What is the main difference between a "Normal" SQL Injection and a "Blind" SQL Injection vulnerability?

Whether the administrator can see the HTTP request to the server is unrelated to the blind vs. normal SQL injection distinction - both attack types produce the same server-side traffic visibility.

Blind SQL Injection does not imply the application properly filters input - the injection still succeeds; the 'blind' aspect refers only to the absence of visible output returned to the attacker.

Whether the administrator sees error messages is not the defining characteristic - the key distinction is that the attacker, not the administrator, receives no direct feedback about the injection results.

In a Blind SQL Injection scenario, the vulnerable application suppresses or does not return database error messages or query results directly to the attacker. The attacker must infer information indirectly - either through boolean-based responses (true/false page differences) or time-based delays - rather than reading data directly from an error or output field.