This reflection extends Behaviour & Responsibility — deciding what should live in the system and what should live in us.
Human beings have always used tools to extend memory. Writing allowed knowledge to survive beyond the mind. Search engines made information immediately accessible. Artificial intelligence can now retrieve, organise and interpret information on our behalf.
The question is no longer whether technology should support memory. It is what humans still need to remember for themselves.
When information can always be retrieved, remembering may appear inefficient. But memory is not simply storage. What we remember shapes how we recognise patterns, understand context and respond without needing to consult an external system.
A person who remembers the history of a relationship interprets a conversation differently from someone who reads a summary. A professional who has internalised foundational knowledge can recognise when a recommendation appears unreasonable. When too much knowledge remains outside the individual, people may become dependent on systems they are less able to evaluate.
AI can provide an answer, but users need enough internal understanding to recognise whether the answer fits the situation.
Cognitive outsourcing can free the mind for higher-level thinking. Calculators did not eliminate mathematics, and search engines did not eliminate knowledge. The risk arises when support becomes substitution before foundational capability has developed.
Students may ask AI to interpret material they have never read. Professionals may depend on summaries without maintaining knowledge of the underlying field. Organisations may lose institutional memory because everyone assumes the system retains it.
The intelligent age requires deliberate decisions about what can be outsourced and what should remain internally understood. AI should carry information that does not need to occupy human attention. Humans should retain the knowledge required to recognise context, challenge error and connect information with experience.
We do not need to remember everything, but we must remember enough to know when the machine has forgotten what matters.