On January 30, Yelp announced that it too had succumbed to the lure of large-scale language models, otherwise known as “AI.” in a press release, chief product officer Craig Saldana said the review website has added 20 new updates, including “AI-powered business overviews that make it even easier to find businesses that fit your needs at a glance.” It states that These summaries basically take existing reviews and summarize them in a few sentences. Akhil Kuduvari Ramesh, Yelp’s senior vice president of consumer products, said reviews alone won’t get you scrammed. tell the vergeBut “a reliable number of recent reviews” recommended by the software Yelp already uses.
Will this make Yelp even more useful? Maybe. The examples Yelp highlights in their press releases make the summary seem more redundant than anything else. The description for the place “Darwin’s Diner” mentions cheeseburgers and reasonable prices. cheeseburger? Is it for dinner? groundbreaking.
Yelp acknowledges that it has been using AI in various features for “years” to “reduce consumer decision paralysis” when choosing restaurants and bars. Primarily, AI helps highlight dishes that are often mentioned in reviews. However, the paralysis of judgment may be coming from within the home. When I use Yelp these days, I find myself surrounded by irrelevant sponsored content (I recently searched for “Chinese food” and Panera Bread came up), unhelpful community questions, and a lot of trouble deciding whether a particular place is worth it. It felt like a lot of difficult and malicious comments were piling up. your time and money. Twenty years later, is Yelp still working?
Databases for all businesses have always been unwieldy, but Yelp started with a decent idea: a place where everyone could express their opinions and experiences equally. This allows locals to defend their favorite restaurants that may have been ignored by local media, or to give a “real” review of a place that has been subjected to perhaps unwarranted hype. It’s done. In the early days, this created a strong community of restaurant enthusiasts, some of whom became part of the Yelp Elite Squad and Verified Users, whose reviews are considered more trustworthy.
Of course, there were problems. Yelp elite members pressured restaurants to provide free meals or cash In exchange for good reviews, this trend continues among Instagram and TikTok influencers. Since then, there has been a proliferation of “ale bombs,” in which many reviewers pile up reviews on restaurants they have never even visited, based on media coverage. For example, Demi Lovato’s Instagram complaint about frozen yogurt store Big Chill sparked criticism of both. positive and negative reviews It had little to do with the quality of the froyo. And by this point, you’ve probably seen the cryptic reviews for everyone’s beloved restaurant. Someone gave my favorite local restaurant one star because they didn’t have Moscato on the wine list and had an allergic reaction to the shellfish they ordered. I hope you’re okay, but that’s your problem!
A summary of these reviews generated by a large-scale learning model (LLM) can help users know whether an Italian restaurant really serves pasta. But the false assumption Yelp makes and sells is that if you give all opinions equal weight, you’ll somehow get the truth about a restaurant. The argument seems to be that one opinion is one opinion, but now 1,000 opinions are data. This is of course science, it is fact, and it is incontrovertible evidence that the accumulated evaluations are accurate and deserved.
Yes, if you find a restaurant with mostly one-star reviews (that actually talk about the food), you might go elsewhere. But the fundamental truth about restaurant reviews, whether they’re on Yelp, TikTok, or by a James Beard Award-winning critic, is that you’ll never be able to eat there . A good critic invites you into their experience by using their experience and expertise to explain why their meals were successful or unsuccessful in their evaluation. His group of 100 reviews by different diners tells you whether a certain number of people liked the dish. But information is not the same as knowing. know Will you enjoy it until you try it? At some point, you have to choose to go. And his two-sentence summary by AI probably won’t be very helpful.
