant tool 'Gemini Code Assist' for individual developers free of charge. Gemini Code Assist operates based on the Gemini 2.0 model and allows users to write code by inputting commands in various languages, including Korean. The usage limit reaches up to 180,000 times per month, which is close to unlimited, and it can be used in integrated development environments (IDEs) widely used by developers, such as Visual Studio Code (VSCode) and JetBrains IDE.

OpenAI introduced Codex, an AI model that converts natural language (English) into computer programming code, in 2021. Codex has the ability to implement user input interactively in more than 12 programming languages. Codex powers GitHub’s code auto-completion AI service 'GitHub Copilot.' GitHub is a distributed version control system (VCS) platform that allows storing program source code online and serves as a kind of cloud service for developers.

As the use of AI development support tools expands, changes are expected in the corporate developer recruitment market. The Software Policy Research Institute noted in its report 'Changes in the Software Developer Recruitment Market and the Impact of Generative AI' released earlier this year that "the tasks of entry-level developers, who mainly handle repetitive and simple tasks during the development phase, currently overlap with some functions of generative AI development tools," and "concerns are spreading that entry-level developers may be replaced by generative AI."

There is also a forecast that developers’ system design capabilities will become more important with the advent of generative AI, since simple coding tasks can be replaced by AI. The institute predicted in the same report that "the use of generative AI tools is expected to raise the baseline capabilities of entry-level developers," and "evaluations will become more important for problem-solving skills using AI tools, integrated system design capabilities, and collaboration skills with AI rather than simple coding ability."