GPT-6 Sol First Impressions: Faster Builds, Smarter Defaults
Peter compares GPT-6 Sol with GPT-5.6 Sol on the same max-reasoning creative builds. See the reported time, token, quality, and reasoning-effort lessons.
Read Article
Library · Full Archive
AI Automation
AI Automation
Case Study Archive.
Every case study, system breakdown, and field note, newest first and filterable by topic. No theory, no demos. Only systems that run in production and what it took to get them there.
At a glance
What is in the JQ AI SYSTEMS case study archive?
This is the complete archive: all 508 posts across 54 topics, newest first, filterable by category. It holds the case studies, the model and tool comparisons, and the field notes from systems I built and run. For the curated view, where the writing sits alongside the free Claude Code skills, start at the Library instead.
The archive leans in four directions: AI Search Visibility (98), AI Agent Architecture (61), AI Tools (38), and AI Coding Agents (27). Those four account for most of what I publish, because they are where most of the client questions land.
Three places to start. The AI search visibility guide is the hub for the largest cluster and links out to every spoke in it. Grok Imagine vs Midjourney is the image model comparison, rebuilt against live leaderboard data rather than left to go stale. OutreachIQ is the longest running system breakdown here, from first prototype through to a public repo.
Claude Opus 5.5 Review: Anthropic's New AI Workhorse
Matthew Berman reviews Claude Opus 5.5 with an Anthropic engineer. Learn why cost per task matters, which effort level to use, and when Fable 5.1 still fits.
Read Article
Claude Opus 5.5 Tested: Four Builds Against Fable 5.1 and GPT-6 Astra
Pat Simmons blind-tests Claude Opus 5.5, Fable 5.1, and GPT-6 Astra across four real builds. Compare visual quality, 3D work, gameplay, reported cost, and runtime.
Read Article
Grok 4.7 Review: Near-Frontier Coding at a Lower Token Price
Grok 4.7 closes much of the gap on coding and agentic knowledge work. Here is what xAI claims, what independent tests show, and where the value case holds.
Read Article
Fake Jev Demos: What the Decision Model Actually Does
Steve Sewell audits viral Jev demos, explains the decision model's real boundary, and shows stronger uses in email classification, browser routing, and Agent-Native tool selection.
Read Article
Jev Explained: Fast AI Decisions and the Reliability Question
Matthew Berman tests Jev, TypeSafe AI's fast decision model. Learn its typed questions, RLCD training, useful routing patterns, demo limits, and why zero hallucinations is not a safety guarantee.
Read Article
I Pushed GPT-6 Astra Computer Use to Its Limits
A ten-app stress test shows where GPT-6 Astra computer use saves real effort, where direct tools win, and how to delegate visual software safely.
Read Article
GPT-6 Astra: 8 Real-World Use Cases Beyond 3D Demos
Eight creator tests show where GPT-6 Astra is useful now: app building, browser and iPhone control, video editing, meeting analysis, writing, and presentations.
Read Article
OpenAI Just Released an AI That Can Actually Do Your Work
Ten GPT-6 Astra workflows show what computer-using AI can do across robotics, Figma, presentations, research, mobile apps, Blender, Sheets, and product analysis.
Read Article
Claude Just Built a Full 3D House in Blender From One Prompt (Fable 5.1)
A practical review of Claude Fable 5.1 across research decks, animated SVGs, an app, Blender, a product website, and a reusable prompting skill.
Read Article
GPT-6 Astra After 24 Hours: A Practical Builder Review
Ras Mic tests GPT-6 Astra on performance, security, analytics, hardware planning, UI, architecture, and computer use after 24 hours of real work.
Read Article
GPT-6 Astra Developer Impressions: Three Builds, One New Workflow
Ben Davis, Peter Gostev, and Tom Krcha test GPT-6 Astra on a playable history of London, matcha-shop design directions, and a DEF CON puzzle with parallel agents.
Read Article
GPT-6 Astra First Look: Computer Use Is the Real Leap
Matt Wolfe tests GPT-6 Astra on SVG coding, a Three.js game, an interactive world, Blender, and Unreal Engine. Here is what the demos prove, what they do not, and how to evaluate Astra safely.
Read Article
Claude Fable 5.1 Tested: Better Worlds, Harder Economics
Arena AI tests Claude Fable 5.1 against Fable 5, GPT-5.6 Sol, Opus 5, Kimi K3, GLM 5.3, Qwen 3.8, Grok, and DeepSeek across 3D worlds, games, SVGs, and research interfaces.
Read Article
Qwen3.8 Max Tested: One Brilliant Build, Several Expensive Misses
Bijan Bowen tests Qwen3.8 Max across C++, CAD, games, writing, and multimodal coding. See what worked, what failed, what $32 bought, and its open-weight status.
Read Article
Claude Opus 5 in the Wild: Brilliant Agent, Slow Expensive Specialist
AI Search pushed Claude Opus 5 through browser software, 3D reconstruction, financial video, Blender, music, vision, medical images, and research. Here is what the tests actually prove, where the model failed, and when its cost is justified.
Read Article
Your System Next
See something that
See something that
fits your problem?
Every post here is about a system that actually shipped. Book a free call and let's talk about what could ship for you.
Book Free 30-min Call