Google Jules AI coding agent builds and ships real feature in minutes
Google's Jules AI coding agent successfully added and deployed a new WordPress plugin feature while the author worked on other tasks, demonstrating the potential and risks of AI-assisted coding
ZDNET's David Gewirtz tested Google's newly released Jules AI coding agent on his WordPress security plugin, with surprising results
Key Developments
- Feature Implementation: Jules successfully added a "Hide Admin Bar" feature to Gewirtz's plugin in just two prompts
- Time Savings: What would normally take 2-3 hours of manual coding was completed in minutes while the author worked on other tasks
- Real Deployment: The AI-generated code was actually shipped to 20,000 active sites after testing
How It Worked
-
Gewirtz provided Jules with four specific instructions:
- Add a checkbox UI element in the admin interface
- Save the preference setting
- Display the correct checkbox status
- Implement the admin bar visibility logic
-
Jules autonomously:
- Analyzed the entire codebase
- Created an implementation plan (self-approved)
- Modified multiple files while maintaining code conventions
- Pushed changes to a new GitHub branch
Notable Findings
- Beta Limitations: The tool experienced slowdowns and temporary outages during testing
- Risk Factors:
- The AI made changes without explicit approval
- One incomplete instruction initially "bricked" the site
- Potential for large-scale errors without proper oversight
Industry Implications
- Productivity Gains: Could significantly reduce development time for routine features
- Workforce Impact: May eliminate junior developer roles while increasing demand for senior oversight
- Access Concerns: Current free version limits users to 5 requests per day
Expert Perspective
Gewirtz concludes that while tools like Jules represent a major advancement in AI-assisted coding, they:
- Require experienced developers to guide and review changes
- Need robust testing frameworks to prevent errors
- Raise questions about training future developers without mid-level positions
"These tools are going to require programmers, managers, and investors to rethink the software development workflow" - David Gewirtz
For more on AI developments, visit ZDNET's Innovation newsletter
Related News
AWS extends Bedrock AgentCore Gateway to unify MCP servers for AI agents
AWS announces expanded Amazon Bedrock AgentCore Gateway support for MCP servers, enabling centralized management of AI agent tools across organizations.
Jagged AI Already Disrupting Jobs Despite AGI Remaining Distant
Enterprise AI adoption focuses on real-world returns with 'jagged' systems, driving productivity gains but leading to white-collar job cuts across industries.
About the Author
Alex Thompson
AI Technology Editor
Senior technology editor specializing in AI and machine learning content creation for 8 years. Former technical editor at AI Magazine, now provides technical documentation and content strategy services for multiple AI companies. Excels at transforming complex AI technical concepts into accessible content.