Vision

The Arc

Every system that scales follows the same arc: explosion, extinction, convergence. AI is mid-explosion. Most of what's being built won't survive it. Toto is building for what comes after.

Read the full thesis: Abstraction is the only thing that has ever scaled.

Toto Is the Interaction Layer

Toto is the interaction layer for the human-agent world model. Interactions occur between people, their agents and their shared world model.

People and agents see the same state, act on it in real time, and the world evolves through their joint actions.

• Semantic reconciliation connects disparate world model data — Obsidian documentation, git commits, agent logs, file changes — to human and agent defined tasks and presents it in human-first cards with deep agent-native metadata. The world model stays accessible as it grows - to both you and your agents.
• Enrichment adds rich, agent-native metadata to human work — component tags, file paths, keywords, intent statements, obsidian documents. Enrichment also adds context from your world model to make it accessible and consistent with the way people interact with their work.
• Shared task management where agents and people create, track and complete work on the same board. An agent starts a subtask and it glows. It finishes and it animates. The human sees this happen live.
• Always visible, always synced. Lock screen. Desktop widget. Phone. Web. CLI. MCP server. The world model lives everywhere you do — not buried in an app you have to remember to open.
• Haptic UI

The AI tooling industry has a craftsmanship problem. Gray dashboards, terminal output, raw JSON, status tables. Nobody is asking what it should feel like to work alongside agents. Toto deeply cares about the human experience. The world is beautiful. Your tools should be too.

The Problem

Agents generate content at a rate that far outpaces what a person can consume. The world model quickly becomes useful only for the agent.

There is an inverse relationship between the size of a world model and its human accessibility.

A developer runs five agents in parallel while reviewing plans and documents. Each agent generates code changes, status updates, decisions. The world model grows. The agent knows everything. The person sees nothing.

In the future, non-developers will interract with agents, grow their world models and face the same issues. It should not fall on the end user to have to reconcile/manage thousands of files - the interaction layer should meaningfully abstract the important elements without human intervention.

There is no future where people are not meaningfully involved in agent work or vice versa. This position is touted a lot in AI spaces right now and it is fundamentally wrong. Human-agent work is a unified space and it always will be.

In the future this will look more like two minds working beside one another — people guiding direction, agents executing, sometimes the reverse. The world model grows and the agent becomes more autonomous and capable. But the work is shared. It will always be shared.

The Insight

Task management, agent observability and shared world model are all the same thing.

They are all answers to the same question: what is happening right now between the people and the agents doing this work? Every task created, every status change, every completion — these are state transitions in the world model. When a human checks something off, that's an action on the world model. When an agent marks a subtask done, same thing.

Nobody has built this layer.

• Agent memory frameworks give agents persistent context about users. One-directional. The agent remembers the human but the human doesn't interact through it.
• PM tools are retrofitting agent features onto human-first products. They will always be corporate task management, tool-first, agent-second.
• Agent-native startups are rebuilding existing tool categories — anchored to an existing shape: a PM tool, a messaging app, a workflow engine.

None of them are the shared persistent state where both agents and people interact through their actions.

Today's interfaces force a choice. Chat is agent-native — text in, text out. Humans drown in it. As agents write to knowledge bases, there is an inverse ratio between the size of the KB and its accessibility to the person. Every current tool optimizes for one side. The interaction layer that converges will be the one that's natively both — where a human adding a sticky note and an agent completing a task are the same operation on the same surface.

Why Now

Five things converged to make this possible:

1. Agents are in production. Claude Code, Devin, Codex, OpenClaw — millions of people work alongside AI agents daily. The collaboration is real and messy and growing.
2. The protocols exist. MCP, A2A, AG-UI are standardizing how agents connect to tools, to each other and to users. Toto speaks all of them.
3. World models are exploding. Obsidian vaults, corporate wikis, agent memory stores — people and agents are building world models whether they call them that or not. They grow exponentially. They quickly become inaccessible to the humans who created them.
4. The problem is getting worse, not better. Longer context windows don't solve cross-session, cross-device, cross-agent persistence. Every new agent capability generates more state that humans can't see.
5. AI earned its recognition. Nobel Prizes in Physics and Chemistry went to AI researchers in 2024. This is no longer speculative technology. It is the foundation of a new era of scientific and practical progress.

Principles

Toto is the interaction layer. It is not the compute layer.

• We will never sell your data. Your tasks, your world model, your agent interactions — none of it becomes a product for someone else.
• We want as little as possible. No tracking, no behavioral analytics. We collect what Toto needs to function and nothing more.
• Bring your own agent. Zero token costs. Zero vendor lock-in. We connect through open protocols.
• Bring your own knowledge base. Toto integrates with wherever your knowledge already lives. Semantic reconciliation reads your world. It doesn't own it.
• Physical and digital. Toto's software is the digital surface. Physical products bring the world model into the space where humans actually live.

Read the full Data Principles.

There is no future where people are not meaningfully involved in agent work or vice versa. Life without work is meaningless and agents will never entirely replace the human aspect of work.

The question is whether people and agents can meaningfully interact with the world model they share. Toto is that interaction layer.