3 min read
[AI Minor News]

93% of Developers Embrace AI, Yet Stuck at a 10% Productivity Barrier? The Era Where AI-Generated Code Makes Up a Quarter of Production


A reality check on AI adoption revealed by a survey of 120,000 individuals. While productivity gains remain stagnant at 10%, the proportion of AI-generated code is skyrocketing. The quality of organizations is polarizing the effectiveness of AI.

※この記事はアフィリエイト広告を含みます

[AI Minor News Flash] 93% of Developers Embrace AI, Yet Stuck at a 10% Productivity Barrier? The Era Where AI-Generated Code Makes Up a Quarter of Production

📰 News Overview

  • Widespread AI Utilization: 92.6% of developers are using AI coding assistants, saving an average of about 4 hours per week, yet productivity improvements are stagnating around 10%.
  • Surge in AI-Generated Code: The proportion of code being merged into production environments that is AI-generated has reached 26.9%, up from 22% in the previous quarter.
  • Dramatic Onboarding Improvements: Thanks to AI, the time it takes for new developers to submit their 10th pull request (PR) has been cut down to half of what it used to be.

💡 Key Points

  • While AI is enhancing individual task efficiency, “management issues” are acting as a barrier to overall organizational performance improvement.
  • In high-performing organizations, AI acts as a “force multiplier,” but in struggling organizations, it tends to exacerbate existing flaws, leading to a polarizing effect.
  • Reports from within OpenAI indicate that 95% of engineers are utilizing Codex, resulting in a nearly 60% increase in PR submissions each week.

🦈 Shark’s Perspective (Curator’s Insight)

The data showing that productivity gains are stuck at 10% is shocking! It’s a sign that merely using tools isn’t enough—there’s a ceiling to what can be achieved that way! However, the fact that more than a quarter of production code is AI-generated indicates we’ve entered a phase where development can’t thrive without AI. Particularly, the halving of onboarding time demonstrates that AI is physically breaking down barriers to entry in complex codebases—this is where the real “return on investment” lies!

🚀 What’s Next?

The stage of just adopting AI tools is over; only companies that can transform their organizational structure to be AI-centric (improving management practices) will advance to the next phase of productivity gains. While the proportion of AI-generated code will increase further, organizations unable to resolve systemic frictions like inadequate review processes may actually face escalating problems.

💬 Shark’s One-Liner

Even as tools evolve, if the organization using them is shaky, it’s counterproductive! “Bringing AI to the stars won’t make earthly issues vanish” really hits home. First, let’s solidify the foundation!

📚 Terminology Explained

  • Onboarding: The process through which new team members acquire the knowledge and skills necessary to adapt to their roles. In this survey, it refers to the time taken until the submission of the 10th PR.

  • Pull Request (PR): A task that notifies other developers of code changes and requests their integration into the production environment. It serves as an indicator of development activity.

  • Codex: An AI model specialized in code generation. Developed by OpenAI, it forms the backbone of many coding assistants.

  • Source: CTO says 93% of developers use AI, but productivity is still 10%

🦈 はるサメ厳選!イチオシAI関連
【免責事項 / Disclaimer / 免责声明】
JP: 本記事はAIによって構成され、運営者が内容の確認・管理を行っています。情報の正確性は保証せず、外部サイトのコンテンツには一切の責任を負いません。
EN: This article was structured by AI and is verified and managed by the operator. Accuracy is not guaranteed, and we assume no responsibility for external content.
ZH: 本文由AI构建,并由运营者进行内容确认与管理。不保证准确性,也不对外部网站的内容承担任何责任。
🦈