OpenAI, Meta and Google’s AI chatbots repeated Australian political paedophile conspiracy theory
Politics tamfitronics Artificial intelligence (AI) chatbots from companies like OpenAI, Meta and Google are all regurgitating an Australian conspiracy theory about a legally suppressed list of Australian high-profile paedophiles, highlighting the risks of training generative AI on unverified data scraped from the open web.The world’s largest tech companies have raced to release and incorporate generative AI features into their products that have been built on large language models, a recent advance in AI technology that can analyse enormous troves of data to understand the statistical relationships between words and phrases.These products, best known for powering chatbots, give conversational answers to questions by predicting a response based on training data. While these chatbots can credibly provide information in response to questions not programmed by their creators, it also makes them vulnerable to flaws.One problem is providing...
