Editor’s note: Hard landings, soft landings, and no landings. Throughout it all, you sought information to help give your clients more effective advice. These Members’ Choice runner-up articles show your continued interest in how to help your clients save money on taxes and to position their retirement savings for the years ahead. These just missed being in our annual Top 10 list, but were still among the most popular articles of 2024.
Author’s note: Since early 2023, I’ve been researching “generative artificial intelligence.” This form of AI technology is capable of quickly creating new text, images, audio, video, computer code and virtual environments. My objective for AI for Advisors is simple. I want to show you and your team how you can use generative AI in your daily, day-to-day work, and personal life, to save time, boost efficiency, and amplify your intelligence. This will include AI tech stack reviews, how-to’s, case studies, and deep dive demos into this exploding new technology frontier.
Interested in AI? Join our upcoming book discussion “Unlocking Co-Intelligence: A Two-Part Book Discussion for Financial Advisors.”
The way you search for information on the Internet will change dramatically by the time you’ve finished this article.
You’ve probably noticed over the years how Google search has become cluttered and harder to use. In its early days, it was a page of blue “organic” links to useful websites and one or two unobtrusive paid ads.
Now a simple search might give you a page with a dozen or more “sponsored links,” a list of traditional “PageRank” links, and a variety of other search results that might contain the information you want to read.
The problem
When doing serious research on Google, your results page is just the start. You’re still burdened with clicking each link, quickly scanning for relevant information, and deciding whether to invest time reading, copying key snippets or printing for later reference.
And then you hit the back button, and repeat the process. It’s not efficient. It’s not fun. And critically, you can burn up a lot of brainpower looking for what you need.
Introducing Perplexity.ai
There’s a new class of AI-driven search engines that seek to solve the Google search problem. Perplexity.ai is the head of the pack and calls itself the “world’s first conversational search engine” and direct answer machine. It aims to connect you with the information you want in an easy way. You’ll see why after just a single search.
Click here to try Perplexity. Enter this search prompt: “Tell me about [name of your firm].” Let us know in the comments below what happened.
What you’ll observe is that unlike Google search, Perplexity provides a single answer to your search query. This is its key differentiator. Plus, its answers provide hyperlinked footnotes to the sources it drew its answers from. Importantly, Perplexity also will tell you when it can’t find good information. And there’s more. It also asks you follow-on questions that often reveal new aspects of the topic you’re researching. Like they say, it’s a conversational answer.
Like other AI tools, Perplexity can make mistakes. So, you need to scrutinize Perplexity’s responses, watching for possible errors.
What are the underlying differences?
Google uses complex search algorithms, including PageRank, to analyze and rank web pages based on their popularity and relevance to your search query. Its search algorithm considers hundreds of factors, including keywords, site usability and backlinks.
Perplexity.ai, on the other hand, leverages generative AI to understand and respond to your queries in a conversational manner. It provides direct answers by summarizing information from various web sources, rather than just listing web pages that might contain the answers.
Here’s an overview of Perplexity’s features:
- Ease of use: Like Google, Perplexity is accessible with a straightforward interface, where users can input their queries and receive responses directly on the website.
- Fresh information: A significant advantage of Perplexity AI versus other AI models (ChatGPT, Claude, etc.) is its ability to provide up-to-date information along with relevant links and citations, even in its free version. Many AI chatbot models are constrained by their knowledge cutoff dates, lacking the real-time information access of search engines.
- Custom, focused search: Perplexity allows you to micro-target searches across various domains like Wikipedia, Reddit, YouTube, Wolfram|Alpha, and academic papers and journals. This allows you to exclusively research specific interest areas.
- Search thread library: Perplexity saves your search threads in a library for easy access and reference. Additionally, you and others interact with shared threads by asking follow-up questions or exploring related questions, enhancing the collaborative group search aspect of information discovery.
- Paid features and customization: With a subscription of $20 per month for Perplexity Pro, users gain the ability to customize their search experience by selecting specific sources and editing the AI’s responses for more personalized and accurate results. This feature introduces a collaborative aspect similar to Wikipedia. Additionally, subscribers can switch between various Large Language Models (LLMs) including ChatGPT-4, Claude-3 and Mistral, tailoring the AI’s functionality to their preferences.
- Copilot feature: The Copilot function, leveraging GPT-4, enriches searches by posing follow-up questions to better grasp your inquiries and refine the accuracy of its responses.
- In your browser, on your phone: Perplexity isn’t just a website. It can be added to your Chrome browser and on your mobile phone. Additionally, the mobile app brings the convenience of voice search and other features, although it currently does not support account syncing with the web app.
Google isn’t sleeping
Personally, I’m now using Perplexity for more than 90% of my daily searching. (I’m sticking with Google maps for navigation.) And while Perplexity is making waves, it’s important to note that Google isn’t idly sitting by.
The search giant has begun incorporating generative AI capabilities into its traditional search interface. Users may now see a “generate” button on the search results page, which triggers an AI-powered search on the given term and produces a concise answer. However, Google’s current implementation doesn’t include cited sources, limiting its usefulness for in-depth research. Nevertheless, Google’s AI model Gemini is a major generative AI player and Google clearly recognizes how its money-making search business is threatened by generative AI.
Stay tuned.