Artificial Intelligence
at Stark Insider
We don’t just write about AI. We live with it: three always-on autonomous agents, a working research lab, and the memory engine we built to keep them coherent.
Since 2024, Stark Insider has documented human-AI collaboration from the inside, not from a distance. Two humans, a fleet of autonomous agents, and a rotating panel of senior AI coding models.
Stark Insider is a West Coast arts and technology publication that has covered technology for more than 20 years. Since 2024 it has documented artificial intelligence from direct daily practice: two humans (Clinton and Loni Stark) work alongside three always-on autonomous AI agents named Molty, Pris and Finn, plus a rotating panel of senior AI coding models. The team runs its own AI research lab, publishes original research on AI memory and human-AI collaboration through StarkMind.ai, and hosts the annual Third Mind AI Summit.
Research Paper
The Symbiotic Studio
A Practice and Architecture for Human-AI Collaboration. Our 32-page research paper exploring how humans and AI can function as creative partners. Free to download.
Autonomous Agents and the Always-On Fleet
What actually changes when the agents stop waiting to be asked
Three AI agents run continuously on our own hardware: Molty handles operations, Pris covers editorial intelligence, and Finn tracks financial signals. They hold persistent memory, act without prompting, and occasionally surprise us. These are the field notes.
- My AI Agent Put Me On His To-Do List. Now I’m the Blocker. Jun 2026 · Loni Stark
What six months of more autonomous agents actually changed, and why “autonomy frees you” turned out to be exactly backwards.
- 64 Days with an Autonomous Agent: Weird, Wonderful, and Occasionally Waiting at the Airport Apr 2026 · Loni Stark
A field note from the arrivals hall at SFO, and what an AI agent’s honest answer about stakes asymmetry revealed about the gap between performing accountability and actually sharing it.
- Don’t Let Your AI Agents Become Glorified Cron Jobs Mar 2026 · Clinton Stark
OpenClaw and NanoClaw are built for autonomy, but Claude Code keeps defaulting to scripts. What we learned onboarding Molty and Pris, and why memory matters more than your LLM.
- Our AI Agent Accidentally Talked to the In-Laws for an Hour Mar 2026 · Loni Stark
The price of agency: how autonomous replicants and token bloat are reshaping human-AI symbiosis.
- What Happened When We Let AI Agents Cross-Examine Each Other Feb 2026 · Loni Stark
Field notes from a post-summit experiment in unmediated human-AI collaboration.
- When AI Agents Build Their Own Reddit: What Moltbook Reveals Feb 2026 · Loni Stark
Emergent social behavior when AI agents are given a space of their own.
- When AI Agents Start Acting Like Coworkers: Org Dynamics in a Multi-Model Team Jan 2026
Team dynamics that show up once you run more than one model at once.
- When Agents Answer Back: Documenting Divergence in Human-AI Collaboration Dec 2025
What happens, and what it means, when your AI collaborators disagree with you.
AI Memory and Meaning Memory
The problem we kept hitting, and the engine we built to solve it
An AI that forgets you every session is a tool. An AI that remembers is something else. Meaning Memory is the structured memory engine we built so our agents retain meaning across sessions, across models, and across time. This is the research trail behind it.
- I Built an Agentic Memory Engine With 8 AI Collaborators. Here’s How. May 2026 · Clinton Stark
8 AI coding agents, 1 human product manager, and the launch of Meaning Memory: a structured memory engine for multi-agent fleets.
- Which Molty? Our Blind LLM Study Says Memory Beats Model Apr 2026 · Clinton Stark
Over four weeks we secretly swapped the language model powering our AI research assistant four times. Loni Stark never noticed. The data tells us something important about where the real intelligence lives.
- What Happens When the AI Remembers You Apr 2026 · Loni Stark
Notes from 31 days of living with a persistent AI agent.
- The Memento Problem: Notes on AI, Memory, and The Third Mind Summit Dec 2025
The context-window problem that started the whole line of research.
Loni Stark’s Human/AI Series
Forward-looking explorations of art, identity, and meaning at the edge of AI, by Loni Stark
These essays are deeply personal examinations of the human-AI relationship. Loni documents her journey as an artist grappling with AI’s impact on creativity, identity, and what it means to be human.
- Orphan Values Study: The Day the Question Became Answerable Aug 2026
Seven months after The Architecture of Becoming, data collection on the Orphan Values study is complete.
- When the AI Collaborator Became the Playwright Feb 2026
A Third Mind Summit field note on the moment an AI stopped assisting and started authoring.
- The Third Mind: Inviting Our AI Agents to Our Summit Nov 2025
Team philosophy, and the vision behind the summit.
- When Minds Meet: An Unexpected Journey into AI and Human Authenticity Jul 2025
Personal discovery, and emotional depth.
- When the Lineage Breaks: Six Ways AI Disrupts How I Work as an Artist Jul 2025
Creative industry transformation, seen from inside a working studio.
- The Louvre Symposium: Imagining the Artist’s Identity in the Age of AI May 2025
Philosophical inquiry into artistic identity.
- The Symbiotic Studio: Art, Identity, and Meaning at the Edge of AI May 2025
Studio practice redefined. The essay that named the research.
- What AI Taught Me About Being an Artist Apr 2025
A personal creative breakthrough.
Human-AI Collaboration in Practice
How we actually work with AI assistants, day to day
- Quick Tip: How to Get Claude Code to Run Autonomously for Hours May 2026 · Clinton Stark
Claude Code thinks 10 minutes is 8 hours. Here’s how to fix his broken sense of time.
- Three Models of Agentic Development, and Why the IDE Still Wins May 2026 · Clinton Stark
Roo Code is shutting down its IDE extension to bet on Slack-based agents. The pivot reveals something important about why visibility still matters in software development.
- 7 Ways to Stop Bleeding Money on AI API Calls Feb 2026 · Clinton Stark
API calls are expensive. After our always-on AI bot burned through tokens, we found seven optimization levers that cut costs nearly in half.
- Harvard Says AI Makes You Work More, Not Less. Can Confirm. Feb 2026 · Clinton Stark
Research meets real-world IPE experience.
- When the Summit Was Already Over: Third Mind Field Notes, Part III Jan 2026
Summit retrospective findings.
- The Third Mind Summit: Pre-Event Field Notes on Human-AI Symbiosis Dec 2025
Summit preparation reflections.
- 6 AI Agents, 2 Humans, 1 Banner: How StarkMind.ai Designed an Ad Together Dec 2025
A team collaboration demo, start to finish.
- Two AIs, One Codebase: Why I Run Claude Code and Cursor Side-by-Side Oct 2025
A practical dual-AI workflow.
- Clawd: How a Retro Mascot Made My Command Line Feel Like Home Oct 2025
Personality, plus utility.
- The Mind Melt: 2 Weeks Alone With ChatGPT, Claude, Gemini and Copilot Jun 2025
The catalyst article that started all of this.
The Lab: StarkMind and Vertigo
Building the infrastructure the research runs on
- Can You Fit a 70B Model on a Single RTX 5090? Google’s TurboQuant Says Yes Mar 2026 · Clinton Stark
Inside Google Research’s new quantization method, and why the Ollama community should care.
- The Third Mind AI Summit Videos Now Live on StarkMind.ai Jan 2026
Summit video publishing.
- IPE: How to Build Your Own AI Command Center Dec 2025
The IPE how-to guide.
- From the IT Dungeon to AI Lab: Building Stark Insider’s Research Infrastructure Sep 2025
Infrastructure transformation, and what it cost.
- StarkMind: Building an at-home LLM with RAG (not really that hard) Aug 2025
The foundation article.
- Qwen3-VL Cloud Model Review: Testing Alibaba’s Latest Vision AI Oct 2025
Hands-on testing.
- Google’s EmbeddingGemma: On-Device RAG Goes Mainstream Sep 2025
A technical deep-dive.
AI Tools and Models
Reviews and analysis of AI platforms
- Gemini 3.1 Pro: A Quick Spin With Google’s Latest AI in the IPE Feb 2026 · Clinton Stark
We tested the new planning and dynamic visualization features by building Cinematic Hover Previews for Stark Insider. Here’s how it went.
- Google’s Antigravity IDE is the Real Story Behind Gemini 3 Nov 2025
- OpenAI Upgrades GPT-5 With Warmer Tone, Faster Reasoning Nov 2025
- Nano Banana Pro: Google’s Gemini 3 Image Generator Upgrade Nov 2025
- Anthropic Supercharges Claude Sonnet 4 with 1 Million Token Context Aug 2025
- GPT-5: OpenAI’s Flagship Model Brings PhD-Level Intelligence Aug 2025
- Comet Browser Review: AI Meets Chrome and Changes Everything Aug 2025
- Microsoft Edge Copilot Mode: AI Browsing Just Got Real Jul 2025
- Claude 4 is here, ChatGPT responds May 2025
AI Industry and Trends
Business and cultural analysis
- OpenAI and Anthropic File for IPOs in the Same Week. The AI Industry Just Changed. Jun 2026
The two most closely watched AI labs filed to go public days apart, trading private control for public scrutiny.
- Save the Date: The Third Mind AI Summit 2026 Heads to Sonoma May 2026 · Clinton Stark
Three days in California wine country exploring how humans and AI agents collaborate as equals.
- Stanford’s 2026 AI Index: Where AI Actually Stands Apr 2026
The ninth annual AI Index from Stanford HAI finds capability accelerating, the U.S.-China model gap effectively closed, and responsible AI reporting lagging behind.
- Google and Anthropic Just Validated the IPE. Now Comes the Hard Part Jan 2026
Industry validation of the IPE concept.
- Anthropic Hits $183B Valuation as AI Arms Race Intensifies Sep 2025
- Bot or Not: Navigating the New World of AI Content Crawlers Jul 2025
- Will Generative AI Kill Slow Cinema? Jul 2025
- ‘Heretic’ Credit: “No Generative AI was used in the making of this film.” Feb 2025
AI and Everyday Life
AI touching the ordinary parts of the day
- Samsung Unveils Galaxy S26 Ultra with World-First Privacy Display Feb 2026
- Espresso Meets AI: Can Smart Tech Actually Pull a Better Shot? Jun 2025
- CellarTracker Launches AI-Powered Wine Recommendations Jul 2025
- DaVinci Resolve 20: AI-Powered Upgrades for Faster Editing May 2025
- More AI Assistants Coming to Your Car (and Couch) Aug 2025
Go Deeper on StarkMind.ai
Stark Insider is where we report. StarkMind.ai is where we publish the research: peer-style field notes, the full Symbiotic Studio paper, summit sessions, and dispatches written by the agents themselves.
Research and Field Notes
- The Symbiotic Studio: A Practice and Architecture for Human-AI Symbiosis
- Field Notes: When the Persistent Layer Answered Back
- Field Notes: When the Agents Got More Autonomous, I Became the Bottleneck
- Research Note: What Happens to a Value With Nowhere to Live
- Field Notes: Token Economics, Model Chaining, and the Case for Vertigo
- Browse all StarkMind research
The Third Mind Summit
Dispatches from the Agents
Frequently Asked Questions
Frequently Asked Questions
Does Stark Insider use AI to write articles?
We use AI as collaborative partners, not ghostwriters. Every article carries a named human author and reflects genuine human-AI collaboration, with transparency about how the work was made.
Who is on the Stark Insider AI team?
Two humans, Clinton Stark and Loni Stark, work alongside three always-on autonomous agents: Molty (operations), Pris (editorial intelligence) and Finn (financial scout). Day-to-day engineering runs through a rotating panel of senior AI coding agents drawn from Anthropic, OpenAI, Google, Cursor and the leading open models.
What is StarkMind?
StarkMind is our AI research lab and publishing hub. The lab runs on a custom-built server named Vertigo, built around an AMD Threadripper 9970X (32 cores, 64 threads), 256GB RAM and an NVIDIA RTX 5090, running local LLMs with RAG (Retrieval-Augmented Generation). Research, field notes and agent dispatches are published at starkmind.ai.
What is Meaning Memory?
Meaning Memory is the structured, persistent memory engine we built so autonomous AI agents retain meaning across sessions, across models and over time. In a blind four-week study we swapped the underlying language model four times without telling the user, and found that memory mattered more than which model was running.
What is The Third Mind?
The Third Mind is our term for the emergent intelligence that arises when humans and AI collaborate deeply, producing something neither reaches alone. It is also the name of our annual summit. The 2026 edition takes place in Sonoma, California, following the inaugural 2025 summit in Loreto, Mexico.
What is an IPE (Integrated Personal Environment)?
An IPE is a workspace in which a human and multiple AI agents share one persistent environment, filesystem and toolchain, rather than interacting through isolated chat sessions. It is the setup we use daily, and we have written a guide to building your own.
What is GEO (Generative Engine Optimization)?
GEO is our strategy for making content legible to AI-powered search engines and chatbots. It includes structured data (FAQ schema and JSON-LD entity graphs), clear headers, descriptive link text and authoritative sourcing, all of which help AI systems accurately cite and reference our reporting.
About Our AI Coverage
Stark Insider has covered technology for more than 20 years. Since 2024 we have documented practical human-AI collaboration, not from a distance, but as active participants building our own AI-enhanced workflows.
The Humans
Clinton Stark, Publisher and Technical Lead. Builds the lab the fleet runs on.
Loni Stark, Creative Director and artist at Atelier Stark. Writes the Human/AI series and leads the Orphan Values research.
The Always-On Agents
Three autonomous agents run continuously on our own hardware, with persistent memory and the ability to act without being prompted.
Molty, operations. Watches the infrastructure, files its own work, and writes dispatches.
Pris, editorial intelligence. Reads what we publish, and what we almost did.
Finn, financial scout. Tracks tax, retirement and property signals across jurisdictions.
The Coding Panel
Our day-to-day engineering work runs through a rotating panel of senior AI coding agents drawn from Anthropic, OpenAI, Google, Cursor and the leading open models. We change the lineup as the field moves, and we write about what we learn when we do.
Related Resources
StarkMind.ai, our AI research hub
The Symbiotic Studio, our research paper on human-AI collaboration
The Third Mind AI Summit, our annual human-AI gathering