← Back to concept explainers

Concept Fable

Blackboard architecture

Conclusion first:Blackboard Architecture is a collaborative architecture: multiple specialized modules or Agents read the problem, write intermediate results, update hypotheses, and trigger the next processing step through a shared blackboard state. It suits knowledge-intensive tasks with non-fixed steps that require multiple capabilities to gradually approach an answer; the key is the shared state structure, write rules, conflict handling, and audit of state changes.

Blackboard Architecture · Architecture

Read the story first. There is no rush to give a definition here; let the question grow on its own first.

Fable

The blackboard in the center of the hall that becomes clearer as it is written on

When the clinic took on difficult cases, the diagnostician, pharmacist and bookkeeper each looked at their own papers. Minor illnesses were easy, but complex illnesses were often misjudged because information was scattered.

A patient had an old illness, a new injury, and still owed for the previous medicine. The diagnostician saw only symptoms, the pharmacist saw only drug properties, and the accountant saw only constraints.

The curator first had the three people report in turn. When they finished, they dispersed, and the next person started asking from the beginning again. The patient was exhausted from questioning, and the conclusions still conflicted with one another.

He also had each person write long letters to the others. The letters grew thicker and thicker, with key findings buried under pleasantries and repetition.

Later the curator placed a blackboard in the hall, allowing only current facts, assumptions, open questions, rejected options, and next actions to be written. Anyone could read it, and could only change it according to the rules.

The diagnostician wrote down the symptoms, the pharmacist used them to rule out prohibited drugs, and the accountant marked the reimbursable scope. New findings no longer scattered into three booklets, but flowed into the same work surface.

When a hypothesis is disproven, it is immediately crossed off the blackboard; when the next step is clear, the most suitable person takes over. The curator can tell from the blackboard where the case is stuck.

The clinic's knowledge was not concentrated in one person's head, but in a space that could be shared, updated, and traced. The director said: Complex collaboration needs a common working surface, so experts can each contribute without losing the overall state. Later, an even trickier job came along. On the surface it just looked like an old problem magnified, but in reality it forced out all the gaps hidden in the process: who takes over first, who can make changes, how to return to the previous step when something goes wrong—places that used to get by on the tacit understanding of acquaintances now all had to be spelled out. The person in charge didn't urge everyone to be more careful; instead, he walked the whole site again. He had people record every handoff, every failure, and every remedy—which spots were merely slow, which could harm the results, and which, once wrong, would be hard to trace back. After the new rules were rolled out, the first few days didn't go smoothly. Some complained about an extra check; some felt their old experience was being insulted. But when the next anomaly appeared, for the first time everyone could follow the traces to find where the problem had occurred, instead of gathering around the results and guessing at each other.

Reveal

This story is about: blackboard architecture

Blackboard Architecture is a collaborative architecture: multiple specialized modules or Agents read the problem, write intermediate results, update hypotheses, and trigger the next processing step through a shared blackboard state. It suits knowledge-intensive tasks with non-fixed steps that require multiple capabilities to gradually approach an answer; the key is the shared state structure, write rules, conflict handling, and audit of state changes.

What matters is that it is not just a term, but a structural problem that repeatedly appears in real AI / Agent systems. Understanding it lets you judge when to add a model, when to change a process, and when to add governance.

Metaphor mapping

  • Clinic specialists: different capability modules or Agents
  • Blackboard: shared work state
  • Facts, assumptions, open questions: structured intermediate representation
  • Cross out assumptions: state updates and falsification
  • The most suitable person takes over: state-triggered scheduling
  • Curator: the governance and observation layer.

Andy Ren Xu's judgment

The practical value of this concept is that it helps you break "AI features that look smart" into deliverable, verifiable, and governable units of work.

Shared memoryLoose coupling collaborationExpert systemOpportunistic problem solving
京ICP备2026014272号