Perplexity AI: With a valuation of over US$1 billion, Huang Jen-Hsun also loves to use AI tools! Claiming to be an “answer” engine, how can you use AI to challenge Google Search’s status?

Perplexity's answer method will attach the most relevant reference sources at the beginning, saving time on manual searches and judging the relevance of the website (the top few rankings on Google searches are usually advertisements)

Attached below is a summary compiled by AI, so that you can see a summary of multiple data sources at the same time, but you won't know where the information comes from!

Small advantages compared to traditional search engines:

1. If the question is not clear enough, it will proactively ask you back: Sometimes we cannot find a satisfactory answer when searching on Google because the question we ask is not good enough or even the wrong question. At this time, Perplexity can help and improve the quality of the question.
2. Search with pictures and files to automatically generate images: Not only can you search using text, but you can also directly upload pictures and files to Perplexity to assist your questions. If you need picture output, Perplexity will also provide corresponding pictures.
3. Systematize every "search" you do: Search records on Perplexity can be organized into "search strings" (Threads). When you search for similar concepts in the future, you can review past search records and share them at the same time. Give your Search History to others.

Summary: The biggest difference with Google

Compared with manually searching and integrating several website sources to filter out the final information, Perplexity's advantage is to use AI to directly provide answers to users.
Through the user's spoken questions, Perplexity will directly reply with a summary and attach reference materials (mainly websites and literature). Srinivas said: "No one needs these 10 blue hyperlinks if they can directly answer the other party's question."

Perplexity breaks away from traditional keyword-based search algorithms and uses machine learning to conduct in-depth research and clarify the real confusion and context behind user searches, providing more accurate results and making the search process more intuitive and user-friendly.

Perplexity AI Disadvantages: Love of brainstorming & copyright disputes

Perplexity was only established in August 2022. Although during its initial launch period, it successfully accumulated more than 50 million users through word-of-mouth promotion of X alone, there are still many potential problems that need to be overcome.

I love imagining, and the generated information is not entirely true:

Like many large-scale language models, Perplexity will still pop up some fabricated information from time to time. Although reference sources are attached above, many users have reported that some of the content is still untrue and requires manual Double Check.

Answering content copyright disputes:

During the training process, Perplexity's model will crawl billions of examples on the Internet as training data.
The New York Times has said this could affect traffic to articles and ads; well-known report The Verge It also said that this would cross the line of paid content and harm the rights of the original author. If Perplexity provides a reply but does not fully provide the source of the information, or quotes but partially adds jealousy or distortion, it is an act that harms the original author. .

Perplexity AI Business Model

Facing super rival Google, how does Perplexity plan to make money?

Perplexity CEO Srinivas said they have no plans to replace Google,
Instead, they targeted the top 10% users with the highest income as their customer base and sold the subscription version of Perplexity Pro.
.
With a monthly subscription of US$20 for Professional Edition users, you can have unlimited use of the Perplexity Copilot function, upload your own information and data, and have it analyzed by Perplexity AI.
Although it targets a niche market, Perplexity has currently accumulated about 100,000 paying users.In the future, it will also target enterprise-level users and launch its own Perplexity search engine within the enterprise.

Investors and future development behind Perplexity AI

NVIDIA CEO Jensen Huang once mentioned that Perplexity is one of his most commonly used AI tools.
NVIDIA also participated in the financing of Perplexity AI earlier this year!

 Now Perplexity has successfully gained the attention of VC IVP (Institutional Venture Partners), and has also received support from technology executives such as NVIDIA and Amazon founder Bezos in the latest round of financing. It has raised up to US$74 million in recent years. The most outstanding performance among the new startups focusing on online search.
Perplexity's fundraising performance also shows that technology giants believe that AI will subvert the way people search for information in the future.

3 Takeaways

1. The rise of Perplexity AI

The founding purpose of Perplexity AI is to provide users with more comprehensive search content through multiple presentation methods. Compared with Google's search engine, Perplexity AI hopes to provide an "answer" engine, providing users with not only website links, but a complete answer. You can think of the operating principle of Perplexity AI as a combination of Google + ChatGPT, providing answers by understanding user needs and converting them into understandable text.

2. Use AI to create differentiated and unique functions

Perplexity AI provides two search methods, Quick Search and Pro Search, which provide quick, short answers and detailed answers similar to essays respectively. This multi-level answer method allows users to obtain information more deeply. Perplexity AI allows users to search for information just like interacting with a ChatBot. The dialog box will keep a record of the conversation and attach corresponding data links to each answer, saving time on manual searches and determining the relevance of the website.

3. False information and copyright challenges

Although Perplexity AI has successfully accumulated more than 50 million users in a short period of time and has received investment from technology giants such as NVIDIA and Bezos, it still faces problems such as incomplete information generated and copyright disputes on the content it answers.
Perplexity AI will expand its business model through subscription and enterprise users, and continue to optimize its AI models to meet future challenges and disrupt the way people search for information.