How it works
The framework treats AI as a tool that appears throughout the course rather than as a separate topic. It can suggest causes and solutions from error messages and diagnostic data.
Use a suggestion as a starting point, then confirm it against what the device actually reports. The verification step is yours regardless of where the idea came from.
Common error
A confident explanation is not evidence. If you cannot point at output that supports it, you have a hypothesis rather than a finding.
Worked example
An AI assistant recommends disabling the firewall after one failed ping. Evaluate that advice using the evidence.
Evidence
ping to server: no reply
HTTPS to server: works
Firewall logs: no relevant deny eventA failed ping can result from ICMP filtering even when the required application works. Disabling the firewall would introduce exposure without testing a demonstrated application fault. Verify the application requirement and inspect relevant rules or logs before proposing any exception.
Exercises
1.5.1
An AI tool says a failed name lookup proves that the cable is broken. Name evidence that would distinguish those explanations.
Show solution
A working link and successful access to an approved destination by address weaken the cable hypothesis. Test the configured DNS resolver and compare name lookup results with a working client.
1.5.2
What should be removed before sharing diagnostic output with an external assistant, and who remains responsible for the final change?
Show solution
Remove passwords, tokens, personal identifiers, and confidential internal details not needed for the question. The operator remains responsible for reviewing the recommendation, obtaining required authorization, testing it, and documenting the result.