Perplexity has quickly become a force in the world of search by using artificial intelligence to provide answers to questions rather than simple pages of links, and Perplexity's founder hopes to build on this to challenge Google and become the place to find information online ...

This is something Google has dipped its toe into with its opt-in "search generative experience" experiment, but has been reluctant to rock the boat from such a dominant position.

Just as Google disrupted Yahoo, Lycos, and Infoseek 20 years ago by offering something faster, more efficient, and different, it may be about to suffer the same fate with smaller, more agile competitors.

Perplexity uses a variety of methods to deliver its results, including custom-built large-scale language models, OpenAI's GPT-4, and Google's own data.

As first reported by The Information, Perplexity and Google appear to display results for similar queries in the same order; CEO Aravind Srinivas, an outspoken opponent of Google, has said that AI is the He admitted that he would use Google's ranking signals if he thought they were the best indicator of link quality.

Founded in 2022 by former Google and OpenAI employees, the company has raised over $100 million, including an investment from AI giant Nvidia, which CEO Jensen Huang uses "almost daily" as both a search engine and chatbot He admits to using it "almost daily.

Its interface is closer to ChatGPT than Google, but at its core it is primarily a search engine, providing a list of search results, but presented conversationally with analysis.

The free version offers simple answers, sources per statement, and images. For a more immersive experience, one can pay for a premium plan that includes image input and generation.

Perplexity offers its own base model and a variety of free models, as well as access to a wide range of powerful models for answering questions, including Claude 3-Opus, GPT-4-Turbo, and Mistral-Large with its Pro Plan.

Aravind Srinivas, CEO of Perplexity, previously argued that Google is considered legacy and old, while his company is considered next-generation and future.

Personally, I am a big fan of how Perplexity works. It's intuitive, simple enough for casual use, and complex enough for those who want to get deep into AI. It gives me answers to the questions I am asking, with actual sources.

It's like combining ChatGPT and Google into a new app, but in a way that actually works, services up information, and respects sources.

ChatGPT and Google Gemini can provide links and sources of information, but it is largely hidden or hard to find; Copilot does a better job, but in a way that Perplexity is not secondary to chatbots It is good at putting information online in a way that is not secondary to the chatbot.

While Google's search signals inform the results, Srinivas says that Perplexity does not actually copy the results themselves, but has its own bots that crawl the web and index information just like Google He says.

Perplexity uses Bing's API for ranking signals and appears to use a third party to retrieve Google's data. These rankings determine the importance of a website for a particular term or topic.

According to Srinivas, they take a lot of data to determine their rankings, including not only Google and Bing ranking data, but also how recently they were updated. This is then fed into the AI, where further analysis is performed and a summary is created.

Alone, this is not a problem at all. If the goal is to answer questions and provide information, the more data points, the better. Still, Srinivas' past boasts about the novelty and value of its technology lose some weight when using Google's "legacy technology".

Perplexity's insistence on using Google and Bing in the process of setting its rankings also does not bode well for the company.

According to The Information, OpenAI is building a dedicated search product that, while powered by Bing, will use its own powerful AI to provide analysis, ranking, and summary of results.

This could be integrated into ChatGPT, providing a fast search product to compete with Google and Perplexity. [But Srinivas may not care. Competition doesn't kill companies; companies destroy themselves by not focusing on users and long-term thinking.