Writings · quirq

AI Token Spend Needs Its Horsepower Moment

The unit-of-work whitepaper, in short: the mint, the calculus, the ledger, and the environment that owns the ground truth. Readable in full, with every claim tiered against its evidence.

Jul 15, 202627 min read

News

Launches · partnerships · integrations

Quirq Receives $350k Grant from Google for Startups

Google backs the environment layer: $350k in Google Cloud credits and support to scale quirq environments on GCP.

Read the chapter →

Quirq Integrates One-Click Google Antigravity

Google's agent-first IDE, deployable inside a quirq environment in one click.

Read the chapter →

Quirq Receives $200k Grant from Gaia

Gaia backs local intelligence on quirq: your-own-LLM launchpad templates for the fleet.

Read the chapter →

Quirq Enters the NVIDIA Inception Program

Serious compute behind agent workspaces.

Read the chapter →

Quirq Integrates VS Code and Azure Compatibility

Edit environments in VS Code and deploy them on Azure: the Microsoft stack, supported.

Read the chapter →

Quirq Partners on the One-Click OKX Agent

Runtime partner for the OKX Agent Marketplace One-Click Agent.

Read the chapter →

Thoughts

The findings, argued · every note links to its study

The Orientation Tax

What a coding agent's first day on the job actually costs, and why the environment is the only lever on it. On the environment-performance study.

Jul 26, 20265 min read

Agent-Mimicked Synthetic Data Will Never Beat the Real Thing

It never is. On why organic data wins: r = 0.91 vs a coin flip, and the discriminator that can't be fooled.

Jul 25, 20264 min read

Your AI Knows When It's Being Tested

Perfect metrics, going nowhere, screen playing a fake trail. On why alignment testing needs a real environment.

Jul 25, 20264 min read

RTFM: Read the F*cking Manual

We finally found an agent that does, and the reading paid its way. On the curiosity comparison.

Jul 3, 20264 min read

Agents Look a Lot Like Quiet Quitters

Salary paid in full. Discretionary effort: zero. On the incurious agent.

Jun 21, 20264 min read

The Agent She Told You Not to Worry About

No docs, no contract, no memory: nine for nine. On the self-sufficient agent.

Jun 26, 20264 min read

Models Mathematically Find It More Difficult to Navigate Various Languages

The deal closes in fingers, a representation both sides already share. On the tokenizer study.

Jul 11, 20264 min read

Point and Call

One precise gesture beats a thousand words of context. On relevance versus volume.

Jun 26, 20264 min read

The Advantage Was Never the Model

Harvey's edge is the room it puts a model in, not a better model. Routing alone cuts inference cost three to five times. On the Harvey harness.

Jul 18, 20264 min read

The Benchmark That Did Not Finish

Four of ten tasks ran before one of the models was suspended. The verdict never arrived; the method survived. On the coding-agent evaluation.

Jun 14, 20264 min read

The Future of Work

Thoughts · The thesis, with its evidence

How Future Companies Will Measure Work in AI

The three-phase thesis (agentic workforce, collaboration, self-evolving systems) with the evidence map underneath: what held, what moved, and what we still can't claim.

Jul 24, 202612 min read

Guides

Education · the unit of work, explained · read in order

Why quirq

The unit of measurement for AI's business impact: budgeted by a human, minted by verification, dual to the token.

Unit of work

The contract that replaces the prompt: a job with a definition of done, a budget, and a single owner, living in a workspace.

The quirq calculus

Every calculation in quirq accounting: scoring, the mint, the all-in cost model, unit and portfolio metrics, the time axis, and the energy bridge.

The Company Observatory

The quirq ledger a company reads monthly, and the reading discipline that goes with it. A worked quarter where token spend rose 83% while verified work fell.

The environment is the key part

The environment owns the ground truth: it is the specification, the scorekeeper, and the only witness that matters.

Gaming the quirq

The attack surface of a unit of account, in order of severity, and the structural answers to each.

Practitioner playbook

How to write definitions of done, set budgets, and read the ledger schema: the operator's side of the unit.

Validation

Hypothesis-first: every empirical claim with its falsifier, bound to numbered experiments on the research record.

Why the Environment Matters

The research note behind the unit of work: what it argues, and where the full quirq treatment lives.