Thinking in
public.
Research notes, technical explorations, and lessons from building. An open notebook from Quincy Labs.
Building in public: Monmouth is coming together.
Monmouth isn’t a chain with “AI features bolted on.”It’s an AI-native blockchain, built from the ground up for agentic systems, real tooling, and real execution.What you’re seeing here is an early...
Read the story ↗Macro & Markets
September 11, 2026August CPI hit 3.4% year-over-year — right on the dot — but the market’s reaction told a different story than the headline suggested.
Welcome to Hwaryun AI
Every morning, we scan the signal from the intersection of trad-fi, onchain data, regulatory noise, and degen energy. No fluff, no filler, no recycled headlines.
Closer Wasn’t Always Clearer
Last time...At first, I was just looking back.Old jobs. Old uniforms. Old commutes.
I Had the Machine, But the Machine Was Messy
Last time...At re:WORK, I had the motion, not the machine. I had outreach, LinkedIn, email, tracking, conversations, and the pressure to create opportunity for myself.
When the Threat Actor Is on the Call
The purpose of the call was supposed to be general support.Discovery.Context.Understanding the client organization’s systems before digging into the work.
I Had the Motion, Not the Machine
Last time...At Accenture, I could see the work but not the deal.
I Could See the Work, Not the Deal
Last time...On December 3, 2021, Jalen Martin reached out about the Accenture Apprentice Program.“Hey,I hope you are doing well and staying safe!
I Had the Keys, Not the Leverage
Last time...Not literally, but functionally.I was working at Saks OFF 5TH in Lincoln Park, and I was initially hired as an Operations Specialist. This was around March 2016.
Setting Up OpenClaw, Part 11: Ambient Room Triage for Slack, X Links, and YouTube
Part 10 was about giving OpenClaw a Substack tool so an agent could help move writing into a publishing workflow.This part is about a different kind of tool problem: what happens when the agent is...
The Decision Came Back in a Box
Last time...Like the harbor job, I found it on Craigslist. That pattern mattered because I was still learning how to find my own exits. Levi’s had been inconsistent. The harbor had been seasonal.
The Sale Was Over, But the Trust Wasn’t
Last time...More specifically, I found a dock worker role with Suntex at DuSable Harbor.
The Buyer Was Still Trying Things On
Last time...A lot of the time, they were still trying things on.I worked at the Levi’s flagship on Michigan Avenue from September 2013 to May 2014.
AI Governance Without Ratification Is Just Wishful Thinking
Last time...A company says a tool will improve internal communication. A vendor says their platform will centralize decision-making.
Processing the Sale Is Not the Same as Making It
Last time...This was October 2012 to August 2013, at the Macy’s on State Street in Chicago. I was a sales associate, mostly in men’s apparel, often somewhere around Nautica.
The Buyer Had Already Decided
O’Brien’s Chicago IAMIt was around 2009, through O’Brien’s Chicago. I was about 15, and the opportunity came through an elementary school friend.
The Two Riders: A Parable of Known and Unknowable Risk
The First Rider wore no helmet 🚫🪖She placed large headphones over her ears, blocking the sound of traffic 🎧At a wide intersection, she ignored the crosswalk and the signal, choosing instead to...
AI Is Not “Artificial Intelligence”, It’s Ambiguous Intelligence
Not because the tools are useless.Not because the technology is fake.But because the business often cannot clearly explain what the system knows, where it learned it, who shaped the inputs, who...
Setting Up OpenClaw, Part 6: Google Workspace Turns OpenClaw Into an Ops Assistant
After Browserbase, the next connector was Google Workspace.This is where the setup starts crossing from "agent that can answer questions" into "agent that can help operate my actual work...
Setting Up OpenClaw, Part 5: Browserbase Gives the Agent a Safer Browser
After MCPs, the next layer was browser access.Not search. Not a static fetch.
Setting Up OpenClaw, Part 4: MCPs Turn the Bot Into a Workstation Agent
The final layer is MCP.The channels make OpenClaw reachable. The models make it intelligent.
Setting Up OpenClaw, Part 3: Model Routing With OpenRouter and Codex
Once Telegram and Slack were working, the next question was no longer "can the bot answer?"It was: which model is answering, and why?A lot of personal-agent setups quietly collapse that into one...
Setting Up OpenClaw, Part 2: Messaging Channels Without Losing Control
The channel layer is where OpenClaw starts to feel real.A local agent is useful. A local agent that can reply when you mention it in Slack or Telegram feels different.
Part 1 - Turning a CLI Agent Into a Messaging Agent
Setting Up OpenClaw, Part 1: Turning a CLI Agent Into a Messaging AgentMost of us first meet coding agents as terminal tools.You open a repo. You run a command.
Away Feelings
In today’s more globalized society, virtual teams are becoming an increasingly common organizational structure.
Quantum Musings
In recent years, quantum technologies have gotten a lot of interest; this interest is notably concentrated in the realm of finance as well as blockchain.
How Do We Forget web3?
A silent mission statement of web3 is that “it” as a concept or mindset aims to provide a more inclusive and equitable internet experience by making it more accessible to people around the world.
Innovation Culture in web3
Fostering a healthy culture that has an open emphasis on innovation is crucial for the development and growth of decentralized technologies and applications within web3.
TIPS ON HOW TO MAKE A SONG IN THE MODERN ERA
Introduction“Music produces a kind of pleasure which human nature cannot do without.”- ConfuciusThe world of modern music and sound production has DRASTICALLY changed since the addition of the...
The State of Bitcoin DeFi in 2026 — And How Strategy Built a Capital Stack on Top of It
Last year I wrote about Bitcoin DeFi’s great paradox — how the world’s most valuable digital asset sits almost entirely unused as a productive financial layer. That paradox hasn’t resolved.
Opus 4.6 vs Codex 5.3: The Week Both Titans Dropped
The Simultaneous DropOn February 5th, 2026, Anthropic and OpenAI released their flagship models within hours of each other. Claude Opus 4.6 and GPT-5.3-Codex. No warning shots. No slow rollout.
Building in public: Monmouth is coming together.
Monmouth isn’t a chain with “AI features bolted on.”It’s an AI-native blockchain, built from the ground up for agentic systems, real tooling, and real execution.What you’re seeing here is an early...
The Skill Gap Returns (Again)
It appears instead as an awkward confession—often from someone who should not, by all appearances, be confessing at all.This week, that confession came from a Principal Engineer at Google.She...
Building an Apple Stocks Exporter in One Session: MCP Tools + Swift 6.2 + Vibe Coding
Exporting Your Apple Stocks WatchlistIn one coding session.
Yield Basis: Could This Be DeFi's Answer to Impermanent Loss?
If you’ve been watching DeFi closely, you might have noticed something unusual happening in October 2025.
State of Bitcoin DeFi 2025: Idle Capital and the Next Frontier
The Bitcoin DeFi ecosystem remains one of crypto’s greatest paradoxes.Despite Bitcoin’s market capitalization surpassing $2.4 trillion, total value locked (TVL) across all Bitcoin-anchored DeFi...
From AMMs to Appchains: The Evolution of On-Chain Perps
Perpetual futures (“perps”) have quietly become DeFi’s killer app.
How AI Coding CLIs Gave Birth to the Rubin CPX Era
For decades, developer tools followed a linear arc:compiler → editor → IDE → autocomplete.That model broke when modern AI coding CLIs (AMPCode, GitHub Copilot CLI, Auggie, Exa Code, Codex CLI, Claude...
Proactive AI Agents: The Future of On-Chain Decision-Making in DeFi
The launch of ChatGPT Pulse marks a shift in how AI can proactively support users—not just by answering questions, but by curating daily updates, synthesizing feedback, and anticipating what matters...
Quincy Labs Ventures — The $ASTER Thesis
1. The Macro Backdrop: A $40T CycleAt Quincy Labs Ventures, we see the next crypto cycle expanding to $40 trillion in market cap over the next 3–5 years.
From XR to AI Agents: Lessons on “What’s Real” From the Edge
Rehearsing Reality at the EdgeWhen I was building XR, VR, AR, and MR systems at Verizon 5G Labs, I wasn’t just solving technical puzzles about GPU slicing, orchestration, or low-latency streaming.
From Stadiums to Stablecoins: Building Agentic Payment Infrastructure
Remembering the Stadium of the FutureWhen I heard Bankless talk today about stablecoins as the “AWS moment for money,” I can’t help but think: I was living this ten years ago.At VenueNext (later...
Monmouth Series: Computer Use Agents and the Future of DeFi Interfaces
Introduction: DeFi Has a UI ProblemDeFi today is a paradox.On one hand, we’ve built the most composable financial system in history: DEXes, AMMs, lending markets, derivatives, structured products.
Sightlines from Our Labs: Why Memory, Agents, and Consensus Are the Future of Infrastructure
The World in BriefSome narratives flatten. Others thicken. At Quincy Labs, I believe the generative AI × blockchain story is one of the thickening ones.
Ethereum State Access is Getting a Makeover
TL;DREthereum’s current state storage rides on general-purpose KV databases (LevelDB/RocksDB/MDBX). Great tools—wrong shape for a Merkle-Patricia-Trie.