Be part of leaders in San Francisco on January 10 for an unique evening of networking, insights, and dialog. Request an invitation right here.
Right this moment, California-based Perplexity AI, the year-old startup trying to reinvent information discovery with an AI-native search engine, introduced it has raised $73.6 million in a sequence B spherical of funding.
The funding comes from IVP with participation from NEA, Databricks, Nvidia, Bessemer Ventures and Kindred Ventures.
Notably, a number of tech visionaries additionally joined the record of backers, together with Amazon founder and former CEO Jeff Bezos, former X VP Elad Gil, GitHub CEO Nat Friedman, Shopify CEO Tobi Lutke and Vercel founder Guillermo Rauch.
Perplexity, which has now raised a complete of $100 million at an estimated valuation of over $500 million, plans to make use of the capital to construct out its AI-driven search engine providing and tackle heavyweights Google and Microsoft within the age of generative AI and huge language fashions (LLM).
What precisely does Perplexity AI have on provide?
Whereas Google and Microsoft have been within the search engine house for years — with the previous largely dominating the sport, Perplexity believes it could take them on (or possibly even dethrone them) by delivering a greater, AI-native expertise.
The startup gives a information search platform that mixes net indexes with a spread of AI fashions to offer solutions to each query a person might have. The expertise is solely conversational, the place there’s simply the knowledge required and no have to click on on completely different hyperlinks (together with adverts/sponsored posts), examine solutions or endlessly dig for info.
To make the expertise even higher, Perplexity additionally offers customers the flexibility to delve into the main points of the knowledge supplied by browsing essentially the most related sources and citations together with the outcomes.
“In an period the place misinformation and AI hallucinations are inflicting growing concern, we’re constructed on the concept accuracy and transparency are stipulations to creating AI-powered search ubiquitous. The occasions of sifting by web optimization spam, sponsored hyperlinks, and a number of net pages shall be changed by a way more environment friendly solution to devour and share info, propelling our society into a brand new period of accelerated studying and analysis,” Aravind Srinivas, co-founder and CEO of Perplexity, wrote in a blog post.
The corporate gives a free model of the platform, the place one can seek for the knowledge they want utilizing the default mannequin of the corporate. Nonetheless, if there’s the necessity to do extra, there’s additionally a Professional model, billed at $20/month, that brings the flexibility to decide on the mannequin of alternative, together with GPT-4, for higher outcomes.
The paid tier additionally offers entry to Perplexity API credit to combine cutting-edge open-source LLMs into tasks in addition to limitless querying help for Perplexity Copilot, the interactive analysis assistant of the corporate.
“Copilot tailors search queries with customized follow-up questions, introducing the idea of generative person interfaces. It removes the burden of immediate engineering and doesn’t require customers to ask completely phrased inquiries to get the solutions they search. This allows customers to realize extra related and complete solutions than different AI chatbots, conventional search engines like google or analysis instruments,” Srinivas added.
Vital development however challenges stay
Perplexity launched its search engine a yr in the past and already claims to have gained important traction with 10 million month-to-month energetic customers and extra tha 1,000,000 cell customers.
The CEO famous that the platform served over half a billion queries in 2023, with its Copilot getting explicit traction from lecturers, college students, and information staff who depend on frequent analysis for his or her day-to-day wants.
Despite the fact that it is a nice begin for the corporate, it’s crucial to notice that it’s nonetheless far off from its core mission “to serve all the planet’s unbounded curiosity”.
Google and Microsoft each have billions of customers and are already shifting to loop in generative AI into their search engines like google. Google Search is testing its all-new Gemini mannequin, which has diminished the latency of responses by 40% for English in the U.S., whereas Microsoft Bing has tapped OpenAI’s GPT-4 for info search and DALL-E 3 for picture era. Each gamers, nevertheless, proceed to floor hyperlinks and adverts of their outcomes — which isn’t the case with Perplexity.
It will likely be fascinating to see how the corporate builds on its efforts – and distinctive options like mannequin flexibility – to tackle these behemoths and scale to allow information discovery within the age of AI.