Books
Science fiction, literary fiction, and creative nonfiction by Issac Daniel Davis. Three titles are published; one more is clearly marked as a draft.
Books · useful AI systems · verifiable work
AetherMoore is Issac Daniel Davis’s independent workshop. Tell Polly what you are trying to do, and the front desk will point you to the right product, tool, evidence page, or human handoff.
Choose by intent
You do not need to understand the whole SCBE, Clay, or AetherDesk stack to use something built from it. Start with what you want to accomplish.
Describe the outcome in ordinary language. Polly will select a public route or prepare a clean handoff.
Ask Polly → 02 / buyUse the short shelf for published books, one-time digital packs, free material, and direct links.
Open shop → 03 / buildBring the problem, current system, constraints, and desired finish line. Start with a fixed scope.
Hire Issac → 04 / verifyInspect public source, measured results, capability boundaries, training receipts, and known nulls.
Check evidence →A simple value ladder
Three levels keep the choice proportionate to the problem. You can support the work with a book, use a finished pack, or begin a scoped engagement.
Science fiction, literary fiction, and creative nonfiction by Issac Daniel Davis. Three titles are published; one more is clearly marked as a draft.
Templates, thresholds, decision records, and starter rollout guidance for teams putting boundaries around AI work.
AI workflow packaging, assistant and site setup, content operations, governance surfaces, or a bounded custom build.
Working products
These systems are at different stages. Some are published packages, some are local tools, and some are working prototypes. Every link opens something you can inspect.
A governed desktop browser with an agent sidepanel, bounded actions, page analysis, and a local control plane. Desktop source is public; the public page is an interactive simulation.
A local-first desktop for bounded AI and human workflows, with registered actions and receipts instead of unrestricted host control.
A packet-and-receipt route for coordinating bounded agents, constrained outputs, checks, and human review.
The public TypeScript and Python package surface for governed automation and risk-scoring experiments.
Runs the same candidate function through Python and JavaScript, reports disagreement with a concrete witness, and can emit a machine-decodable verdict tone.
In active research: Clay is being trained and evaluated, but it is not advertised here as a finished frontier model. Measured progress belongs in the evidence lane.
Evidence before theater
The public site separates what exists, what was measured, what failed, and what still needs work. Use these records before relying on a headline.
The rest of the workshop
The archive still exists. It is grouped by job now, so demos, research, support, story, and system detail no longer compete for the same attention.
Still unsure?
You do not need the right vocabulary. Say what you are trying to finish, what is blocking it, and whether you want to learn, buy, verify, or build. The routing layer handles the organization.
Quick answers
An independent body of books, open-source AI governance work, public research, digital tools, and client services built by Issac Daniel Davis.
No. The public assistant and the four intent routes are designed to get you to a useful surface before exposing deeper architecture.
Yes. Start with the public source, operations evidence, manuals, tools, and delivery proof linked above. Paid downloads remain separate from the open-source repository.