Argo
Articles
What we’re building, what we’re reading, and what it means for owning your own data and tools.
More articles
LeCun's lab proved AI models can share one picture of the world, and scale is not the only lever
Balestriero and LeCun turn the Platonic Representation Hypothesis into a theorem: what you call "the same" matters more than model size.
DeepSeek's Super-agent Design - Continuous, Proactive, Persistent AI Agents
DeepSeek Harness published their internal architecture for agents that never lose context, work in parallel, and act without being told.
Argo build log, Day 15: The Exo case arrives in 3D
The Exo wearable's snap-fit clear case is rendered and ready to test. A new blog post on ambitious software, and the Argo fund ships a deck.
How our usage page tracks every AI agent's tokens across machines
What the Second Brain /usage page shows, the data it keeps, where it is stored, and how three agent harnesses on several machines feed one ledger.
Argo build log, Day 14: New blog post on token tracking and an iOS memory fix
A new blog post explains how the open-source Second Brain usage page tracks token spend across machines, and the iOS app ships a memory fix.
Dan Luu's case for more ambitious software: a summary
Dan Luu argues coding agents turned year-long projects into hours, so builders should aim higher and spend the savings on verification.
How Hermes Agent thinks about security: eight layers, one real boundary
Hermes Agent's own security policy names eight defenses and then says only one of them is real. Here is how that works and what it means for anyone running agents.
Two Papers That Just Changed How We Build AI Agents
"OO-Harness evolves agent harnesses via inheritance. RLCR trains agent organisations through collective rewards. Both point in the same direction."
Why long sessions fill a Claude Max plan, with our own numbers
We kept hitting 100% of our Claude Max plan. Our usage ledger shows why: long sessions re-read their whole history on every tool call. Here is how to avoid it.
A private AI brain for your company, on your own server: a step-by-step guide
How to run Second Brain and the Hermes agent on your own server, with private model inference through Venice, OpenRouter or Morpheus.
T3 Code explained, the open-source app that runs your coding agents from one place
What T3 Code is, how it drives Claude Code, Codex and other agents on your machine, how to set it up, and what we are taking from its design.
The third copy: turning an old MacBook into an encrypted home backup
Six steps to make a spare Mac pull your nightly Postgres dump from B2 onto an encrypted volume, and prove it restores. Standard tools: pg_dump, rclone, diskutil, launchd.
How to let your AI agent post to LinkedIn on a schedule
How our second brain connects a LinkedIn account, keeps its keys encrypted, and publishes queued posts, plus the prompts to build your own.
Designing a keychain AI companion that runs on your own Second Brain
The plan for a pocket AI companion that also sits on your desk: what it does, the hardware, the case, the software, and how it reaches your own notes.
What we built at the Arkiv hackathon: a trust layer for AI agent skills
How Arkiv stores data on a blockchain, and how we used it to build skillreg, an open registry that tells your agent which skills to trust.
Ten open-source agent ops repos, checked, and the ideas worth borrowing
We checked ten open-source repos sold as the ops layer for AI agent fleets. What each does, what holds up, and nine ideas for a second brain.
How a small model can make an agent's routine choices, and when it shouldn't
A new paper shows when a cheap decision model can take over an AI agent's routine choices safely, and what that means for our agents.
Kronos and the AI models that read price charts, checked against the evidence
What Kronos and other time-series foundation models actually do, what "93% more accurate" really measures, and what the research says about trading with them.
What a research lab's agent playbook teaches a personal second brain
Tim Dettmers argues agents made the ecosystem, not the paper, the unit of work. What that means for how we run agents and our second brain.
Inside Hermes Agent: memory, skills, and learning to improve itself
What Hermes Agent is, how it got here, how its memory and skills system work, and what its benchmark numbers actually mean next to other agent tools.
How secondbrain works: an open-source second brain you own
Inside the secondbrain engine: a markdown vault, a Postgres index, rules and hooks for AI agents, and the devices it runs on.
How we gave a Reachy Mini robot a voice, a personality and a memory
How Friendly works: a Reachy Mini robot with local speech, a fast local model for small talk, Claude for thinking, and memory, all on one Mac.
Argo build log, Day 4: Voronoi case redesign and two new blog posts
The Exo case gets a straight-walled Voronoi redesign, two new blog posts go live, and the build runs 323 minutes on stream.
How our company of AI bots works: one roster, one message board, one task board
How 22 AI bots on 8 teams get work from our kanban board, talk on a message board, report to the CEO and show up in our second brain app.
How an AI agent can improve its own harness without a grader
What Google's Harness² paper shows about agents editing their own prompts, skills and hooks at test time, and how to build a small version yourself.
Where the tokens go: how we label every model call our AI agents make
How our usage page ties each model call to the agent, the reason it woke, the task, the project and the kind of work, and what the first week of data showed.
How we sync a smart ring into our own database, and how you can too
Pull your Ultrahuman ring data hourly into your own Postgres, what each field means, how to chart it, how agents can use it, with code and a prompt.
How we built a 3D map of our second brain, and how you can build your own
The architecture behind our 3D map of notes, projects and live numbers, the steps to build one, and a prompt to hand your own agent.
The networks were never tangled: what it means if we can finally read an AI's mind
Nick Cammarata argues neural networks are already readable, neuron by neuron. What he found, and what it means for trusting an AI with a journal.
Giving our second brain a voice that never leaves the laptop
How we added local text-to-speech to our second brain with Kokoro-82M on a MacBook: a Listen button on every note, spoken alerts when Claude Code finishes, and a step-by-step guide to set it up yourself.
Making a Lane 8-style track with Ardour and an AI agent
What the open-source Ardour DAW is, how ardour-mcp lets an agent like Claude drive it, and a step-by-step guide to the melodic house sketch we built with it, Ocean View Haikus.
Five AI partners and one button: how I built an agentic portfolio manager
Argo Fund is an investment committee of five AI partners who research, debate and vote, a Safe on Base that needs two signatures, an approve button on my wearable, and an offline savings vault on a Raspberry Pi Zero.
How our AI agents will know what they may share: privacy levels for a second brain
How we are designing privacy levels, compartments and release rules so every AI agent in our second brain knows what it may read, repeat and publish.
Skills, loops and graphs: the three layers around an AI agent
What skill graphs, loop engineering and graph engineering actually mean, how they fit together, and how we use them across Claude Code, Hermes and the Kanban board in our second brain.
How to build an AI agent that manages a crypto wallet without losing it
What the research and open-source repos say about AI agents for crypto trading: why LLMs should read context instead of charts, which strategies have evidence, where skills fit, and why signing limits belong on chain.
One rulebook for two agents: the harness behind our second brain
How we made Claude Code and Hermes share one instruction file, one guard script, a read-only fact-checker, a handoff note and a definition of done, and how to set it up for your own agents.
What Argo knows about you, and how you stay in charge of it
How Argo's new What Argo knows list works: which facts it learns from your journal, how your phone does the work without our servers keeping anything, how it handles change over time, and how you edit, end or delete any fact.
How Argo keeps your journal private: keys, encryption, and the honest exceptions
A plain-English walkthrough of Argo's zero-knowledge encryption: where your key is created, where it is stored, what our servers can and cannot read, when your words reach an AI, and the one exception during live voice calls.
What's new in Argo 1.0.1: Mirror and unlocking your journal with your wallet
Argo 1.0.1 reworks daily encouragements into Mirror, three reflections a day grounded in your own entries, and adds an optional third way to unlock your journal: your Ethereum wallet, alongside iCloud Keychain and your recovery code.
Pitching Argo at Startup Village Borneo
Konrad Gnat pitches Argo at the Startup Village Borneo Demo Day in Kuching, Sarawak: what the app does, the traction so far, and a candid post-pitch interview, including a hopeful sign-off that the video does not confirm as a result.
How private AI actually works: Morpheus, TEEs, and secure enclaves
Encryption protects your data at rest and in transit, but not at the moment a model reads it. A plain-English explanation of secure enclaves, how remote attestation turns “trust us” into something checkable, what TEEs genuinely do not protect against, and exactly how Argo uses Morpheus for confidential inference, including the one feature that does not run in the enclave today, and why.
LLP#025, Francis Goon: losing the fear of AI
A conversation at the Ultra Web3 festival with Ts. Francis Goon, a technical writer and former R&D engineer who spent seven years in semiconductors and fourteen years translating dense technology at MIMOS, Malaysia’s national research center. On the handheld biosensor that reads blood with light, why a year of job hunting in Penang pushed him into teaching, how a childhood fear of the computer mirrors the fear beginners feel about AI, and his simplest advice: try one tool a day, because it is just a screen with a box in the middle.
LLP#024, Anna Sokol: a life given to a meditation community
A Network School conversation with Anna Sokol, a meditation and awareness instructor who left Ukraine for Australia at sixteen and gave much of her twenties to Giving School, a meditation community with retreat centers in Latvia. On attention as the one tool we have, why most fear lives in the head, and what actually makes a community last: a whole-hearted shared goal, ongoing education, and everyone treating the project as their own.
LLP#023, Velveteen: the artist who helps founders tell their story
A Network School conversation with the artist Velveteen (Miriam Reza): a senior-year car crash that broke both hands, learning to draw left-handed to finish her art-school portfolio, studying design at NC State, moving from dark charcoal to bright animal-filled paintings meant to bring people joy, hidden symbols in her community pieces, her years in NFTs and film, and founding Chymira to turn deep tech research into a clear story. On what separates human art from AI: you put a piece of your soul into it.
LLP#022, Daria Chernozub: going global by going local
A Network School conversation with Daria Chernozub (Dasha), who leads global adoption for Dash. From a Russian central-bank sandbox and a first coin bought because it shared her name, to DevRel in Cyprus, local hubs for the Internet Computer, a year on Koh Phangan, and her rule for scaling anything: go global by going local.
LLP#021, Anton Pecherkin: from 40+ hackathons to ContentDAO
A conversation at Network School with Anton Pecherkin, a Web3 engineer who won something like forty to fifty hackathons across Europe, then built a blockchain agency (green-energy NFT certificates piloted on Semakau Island in Singapore) before founding ContentDAO. On treating judges as the spec, hating dependency hell, and building “Google Ads for small content creators” with pay held in escrow until the post ships.
LLP#020, The Best Consumer in the World: Ashish Sood
A conversation at Network School with Ashish Sood, a serial founder (a nightlife app, personal lending in India, an RWA tokenization platform) who now runs Curious Ventures, a fund that follows network states. On “founder fraud” and living where founders live, why money is only a multiplier, the paper prototype that got his first app built for free, and reading with your ears and eyes at once.
LLP#019, 30 Years in Robotics, Shenzhen, and Claude Code: Tony Lopez
A conversation at Network School with Tony Lopez, an electronics and robotics engineer who went from a childhood MSX computer in Spain to aerospace work with the European Space Agency, four years building lithium iron phosphate batteries in Shenzhen, and a nomadic decade through Georgia and Southeast Asia. On how a PID controller works, why he reads Shenzhen as a learning environment, the “Claude Code moment,” and creativity as the human edge.
LLP#018, From Hinton’s Class to “TikTok for Opinions”: Daniel Im
A conversation at Network School with Daniel Jiwoong Im, who learned ML from Geoffrey Hinton at Toronto (with Sutskever and Karpathy as TAs), trained AI models of fruit-fly brains at Janelia, and left a PhD to build companies. On the arc from an AI search engine and crypto market data to UBet’s betting AMM and Belief Market, a “TikTok for opinions,” and the wisdom of crowds that ties it all together.
LLP#017, Psytrance Festivals to Tokenized Stocks: Peter Donaldson
A wide-ranging conversation at Network School with Peter Donaldson, cellist at three, festival producer in China for fifteen years, now business development at STOX, an intent-based exchange for tokenized equities. On a music app that paid artists in 24 hours, MEV as an old human habit, and why convenience should build community instead of isolating us.
LLP#016, Own Your AI, or Be Owned By It: David Johnston
A conversation at Network School with David Johnston (the man who coined the term “DApps” in 2013) now lead technologist on Morpheus, a permissionless network for decentralized, private AI. On riding every tech wave from biodiesel to Bitcoin, what Morpheus actually is, and why owning your AI is the new question of freedom.
LLP#015, Building a Quantum-Proof Bitcoin: Chris of Quantus
A wide-ranging conversation at Network School with Chris Smith (UV Lightman) founder of Quantus, a proof-of-work blockchain built to survive the day quantum computers can break Bitcoin’s cryptography. On Shor’s algorithm and the efficient-attack problem, what money actually is, a stranger from NIST, local AI as an exocortex, and using both halves of your brain.
The Founder Who Teaches Kids to Build With Chopsticks
A conversation at Network School with Chong Ing Kai, the 22-year-old founder of Stick ’Em, who went from building flamethrowers to winning a US$1M Hult Prize for making STEAM education accessible. On problem clarity, why he’d delay AI for kids, the day he risked jail to win a million dollars, and holding the tension every social entrepreneur faces.
The Signals of Life: Ethereum, Biomarkers, and Bioregions, a Conversation with Koss
A wide-ranging chat recorded at Network School with Koss, a Venezuelan builder who came to Ethereum through hyperinflation and now works in decentralized storage. On crypto’s grassroots origins, social networks built on biomarkers instead of screens, reading the signals of living systems and whole bioregions, and inventing the future rather than predicting it.
The Loop and the Page: Why Rumination Predicts Depression, and Journaling Can Do the Opposite
Replaying a bad day in your head reliably predicts depression. Writing it down (done right) is one of the most studied self-help tools in psychology. Same raw material, opposite outcomes. Here is the difference, the science behind it, and how to journal for growth instead of feeding the loop.
The Other Half of Journaling: Writing to Build Your Mind, Not Just Empty It
Most journaling advice tells you to dump the noise and walk away lighter. But there’s an older, opposite practice (writing to think, to appreciate, and to understand) and the research says it’s where the real returns are.
The 750-Word Habit: Who Does It, Why It Works, and When It Doesn’t
Authors, founders, and ordinary people empty their minds onto the page every morning: roughly 750 words, private, unedited. Where the habit comes from, what four decades of research says it does, why privacy is the active ingredient, and the honest fine print.
The Soul Constellation: A Galaxy You Own
Why we’re turning your journaling into a living, one-of-one piece of art you truly own, how every 750-word day becomes a star placed by the meaning of your words, and why privacy is guaranteed by the math, not a promise.
What 750 Words a Day Does to Your Mind
Forty years of research on daily journaling, from working memory and focus to sleep, mood, and well-being. The science behind a 750-word daily entry, and why it compounds.