|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
15.12.25 - 15:54
|
AI′s Political Persuasiveness Is Dangerously Good: Parmy Olson (Bloomberg)
|
|
|
Scientists found that the latest AI models are getting better at persuading people, but at the expense of the truth, using a debating tactic known as Gish galloping.
Researchers tested 19 language models on nearly 80,000 participants and found that when chatbots were prompted to overwhelm users with information, their factual accuracy declined.
Bloomberg Opinion columnist Parmy Olson discusses the impact of AI and the power of Persuasion on The Pulse. (Source: Bloomberg)...
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|