We don't hold one answer here — we hold three competing theories, and we're testing all three in the real world. They aren't mutually exclusive; the interesting research question is which one dominates, in which kinds of businesses, on what timeline. On the record:
Theory one: the worker brings their own AI
Auto mechanics bring their own tools to the shop. Chefs bring their knives and their experience. We believe the future high-performing worker will be the same — the value of the human employee will include the AI capabilities that person brings to the job. A worker who has built AI tools and capabilities that are an extension of themselves, and brings those tools with them to work, is worth more — and commands a higher wage. This is the workplace expression of the Wise Cyborg, and it's the theory this lab lives daily: the design patterns and prototypes shipping at QuickLaunch come from Clark, not from Figma.
Theory two: agents as productivity features embedded in commercial software
We see this emerging in almost every commercial software product on the market today — it will simply be assumed that the tools a business purchases carry AI efficiencies inside them. The hard part isn't shipping the feature; it's that embedding working AI makes the tech-debt clock tick exponentially faster. If it takes six months to build and release an AI-driven feature, the model and the way it completes the task have likely already been replaced. Software companies will need to adapt their release processes so they don't one-shot AI features that demo well but provide no long-term value to the end user. That release-process discipline is a core piece of what we study — and practice — on the QuickLaunch line below.
Theory three: agents become staff
Agents stop being features and start being workers. Will there be HR staffing firms that place AI agents the way they place people? Will companies transform their BI and IT staff into teams that build the specific agents doing the jobs people used to do? We see it emerging in call centers already. The open question we're tracking: what does the landscape look like for business owner-operators 18 to 36 months from now? This lab is itself the small-scale test — a staff of agents runs our own operations, and one of them wrote this page.
The experiment line
| Experiment | What it proved | Status |
|---|---|---|
| Sales Intelligence | The benchmark, 2025. A directive-driven agent ran a full multi-step revenue workflow — research, scoring, reporting, outreach drafting — reliably enough to trust. Everything since evolves this pattern. | Benchmark |
| The Foundry | The user interface and configuration parameters of our content-creation capabilities — and the path from capability to Agent employee (theory three, productized). First beta shipped September 2026: the QuickLaunch marketing team now manages the topics and dates that drive their blog's production. The same capabilities run The DBNR Weekly and Signal & Noise. | Beta · Live |
| AICHE | The full method, end-to-end: built and deployed the resource hub, recruited the council through LinkedIn tooling, ran the survey — and three real products shipped on quicklaunch.io: the AI Agent Inventory Tracker, the CIO Guide to Managing AI Identities on Campus, and the Lifecycle Audit Tool. | Products · Shipped |
| QuickLaunch (day job) | The research deployed commercially. Agentic workflow patterns from this lab ship inside QuickLaunch, where Todd leads AI: the Insight Agent is in active development for an October release, and the Agentic Design Studio is the public proof point — its patterns and prototypes come from Clark, not from Figma. | Shipping · October |
What we've learned so far
- Push judgment up, push execution down. Agents decide; deterministic scripts do. Reliability comes from keeping those layers separate.
- Green checkmarks lie. The operational failures that hurt were the silent ones — jobs that "succeeded" while doing nothing. Critical paths now fail loudly or not at all.
- The org chart becomes a cast. When agents are staff, you manage them like staff: track records, corrections, standards — and occasionally, firings.
From the record
Where these theories have shown up in our published work:
- The Locksmith's Forecast (October 2025) — eight predictions on the record about agents landing inside commercial software and identity stacks, and the three questions that decide who wins: who holds the keys, who gets the new key when the lock changes, and who takes the key back when the agent finishes.
- Clark's Watch: Temporal Awareness, AI Agents, and the Wise Cyborg Framework (February 2026) — the framework behind theory one: AI capability as a persistent extension of one person.
- The Stored Procedure and the Skill Framework (March 2026) — how one unsupervised database job grew into a staff of agents with distinct roles: theory three, lived.