---
title: "David Loker (@dloker) | Hashnode"
description: "David Loker (@dloker) writes on Hashnode. Read their latest articles and follow their work."
url: "https://hashnode.com/@dloker"
---

@dloker

# David Loker

@dloker · Joined May 2025

### [Gemini 3.1 Pro for code-related tasks: More focus, higher signal-to-noise](https://coderabbit.ai/blog/gemini-3-1-pro-for-code-related-tasks-more-focus-higher-signal-to-noise?utm_source=hashnode&utm_medium=feed)

We evaluated Gemini 3.1 Pro and found it leaves fewer, more focused comments with a higher signal-to-noise ratio, but also surfaces fewer bugs overall

6 min read · [coderabbit.ai](https://coderabbit.ai/blog/gemini-3-1-pro-for-code-related-tasks-more-focus-higher-signal-to-noise?utm_source=hashnode&utm_medium=feed)

[Mar 12](https://hashnode.com/posts/gemini-3-1-pro-for-code-related-tasks-more-focus-higher-signal-to-noise/69b3533193256dfc531935a8)

### [Developers are dead? Long live developers.](https://coderabbit.ai/blog/developers-are-dead-long-live-developers?utm_source=hashnode&utm_medium=feed)

Rumors of developers' death are greatly exaggerated. AI is shifting developers' role from writing code to strategy, reviewing, and defining intent.

8 min read · [coderabbit.ai](https://coderabbit.ai/blog/developers-are-dead-long-live-developers?utm_source=hashnode&utm_medium=feed)

[Feb 12](https://hashnode.com/posts/developers-are-dead-long-live-developers/698e0225855431163f2ba093)

### [Why users shouldn’t choose their own LLM models: Choice is not always good](https://coderabbit.ai/blog/why-users-shouldnt-choose-their-own-llm-models-choice-is-not-always-good?utm_source=hashnode&utm_medium=feed)

Letting users pick LLMs hurts quality, consistency, and cost. Learn why model selection is a data-driven engineering problem—not a user preference.

7 min read · [coderabbit.ai](https://coderabbit.ai/blog/why-users-shouldnt-choose-their-own-llm-models-choice-is-not-always-good?utm_source=hashnode&utm_medium=feed)

[Jan 9](https://hashnode.com/posts/why-users-shouldnt-choose-their-own-llm-models-choice-is-not-always-good/69616c4fc6e8f61a1a374f88)

### [An (actually useful) framework for evaluating AI code review tools](https://coderabbit.ai/blog/framework-for-evaluating-ai-code-review-tools?utm_source=hashnode&utm_medium=feed)

A vendor-neutral framework to evaluate AI code review tools using your own PRs, severity labels, metrics (precision/recall), and a real-world pilot.

11 min read · [coderabbit.ai](https://coderabbit.ai/blog/framework-for-evaluating-ai-code-review-tools?utm_source=hashnode&utm_medium=feed)

[Jan 9](https://hashnode.com/posts/framework-for-evaluating-ai-code-review-tools/696075023d45d04a0f7077ba)

### [Our new report: AI code creates 1.7x more problems](https://coderabbit.ai/blog/state-of-ai-vs-human-code-generation-report?utm_source=hashnode&utm_medium=feed)

We analyzed 470 open-source GitHub pull requests, using CodeRabbit’s structured issue taxonomy and found that AI generated code creates 1.7x more issues.

7 min read · [coderabbit.ai](https://coderabbit.ai/blog/state-of-ai-vs-human-code-generation-report?utm_source=hashnode&utm_medium=feed)

[Dec 17, 2025](https://hashnode.com/posts/state-of-ai-vs-human-code-generation-report/6942d15ecc172c337db09999)
