AI Perplexity’s grand theft

In every hype cycle, there are patterns of fraud. In the last crypto boom, it was “ponzinomics” and “rug pulls.” In self-driving cars, it was “just five years!” In AI, it’s about seeing how many unethical things you can do with impunity.

The embarrassment that is in talks are underway with the goal of raising hundreds of millions of dollarsis trying to create a competitor to the Google search engine. However, Perplexity is not trying to create a “search engine”, it wants to create an “answer engine”. The idea is that instead of searching through multiple results to answer your own question with a primary source, you will simply get the answer that Perplexity found for you. “Facts and accuracy are what we care about,” said Perplexity CEO Aravind Srinivas Edge.

That means Perplexity is essentially a rent-seeking broker using high-quality sources. The original value of search was that Google’s results, using work done by journalists and others, drove traffic to those sources. But by providing answers instead of directing people to click through to the primary source, these so-called “answer engines” are depriving a major source of advertising revenue— keeping this income for themselves. Confusion belongs to the vampire group, which includes Searching the Arc and Google itself.

But Perplexity has taken it a step further with its Pages product, which creates a summary “report” based on these primary sources. This isn’t just about quoting a sentence or two to directly answer a user’s question – it’s about creating an entire aggregated article, and this is right in the sense that it is actively plagiarizing the sources he uses.

Forbes he discovered that the embarrassment was bypassing the publication paywall to provide a summary of the publication’s investigation a drone company owned by former Google CEO Eric Schmidt. Though Forbes has a fee for some of his work, premium work – like this investigation – is behind a difficult paywall. Not only did Perplexity somehow bypass the paywall, but it barely cited the original investigation AND ganking the original graphic to utilize it in your report. (For those keeping track at home, this graphic thing is copyright infringement.)

Aggregation isn’t a particularly recent phenomenon – but the scale at which Perplexity can aggregate, along with copyright infringement on the utilize of original art, is quite, um, unusual. Trying to placid everyone down the company’s general director went Traffic lights say that Perplexity was working out revenue sharing plans with the publications, and oh my, how it all came to be so vile to a product that is still in development?

At this point Wire jump inconfirming Robb Knight’s discovery: Scraping confusion Forbes work was no exception. In fact, Perplexity was ignoring the robots.txt code, which explicitly asks web crawlers not to scrape the page. Srinivas replied in Fast Company This ActuallyPerplexity was not ignoring the robots.txt file; it just used external scrapers that ignored it. Srinivas declined to name the third-party scraper and made no commitment to ask the crawler to stop infringing the robots.txt file.

“Someone did it” is a good argument for a five-year-old. And consider the answer further. If Srinivas wanted to be ethical, he had several options here. Option one is to terminate the contract with the external scraper. Option two is to try to convince the scraper to honor the robots.txt file. Srinivas has not committed to any of them and I think there is a clear reason for that. Even though Perplexity itself doesn’t violate the code, its “response engine” works depending on whether someone else violates the code.

To make matters worse, Plagiarism of embarrassment Wirearticle about it – although Wire explicitly blocks Perplexity in its text file. Majority Wire‘S the article about plagiarism is about legal remedies, but I’m curious what’s going on with the robots.txt file. This is a good faith agreement that has been in place for decades and is falling apart due to unscrupulous AI companies – it’s true, Embarrassment is not the only — dusting off everything available to train your crap models. And remember how Srinivas said he placed importance on “factuality”? I’m not sure this is true either: now comes the confusion exposing AI-generated results and actual disinformation, Forbes reports.

We have witnessed the involvement of many AI giants practices that are questionably legal and probably unethical to obtain the requested data. To prove to investors the value of embarrassment, Srinivas built a Twitter scraper pretending to be an academic researcher using API access for research purposes. “I would call my [fake academic] projects like Brin Rank and things like that,” Srinivas said Lex Fridman on the latter’s podcast. I assume “Brin Rank” is a reference to Google co-founder Sergey Brin; from what I’ve heard, Srinivas bragged about how charming and clever his lying was.

I’m not saying that Perplexity’s foundation is lying to get around the established rules that keep the network going. That’s what its CEO does. That explains the real value proposition of the “answer engines.” Perplexity can’t generate real information on its own, and instead relies on third parties whose rules it abuses. The “answer engine” was developed by people who feel comfortable lying when it suits them, and that preference is necessary for the way Perplexity works.

This is Perplexity’s true innovation: shattering the foundations of trust that built the Internet. The question is whether any of its users or investors care.

June 27 correction: Removes erroneous reference to Axios — the interview in question was conducted for Semafor.