Raw Content

Article Extract: “He Built an AI Ghostwriter With Taste”

Overview

This article features an interview with Danny Aziz, general manager of Spiral, an AI writing tool developed in collaboration with Every’s editorial team. The podcast episode explores how to build AI that enhances thinking rather than replacing it.

Key Themes

The Problem with AI-Generated Content Danny Aziz rejects the current trend of creating “slop”—content that sounds acceptable but lacks substance. The interview emphasizes that thoughtful creation requires deliberate effort, not shortcuts.

Spiral’s Design Philosophy Rather than rushing to generate multiple drafts, Spiral intentionally slows the writing process. The tool asks clarifying questions before drafting, mirroring how professional ghostwriters work. “We’re calling it a writing partner,” Aziz explains, with “partner” being central to the design.

Multi-Agent Architecture Spiral uses multiple AI agents working in parallel—an interviewer agent that asks questions and a writer agent that drafts. This ensures context isn’t lost when transitioning between thinking and writing phases.

Preserving Creative Exploration The tool presents three different directional options rather than a single “best” answer, allowing users to explore possibilities before committing to a path forward.

Craftsmanship in the AI Era

Aziz argues that craft is about judgment, not authorship. His engineering workflow relies on tools like Droid, running multiple AI agents simultaneously across terminal panes to maintain momentum while staying in control of decisions.

Key Insight: Quality requires making intentional choices, whether the tools are traditional or AI-powered. The focus shifts from who writes code to whether the output meets standards.