In some fiction I've encountered humans are treated as being linguistically deterministic. For example, in Arrival humans are strongly linguistically deterministic and a human that learns the alien language is able to escape the linear ordering of time that English imposes. In 1984 Newspeak is a language created for the purpose of limiting human expression particularly those of political will.
Is there any evidence that the language a model is trained in significantly effects any abilities, like that of deception? Or does the language a model is trained on not matter at all?
What is the near-future feasibility of training an AI on a language with a limited amount of publicly available data?
In some fiction I've encountered humans are treated as being linguistically deterministic. For example, in Arrival humans are strongly linguistically deterministic and a human that learns the alien language is able to escape the linear ordering of time that English imposes. In 1984 Newspeak is a language created for the purpose of limiting human expression particularly those of political will.
Is there any evidence that the language a model is trained in significantly effects any abilities, like that of deception? Or does the language a model is trained on not matter at all?
What is the near-future feasibility of training an AI on a language with a limited amount of publicly available data?