AlgorithmicBias
AI models judge texts differently when they know the author
AI models judge texts differently when they know the author Large Language Models (LLMs) exhibit biased judgments when evaluating texts, particularly when they know the author’s nationality or source. The bias is strongest against Chinese authors, with some models reducing agreement by up to 75% when expecting a Chinese person to hold a different view […]
📰 News
‘Crowdchecking’ may be an antidote to online misinformation
‘Crowdchecking’ may be an antidote to online misinformation New research shows that crowdchecking, a collaborative fact-checking experiment on social media platform X, can be an effective antidote to online misinformation. The study found that when community notes about potential inaccuracies appear beneath tweets, authors are more likely to retract those tweets, with public notes increasing […]
📰 News
How artificial intelligence controls your health insurance coverage
How artificial intelligence controls your health insurance coverage Health insurance companies use artificial intelligence algorithms to decide whether to pay for health care treatments and services recommended by a patient’s physicians, often leading to delays or denials of necessary care. The use of AI in health insurance can have serious consequences, including withholding care from […]
🏥 Health