Perplexity AI
Perplexity AI is a chat-based search engine that uses large language models (LLMs) to answer questions. It provides answers and cites sources from the web. The company behind it, Perplexity AI, Inc., is based in San Francisco, California.
 | |
| Industry | Artificial intelligence |
|---|---|
| Genre | Search engine |
| Founded | August 2022 |
| Founder(s)s |
|
| Headquarters | , US |
Key people | Aravind Srinivas (CEO) |
| Services |
|
Number of employees | 100 [1][2] (2024) |
| Website | perplexity.ai |
HistoryEdit
Perplexity was set up in 2022 by four engineers: Aravind Srinivas, Andy Konwinski, Denis Yarats, and Johnny Ho. Here's a bit about each of them:
- Aravind Srinivas, the CEO, used to work as an AI researcher at OpenAI.
- Konwinski, part of the founding team at Databricks, brings expertise in back-end systems.
- Yarats, the CTO, was an AI research scientist at Meta.
- Ho, the CSO, has worked as an engineer at Quora and later as a quantitative trader on Wall Street.[3]
ServiceEdit
Perplexity offers its services through a freemium model and also has an enterprise version.
free planEdit
Perplexity uses a standalone language model based on GPT-3.5 that includes browsing capabilities. This model uses the context of user queries to provide personalized search results. Perplexity not only summarises the search results but also includes inline citations in the generated text. Additionally, users can use the Pages feature to create customizable webpages and research presentations based on their prompts.
ProEdit
Perplexity Pro offers several advanced features. It provides API access, allowing users to search both internal files and web content. The service asks clarifying questions to refine user queries and enhances accuracy. Users can upload and analyze local files, including images. Perplexity Pro uses powerful language models like GPT-4, Claude 3.5, Grok-2, Llama 3, and in-house Perplexity LLMs. Additionally, it can generate images using AI through platforms like Playground v3, DALL-E, Stable Diffusion, and FLUX 1.
ControversiesEdit
ForbesEdit
In June 2024, Forbes publicly criticized Perplexity for using their content. Forbes claimed that Perplexity published a story that was largely copied from a proprietary Forbes article without properly citing or mentioning the source. In response, Srinivas acknowledged that the feature had some "rough edges" and accepted the feedback. However, he maintained that Perplexity only "aggregates" information and does not plagiarise.[4][5]
WIREDEdit
In June 2024, separate investigations by the magazine Wired and web developer Robb Knight revealed that Perplexity does not respect the robots.txt standard, which is used by websites to prevent web crawlers from scraping content. This was reportedly despite Perplexity claiming otherwise. Perplexity publicly lists the IP address ranges and user agent strings of their web crawlers, but according to Wired and Robb Knight, they use undisclosed IP addresses and spoofed user agent strings when they ignore the robots.txt rules.[6][7]
Wired also stated that, in some cases, Perplexity may be summarizing:
"not actual news articles but reconstructions of what they say based on URLs and traces of them left in search engines like extracts and metadata, offering summaries purporting to be based on direct access to the relevant text."[6]
In response, Srinivas stated in a phone interview that:
"Perplexity is not ignoring the Robot Exclusions Protocol... We don't just rely on our own web crawlers, we rely on third-party web crawlers as well."
Srinivas explained that the web crawler identified by Wired was owned by a third-party provider.[8]
When asked if Perplexity would stop scraping content from WIRED using third parties, Srinivas replied that it's complicated.
AmazonEdit
Amazon Web Services, which hosts Perplexity's crawler, has a terms of service clause that prohibits users from ignoring the robots.txt standard. Amazon has started a "routine" investigation into Perplexity's usage of Amazon Elastic Compute Cloud.[9]
Legal casesEdit
In October 2024, The New York Times (NYT) issued a cease-and-desist notice to Perplexity, requesting the company to stop accessing and using NYT content. The NYT alleged that Perplexity is violating its copyright by scraping data from its website.[10] Additionally, the NYT has filed lawsuits against OpenAI and Microsoft, claiming copyright infringement for using millions of its articles to train the large language models that power ChatGPT.[11]
The cease-and-desist notice sent by NYT lawyers read in part:
"Perplexity and its business partners have been unjustly enriched by using, without authorization, The Times's expressive, carefully written and researched, and edited journalism without a license."[12]
Perplexity plans to respond to the notice by October 30, 2024.[10]
In the same month, Dow Jones and the New York Post filed a lawsuit against Perplexity, accusing the company of copyright infringement. The lawsuit also claims that Perplexity attributed quotes to an article about F-16 jets for Ukraine that were not present in the original article.[13]
ReferenceEdit
- ↑ https://www.linkedin.com/pulse/linkedin-top-startups-2024-50-us-companies-rise-linkedin-news-hxote/
- ↑ Cite error: Invalid
<ref>tag; no text was provided for refs named:1 - ↑ "AI-powered search engine Perplexity AI lands $26M, launches iOS app". TechCrunch. April 4, 2023. Archived from the original on March 5, 2024. Retrieved May 5, 2024.
- ↑ O'Brien, Matt (June 15, 2024). "AI startup Perplexity wants to upend search business. News outlet Forbes says it's ripping them off". Associated Press. Archived from the original on June 20, 2024. Retrieved June 20, 2024.
- ↑ Lane, Randall (June 11, 2024). "Why Perplexity's Cynical Theft Represents Everything That Could Go Wrong With AI". Forbes. Retrieved June 20, 2024.
- ↑ 6.0 6.1 Mehrotra, Dhruv; Marchman, Tim (June 19, 2024). "Perplexity Is a Bullshit Machine". Wired. Archived from the original on June 20, 2024. Retrieved June 20, 2024.
- ↑ "Perplexity AI Is Lying about Their User Agent". Robb Knight. June 15, 2024. Archived from the original on June 20, 2024. Retrieved June 20, 2024.
- ↑ Sullivan, Mark (June 21, 2024). "Perplexity CEO Aravind Srinivas responds to plagiarism and infringement accusations". Fast Company. Retrieved June 24, 2024.
- ↑ Mehrotra, Dhruv; Couts, Andrew (June 27, 2024). "Amazon Is Investigating Perplexity Over Claims of Scraping Abuse". Wired. Retrieved July 3, 2024.
- ↑ 10.0 10.1 Davis, Wes (October 15, 2024). "The New York Times warns AI search engine Perplexity to stop using its content". The Verge. Retrieved October 17, 2024.
- ↑ Complaint, New York Times, Co. v. Microsoft Corp., No. 1:23-cv-11195 (S.D.N.Y. December 27, 2023).
- ↑ Bruell, Alexandra (October 15, 2024). "New York Times to Bezos-Backed AI Startup: Stop Using Our Stuff". The Wall Street Journal. Retrieved October 17, 2024.
- ↑ Bruell, Alexandra (October 21, 2024). "Wall Street Journal, New York Post Sue AI Startup Perplexity, Alleging 'Massive Freeriding'". The Wall Street Journal. Retrieved October 21, 2024.