24 June 2026

AI hallucination and creativity

The topic provides conceptual depth on AI limitations and ethical governance challenges, serving as excellent context for GS3 technology questions or GS4 ethics in administration essays.

Questions

1. Examine the technical relationship between creativity and hallucination in Large Language Models. How do the mechanisms of probabilistic text generation challenge the pursuit of absolute accuracy in AI systems? 150 words
   Attempt this — 150 words in 8 min

   0 / 150 words 8:00 Reset
2. The emergence of generative AI necessitates a robust framework for verification and accountability. Discuss how institutional mechanisms, similar to those used in science and journalism, can be adapted to mitigate the risks of AI-driven misinformation while preserving the potential for technological innovation. 250 words
   Attempt this — 250 words in 11 min

   0 / 250 words 11:00 Reset

Prelims

1. In the context of Large Language Models (LLMs), what does the 'temperature' setting primarily control?

   1. A The degree of randomness and adventurousness in word prediction
   2. B The total number of parameters used during the training phase
   3. C The speed at which the model retrieves data from a database
   4. D The physical heat generated by the server hardware

2. According to recent research, why are hallucinations considered a 'statistical inevitability' in current LLMs?

   1. A Because training methods reward confident guessing over admitting ignorance
   2. B Because models are designed to prioritize factual accuracy over creative expression
   3. C Because current hardware lacks the memory capacity to store all known facts
   4. D Because LLMs are programmed to intentionally deceive users for security testing