The Truth Was a Setting
xAI wants to build a truth-seeking AI. But a language model is a coherence engine, and "truth" is a word that hides which kind you mean. What would have to hold for the output to be true?
llms
1essay tagged "llms"
1essay tagged "llms"
xAI wants to build a truth-seeking AI. But a language model is a coherence engine, and "truth" is a word that hides which kind you mean. What would have to hold for the output to be true?