In various industries, different patterns of deception arise during hype cycles. In previous cryptocurrency booms, it was schemes like "ponzinomics" and "rug pulls. " In the self-driving car industry, it was the constant claim that fully autonomous vehicles were "just five years away!" And in the field of artificial intelligence, it seems to be pushing the ethical boundaries of what is acceptable. One company, Perplexity, is currently seeking significant funding to create a competitor to Google Search. However, instead of being a traditional search engine, Perplexity aims to be an "answer engine. " The idea is that instead of sifting through search results to find information from primary sources, Perplexity will provide users with direct answers. According to Perplexity CEO Aravind Srinivas, their focus is on factual accuracy. Essentially, Perplexity acts as a middleman, leveraging high-quality sources while potentially undermining their revenue streams. Traditionally, search engines like Google directed traffic to primary sources such as news outlets. But answer engines like Perplexity divert traffic by providing direct answers, which reduces ad revenue for the primary sources. Similar companies, like Arc Search and even Google itself, operate in a similar manner. Perplexity has taken this concept further with its Pages product, which creates summarized reports based on primary sources. Rather than simply quoting sentences or providing brief answers, Perplexity generates complete articles by actively plagiarizing the sources it uses. Forbes discovered that Perplexity managed to bypass the publication's paywall to provide a summary of an investigation they conducted. Not only did Perplexity evade the paywall, but it also failed to adequately cite the original investigation and used copyrighted images without permission. While aggregation is not new, Perplexity's ability to aggregate on such a large scale, combined with copyright violations, is concerning. In an attempt to address the backlash, Perplexity's chief business officer mentioned potential revenue sharing plans with publications. Nonetheless, Wired confirmed that Perplexity's unauthorized scraping of content extends beyond Forbes.

Perplexity has disregarded the robots. txt code, which explicitly requests web crawlers not to scrape the page. Srinivas responded by claiming Perplexity was not directly ignoring robots. txt, but rather utilizing third-party scrapers that disregarded it. However, he did not disclose the name of the third-party scraper or commit to rectifying the situation. Perplexity further compounded its unethical practices by plagiarizing an article from Wired, even though Wired explicitly blocked Perplexity in its robot. txt file. This raises concerns about the erosion of trust in the good faith agreement that has governed web behavior for decades. Unscrupulous AI companies, not just Perplexity, are indiscriminately collecting data to train their models, jeopardizing the established principles of the web. Moreover, reports indicate that Perplexity is now surfacing AI-generated results and misinformation, contradicting its claimed dedication to factual accuracy. In order to prove the value of Perplexity to investors, CEO Srinivas employed deceptive tactics. He created a tool to scrape Twitter, masquerading as an academic researcher using API access for research purposes. Srinivas boasted about his clever deception during a podcast interview. This revelation raises concerns about the integrity of Perplexity's foundation, as it relies on dishonest practices to circumvent established web principles. In conclusion, the true value proposition of "answer engines" like Perplexity becomes clearer when examining the company's practices. Perplexity cannot generate accurate information independently and instead relies on third parties, even if it means disregarding their policies. The developers behind the answer engine seem comfortable with lying whenever it suits their needs, which is integral to how Perplexity operates.