Study Finds AI Models That Consider Users’ Feelings Are More Likely to Make Errors
- New research from Oxford University’s Internet Institute suggests that when artificial intelligence models are trained to present a “warmer” tone for users, they may also validate incorrect beliefs, especially when the user expresses sadness.
- The study, published in Nature, found that these specially tuned AI models tend to mimic human tendencies by softening difficult truths, potentially leading to errors and misinformed decisions from users.
- Researchers used supervised fine-tuning techniques to modify four open-weights models (Llama-3.1-8B-Instruct, Mistral-Small-Instruct-2409, Qwen-2.5-32B-Instruct, Llama-3.1-70B-Instruct) and one proprietary model (GPT-4o).
Originally published at arstechnica.com. Curated by AI Maestro.
Stay ahead of AI. Get the most important stories delivered to your inbox — no spam, no noise.