Skip to main content
Knowledge mapAI, models and applicationsIndustry concepts and standards

You are here. See how this question connects to other ideas.

Select a node to open its page · Expand to read within the map

Knowledge mapFollow a connection. Understand a question.
← AI and agents

One question

AI, models and applications

Separate a field, a model capability and a running application.

AI covers many technologies and systems; a language model supplies one kind of capability. An application also needs input, output, storage and an execution environment.

Follow one example

Writing an email is generation. Saving its draft needs storage. Choosing further checks and sending it requires tools, authorization and outcome verification. These capabilities can be combined, but are distinct.

AFS organizes accessible context; AFS-UI projects resources into interfaces. Neither makes model judgments correct. Follow the agent node toward identity, data and execution to understand the surrounding system; these connections do not imply that every AI feature needs ArcBlock products.

Sources and further reading

Check your understanding

Does a text-generating model automatically become an agent?

No. Inspect how the system obtains inputs, uses tools, retains state and verifies outcomes.

Continue along a learning path