Ethen

Live

Talk through the work. Keep the thread.

Ethen Chat is the conversational workspace: choose a model, keep artifacts and sources in one thread, and come back to the work.

Open Chat Read the docs

What are you working on?

Interactive replay. No model is called; each scene is scripted from the real Ethen interface.

Continue in Ethen

Work

One thread, from the first question to the finished piece.

Ask, follow up, change direction. The plan, the drafts and the decisions stay in the same conversation, so a longer piece of work does not restart every time you return.

I'm writing a launch brief for our community workshop series. Goal: one page, three audiences, a clear ask. Constraints: no prices yet, dates TBC.

Plan: (1) outline the three audience sections, (2) draft each in its own pass, (3) check the draft against your constraints, (4) keep the decisions we make as we go.

Good. Draft the volunteer section first.

Draft ready. Check against the request: one page ✓, no prices ✓. Missing: the brief says "spring" in one place, but dates are TBC. I've flagged it rather than guessing.

Decisions so far: tone is warm and direct; the ask is "sign up to host one session"; dates stay TBC until you confirm.

Interactive replayScripted example conversation in the Ethen interface.

Files & context

Bring the material. Keep the context.

Attach documents, spreadsheets and images to a message, and group related conversations in Projects so the material stays with the work.

workshop-notes.mdMarkdown · 12 KB
signups-2026.csvCSV · 48 KB
venue-photo.jpgJPG · 1.2 MB

Compare the sign-ups with the notes and tell me which sessions to run first.

Deep Research

Watch a question become a sourced report.

Deep Research in Chat plans the question, searches, reads sources, checks gaps and writes a report with citations and caveats. It is fast and limited; a full research workspace is planned.

What makes an LLM verifier trustworthy?

Deep Research

Research complete3 sources read

Planning (done)
Searching (done)
Reading sources (done)
Analyzing evidence (done)
Checking gaps (done)
Verifying (done)
Writing report (done)

What makes an LLM verifier trustworthy?

A verifier is trustworthy for a task class when its false-accept and false-reject rates are measured on a gold set and stay stable under adversarial inputs [1][2]. Better base models do not remove the need for this; they raise the value of measuring it [3].

Caveats. Sources are Ethen Research Lab publications; none reports measured Ethen results.

How Should We Measure the Reliability of LLM Verifiers? Evaluating the Evaluators: Reward Integrity for AI Agents Why Better Foundation Models May Make Evaluation More Valuable

Artifacts

Make something you can keep.

Briefs, documents and code open in an artifact panel beside the conversation. Switch between the rendered preview and the source, copy it, and come back to it later from Artifacts.

Turn our decisions into a one-page launch brief.

Here’s the brief as a document you can keep editing. Dates and pricing are left as open questions, as agreed.

Workshop series — launch briefDocument · Open in panel

Workshop series — launch briefDocument

Workshop series — launch brief

Who it's for

Neighbours who want to learn, volunteers who want to teach, and local groups who can host.

The ask

Host one session this season. We bring the materials; you bring the room.

Open questions

Dates are to be confirmed. Pricing is intentionally not included yet.

Models

Choose how hard Ethen thinks. Know where each tool lives.

Faros is Chat's conversational model. Fast, Think and Deep set how long it deliberates on your next message. Web and Deep Research run inside Chat; other Ethen products open in their own surfaces.

Ethen tools

WebSearch and read live pagesIn Chat

Deep ResearchMulti-source research with citationsIn Chat

Ethen products

ResearchOpens its own surface CodeOpens its own surface StudioOpens its own surface FlowOpens its own surface ComputerOpens its own surface

Choose models for the work.

Use supported models for different stages. Local execution depends on runtime and configuration. Direct model choice remains available.

Ethen Chat is live. Computer, Flow, Studio and the full model catalogue are separate Ethen products with their own surfaces.

Continue in Ethen.

Review

From intent to outcome.

Review the plan, stage outputs, and available verification notes. Sensitive or state-changing steps need human approval where required. Verification can surface uncertainty; it does not guarantee correctness.

Review · Workshop series launch brief

Plan DoneFour passes, agreed at the start
Stage outputs DoneThree section drafts kept in the thread
Verification note Flagged"Spring" conflicts with dates TBC — flagged, not guessed
Needs your approval Waiting for youShare the brief with the host list

Voice

Speak, and keep the thread.

Voice in Chat keeps spoken exchanges in the same conversation as text: say it, see it in the thread, and follow up either way. It runs on Ethen Voice, the dedicated platform for speech and voice AI.

Explore Ethen Voice Open Chat

“Move the volunteer section above the ask.”

Spoken · added to the thread

One system

Research, intelligence and product compound together.

The loop runs in every direction. Product use raises research questions, evaluations reopen old answers, and some research ends in knowledge that never becomes a product.

Discovers and validates
Ethen Research Lab
The research organization studying intelligence, computing and science, with an explicit evidence status on everything it publishes.
Explore Ethen Research Lab
Turns research into intelligence
Faros
Ethen's model and intelligence-development program: evaluation, routing, data and adaptation, building toward increasingly proprietary capability.
Explore Faros
Puts intelligence to workYou are here
Ethen
The product people use to think, research, write and build with AI.
Closes the loop
Evidence
Outcomes, failures and evaluations become the questions the next round of research starts from.

FAQ

Is this product Faros?

Ethen Chat is the conversational workspace. Faros is the name of Chat's default conversational experience, which currently runs on Meta's Muse model.

How does Ethen differ from Flow?

Ethen helps structure complex work. Flow is designed around processes that can be reviewed and repeated.

Does work run without review?

The product direction centers on bounded steps, visible outputs, and human checkpoints. Autonomous production execution is not established.

Start with the whole task.

Bring your goal into Ethen Chat and work through the next step.

Open Chat Read the docs

FAQ

Is this product Faros?

Ethen Chat is the conversational workspace. Faros is the name of Chat's default conversational experience, which currently runs on Meta's Muse model.

How does Ethen differ from Flow?

Ethen helps structure complex work. Flow is designed around processes that can be reviewed and repeated.

Does work run without review?

The product direction centers on bounded steps, visible outputs, and human checkpoints. Autonomous production execution is not established.

Talk through the work. Keep the thread.

One thread, from the first question to the finished piece.

Bring the material. Keep the context.

Watch a question become a sourced report.

What makes an LLM verifier trustworthy?

Make something you can keep.

Who it's for

The ask

Open questions

Choose how hard Ethen thinks. Know where each tool lives.

Choose models for the work.

From intent to outcome.

Speak, and keep the thread.

Research, intelligence and product compound together.

Ethen Research Lab

Faros

Ethen

Evidence

FAQ

Start with the whole task.

Talk through the work. Keep the thread.

One thread, from the first question to the finished piece.

Bring the material. Keep the context.

Watch a question become a sourced report.

What makes an LLM verifier trustworthy?

Make something you can keep.

Who it's for

The ask

Open questions

Choose how hard Ethen thinks. Know where each tool lives.

Choose models for the work.

From intent to outcome.

Speak, and keep the thread.

Research, intelligence and product compound together.

Ethen Research Lab

Faros

Ethen

Evidence

FAQ

Start with the whole task.