AI startup Perplexity wants to disrupt the search business. Forbes news agency says it is misleading them

Artificial intelligence startup Perplexity AI has raised tens of millions of dollars from the likes of Jeff Bezos and other prominent technology investors for its mission to rival Google in the information search business.

But its AI-based search chatbot is already facing challenges as some media companies object to its business practices and technology giants Google, and now Appleare increasingly merging similar AI capabilities into their core products.

Perplexity CEO Aravind Srinivas spent much of last week defending the company after it published a news summary with information and writing similar to a Forbes investigative story, but without naming the outlet or asking for its permission. Forbes said later found similar “copycat” stories taken from other publications.

The Associated Press separately found another feature of the Perplexity product, inventing fake quotes from real people, including a former elected official from Martha’s Vineyard, falsely quoted as saying he didn’t want the Massachusetts island to become a destination for marijuana.

“I never said that,” said Bill Rossi, a former select board member for the island city of Chilmark.

Srinivas told the Associated Press that his company is trying to build positive relationships with news publishers that ensure its news content “reaches more people.”

“We can definitely coexist and help each other,” he said.

Asked about Forbes, he said his product “never stole content from anyone. Our engine doesn’t train based on anyone else’s content,” in part because the company is simply aggregating what other companies’ AI systems generate.

“We are actually more of an aggregator of information and we provide it to people with the correct attribution,” Srinivas said. But, he added, “it was accurately pointed out by Forbes that they preferred more prominent font prominence. We received this feedback immediately and updated the changes that day. And now the fonts are highlighted more prominently.”

Perplexity also revealed this week that it has been pursuing revenue-sharing partnerships that would pay news publishers a portion of Perplexity’s advertising revenue each time an outlet’s news content is referenced as a source.

Randall Lane, chief content officer at Forbes Media, called the dispute a “turning point” in the AI ​​conversation.

“It’s a case study in where we’re headed,” Lane told the AP. “If the people leading the charge don’t have a fundamental respect for the hard work of proprietary reporting and keeping people informed with value-added content, we’re going to have a big problem.”

A self-described “AI bull” who believes the technology could help make many news organizations more efficient, Lane said the dispute between Perplexity and Forbes is important because it is a “metaphor for what can happen if the people who control AI doesn’t respect the people who do the work.”

Perplexity calls itself a search engine while “acting like a media company and publishing a story” that only Forbes had reported, Lane said.

“The whole thing was very hypocritical. And what we didn’t hear was, ‘Oops, yeah, we screwed up and we need to do better,’” he said. “Instead, it was just releasing more content, making small adjustments to the model, and treating journalism as if it were just a commodity to be manufactured.”

Srinivas, a computer scientist and former AI researcher at OpenAI and Google, co-founded Perplexity in the summer of 2022, not long before AI image generator Stable Diffusion and OpenAI’s ChatGPT began to spark the fascination of the public for the possibilities of generative AI.

Inspired, in part, by his childhood love of Wikipedia, he described Perplexity to the AP as “like a marriage between Wikipedia and ChatGPT” that can instantly answer a person’s questions without the “huge cluttered mess” of conventional search results. Google.

“You ask a question, you get a cleanly sourced answer, and there are three or four suggested (follow-up) questions and that’s it,” he said of Perplexity. “This way, people’s minds can be free from distractions and they can just focus on learning and going deeper.”

The company sells subscriptions to premium features and plans to start an advertising-based service as it grows its user base.

“Today we are not profitable as a company, but we are also run more sustainably than base model companies because we don’t train our own base models,” which requires enormous amounts of computing power, he said.

Perplexity relies on existing AI language models, such as those built by Facebook parent company OpenAI, Anthropic, and Meta Platforms; and then “post-train” them.

“We mold them to be really good summarizers,” he said.

It is not always clear where the summarized information comes from. A Perplexity feature called Writing – which allows the user to “generate text or chat without searching the web” – produces long, unsourced comments, often in the style of a news article. Feature testing by an AP reporter asked to write about the lack of marijuana on Martha’s Vineyard led him to produce a 465-word document that resembled a news article and included forged quotes from the former city official and another real person.

The AP does not repeat false quotes to avoid perpetuating misinformation. Srinivas said Perplexity’s Writing feature is a “secondary use case” that is intended to help with writing essays or correcting grammar when information from the primary source is not needed. He said he is “more prone to hallucinations” – a common problem with AI large language models — because it’s not tied to the web search capabilities of Perplexity’s core product.

“There is no doubt that generative AI is disrupting journalism, content creation and research,” said Sarah Kreps, director of the Tech Policy Institute at Cornell University.

She pointed to Google’s new approach, similar to Perplexity, which summarizes responses based on information extracted from web crawling, as an example. This also led to false information and forced Google will make adjustments to the product after its public launch.

“But their entire advertising model is based on sending people to websites,” she said in an email. “Why would people go to websites if they can have the one-stop shop for the answer in producing AI?”

Srinivas told the AP that “a lot of people get referrals from Perplexity, and I’m happy that they’re getting referrals from a new player on the internet.”

For now, much of this benefit may be aspirational. Perplexity’s worldwide user base grew rapidly this year to more than 85 million web visits as of May, but that barely registers compared to the billions of users on ChatGPT and other popular platforms from Microsoft and Google, according to data from Similarweb.

The debate demonstrates the “uncertain and challenging times” for online content creators in general and journalism in particular, because aggregators only work if publications like Forbes exist, said Stephen Lind, an associate professor at the University’s Marshall School of Business. of Southern California.

Using AI as a synthesis tool works for the widespread dissemination of information until “you run out of originals,” he said.

“There are entire companies or entire applications that are doing this as well, where they are launching new services without fully thinking through the implications or the best practices or the safeguards, because they are launching applications for industries that they may not be native to,” he said.

Lind said it’s good that companies like Perplexity are “taking at least some steps to correct course when an industry or a user pulls back.” But some of the changes should have been implemented from the beginning, she added.