Sign up to see the future, today<\/p>\n
Can\u2019t-miss innovations from the bleeding edge of science and tech<\/p>\n
Planning your next getaway? You\u2019d be wise not to trust the advice of an AI chatbot \u2014 even if they\u2019re explicitly designed to draw on the testimonies of real travelers.<\/p>\n
As The Guardian reports<\/a>, an AI tool from Tripadvisor designed to summarize the reviews of hotels and other locations on its platforms routinely glosses over some of the horrifying complaints that guests have made about their stays.<\/p>\n The AI described one hotel in Cape Verde as \u201cspotless,\u201d failing to mention that it was being sued for mass food poisoning, according to the findings of the consumer campaign organization Which?, reported in the newspaper. <\/p>\n And for a resort in Turkey facing multiple complaints of sexual harassment, the AI waxed enthusiastic about its \u201cfriendly service,\u201d merely alluding to those serious allegations as \u201clapses [in service] noted by a few.\u201d<\/p>\n \u201cThe platform has a responsibility to revisit the accuracy of its AI summaries and AI chatbot,\u201d Rory Boland, the editor of Which? Travel, said in a statement, per The Guardian. \u201cIn the meantime, users should scroll past these summaries and look at guest reviews, particularly one-star ratings, and at reviews on other sites, to make sure their next stay is a safe one.\u201d<\/p>\n It\u2019s another example of how AI summaries, when they\u2019re not outright spewing<\/a> misinformation<\/a>, frequently fail to give the complete picture. And it\u2019s more egregious when you consider that the Tripadvisor tool\u2019s purpose is to give a roundup of what other guests have said, as opposed to a general purpose chatbot that may draw from more generic information on the broader web, a lot of which may be marketing copy from the hotels and resorts in question.\u00a0<\/p>\n In other words, you\u2019d hope Tripadvisor\u2019s AI would cut through the BS, which is the whole appeal of the customer reviews it\u2019s trained on in the first place. We all jump to the one star ratings to see how bad a place can get, and make a judgment call on whether they\u2019re truly representative of the average person\u2019s experience.<\/p>\n But it seems the chatbot isn\u2019t doing that, which is symptomatic of a problem with the tech at large. According to Duncan Brumy, a professor of human-computer interaction at the University College London, AI tends to \u201csanitize and rub off the edges\u201d of harsher viewpoints, which is likely a consequence of how most of its training data contains bland observations, he said.<\/p>\n \u201cHere you have guests describing a really negative experience, but the AI has decided to tone it down,\u201d Brumby told The Guardian. \u201cIt\u2019s as if it\u2019s being polite.\u201d<\/p>\n Maybe this is by design. Chatbots are infamously sycophantic, tending to agree with the user on top of plying them with flattery. Their safeguards could also push them away from broaching darker topics like sexual harassment, even if that info would be crucial for wary travelers.<\/p>\n Tripadvisor, in a statement to The Guardian, said it automatically suppresses AI summaries for locations that have received complaints about serious safety incidents, including death, drugging, and sexual assault, adding that it\u2019s \u201clooking into the examples where reviews did not match the intended property.\u201d Still, it maintains it\u2019s confident that its AI features \u201care delivering exactly what they were designed to do: help travellers quickly understand the breadth of feedback while making it easy to explore the underlying reviews in full.\u201d<\/p>\n This isn\u2019t the only example of how AI can sabotage tourists, nearly a quarter of whom now rely on it for trip planning<\/a>. AI chatbots have been sending travelers to hallucinated locations<\/a> and putting their safety at risk, such as a pair of Peruvian hikers in the Andes who were turned around by an experienced tour guide when he realized they were in search of a nonexistent \u201csacred canyon\u201d dreamed up by an AI. If you want to know your fellow humans\u2019 honest thoughts on something, best not let it be conveyed to you by a mealy-mouthed machine.<\/p>\n