#22 - AI Coding: Better Than Humans? 🤖

Hey readers! AI coding tools are evolving at breakneck speed, and this week we've got a fascinating lineup of stories that showcase both the impressive capabilities and growing pains of these technologies. From head-to-head comparisons of popular coding assistants to surprising survey results about developer attitudes, there's plenty to unpack in the AI coding landscape.

🔍 This Week's Highlights

Vibe Coding Python: Testing Copilot vs. CodeGPT vs. Tabnine — Jessica Wachtel put three leading AI coding assistants to the test by building a simple Python calculator app with each. GitHub Copilot emerged as the favorite due to its chatbot interface and efficient bug-fixing capabilities, while CodeGPT required step-by-step prompts, and Tabnine offered continuous suggestions but struggled with error handling. — Jessica Wachtel

Warp Code: this is the most innovative AI coding tool — Warp has evolved from a terminal into a unified development environment that integrates coding, AI agents, and terminal tools in one workspace. It excels at supporting agentic workflows by allowing developers to multitask with multiple AI agents simultaneously and understand complex project structures. — Engineering with Utsav

How Warp Went From Terminal to Agentic Development Environment — Warp's transformation includes features like a built-in file editor, code review tools, and WARP.md files to guide agents with project-specific rules, addressing common concerns about bugs and security in AI-generated code. — Frederic Lardinois

Exploring Cursor, Windsurf and Copilot with GPT-5 — This comparative analysis found that all three IDEs successfully generated working code, but differed in workflow ergonomics and developer experience. Cursor excels in polish and multi-file edits, Windsurf leads in context retention, and Copilot offers strong test generation.

📊 AI Impact on Development

How AI can improve the speed and quality of software development — AI tools are automating routine tasks like code generation and debugging, allowing developers to focus on complex problem-solving. AI-driven static analysis prioritizes bugs by severity and context, while predictive testing identifies modules likely to fail before production. — Backlinkshub

Does AI improve developer productivity? — Gaurav Sen reviews a Stanford study analyzing AI's impact across 50,000+ engineers and 2 billion lines of code. The findings show AI boosts productivity by 35-40% in greenfield, low-complexity tasks but offers minimal gains in complex brownfield projects. Benefits are higher with popular languages like Python and Java.

AI Copilot Code Quality: 2025 Data Suggests 4x Growth in Code Clones — GitClear's report reveals concerning trends: copy/pasted code has risen from 8.3% in 2021 to 12.3% in 2024, while refactoring has declined sharply. Despite AI boosting productivity, this trend raises concerns about maintainability and the erosion of DRY principles.

👨‍💻 Developer Perspectives

More than half of developers say 'LLMs can code better than humans' — A survey of 800 senior developers found that 53% believe LLMs can code better than most humans, with 78% using AI tools regularly. While sentiments are mostly positive, concerns remain about data privacy, job security, and code accuracy.

Senior developers let AI do more of the coding — survey — Fastly's survey reveals that 32% of senior developers report over half their shipped code is AI-generated, compared to just 13% of junior developers. Seniors are also more likely to invest time fixing AI code, resulting in faster overall coding speeds. — Paul Krill

Indian-American Techie Weighs In On Impact Of Artificial Intelligence On Software Development — Google ML engineer Harsh Varshney discusses how AI has automated routine tasks, enabling engineers to focus on building smarter systems. While AI tools have changed workflows, human engineers remain essential for complex problem-solving and innovation.

🛠️ Code Review & Quality Tools

Code review bottlenecks usually come from three things — Augment identifies three main code review bottlenecks: noisy/inconsistent checks, time zone delays, and large PRs getting rubber-stamped. Their CLI review bot aims to streamline these processes.

Using the Auggie CLI for automated code review — Auggie CLI automates the first pass of code review, reducing reviewer cognitive load and catching obvious issues early. The article outlines installation, custom rule creation, and CI pipeline integration. — Matt Ball

Without context, code reviewers miss duplication, conventions, and design intent — QodoAI's custom RAG pipeline retrieves relevant code examples from across repositories to provide context during reviews, improving their effectiveness.

Qodo now powers context-enriched code reviews in Bitbucket — Qodo has integrated its AI-powered code review capabilities into Bitbucket Cloud and Data Center, with AI agents that analyze entire repositories to understand best practices and coding conventions.

🔮 Future Trends

Virtual panel: How software engineers and team leaders can excel with artificial intelligence — This panel discusses how AI is transforming development while introducing new complexities. Developers need AI/ML knowledge and prompt engineering skills, while leaders should foster collaboration between humans and AI and establish governance frameworks.

The Impact of AI-augmented Software Development on Testing & DevOps — AI is enabling semi-autonomous software systems that self-monitor and fix issues, human-guided AI development for code generation, and automatic test generation from natural language descriptions, streamlining test-driven development.

Best AI Coding Assistants as of September 2025 — Shakudo categorizes AI coding tools into closed-source commercial products (GitHub Copilot, Amazon Q) and open-source projects (Cline, OpenHands). Many organizations adopt a hybrid approach, leveraging strengths of both categories based on security and customization needs.

We're Joining OpenAI — The AlexCodes.ai team announced their transition to OpenAI's Codex project, signaling continued consolidation in the AI coding assistant space.

Made with ❤️ by Data Drift Press. Have questions, comments, or feedback? Just hit reply!