Simplifying Tasks with AI: Three Practical Uses and a Bonus Tip for Everyday Tasks

Searching the web in 2024, AI vs Google Search Engine

Until recently, Google dominated web searches. However, with the advent of ChatGPT 3.5, this landscape shifted dramatically, prompting Google’s ‘Code Red’ and Microsoft’s integration of GPT into Bing. But how exactly does ChatGPT challenge traditional search engines?

Traditionally, a Google search involved navigating to the site, entering key terms, scanning through results, and reading articles to find the needed information. ChatGPT revolutionises this process. You simply ask your question in natural language and receive a direct response, with the added advantage of continuing the conversation for more in-depth exploration.

Enter Perplexity AI

One of the challenges with AI models today is their occasional inaccurate, sometimes resulting in the generation of incorrect facts, a phenomenon known as “hallucination”.

This is where Perplexity AI steps in. Perplexity aims to combine the usefulness of search engines when it comes to research, with the power of language models.

It works very similarly to Google, where the user can enter the search terms, which brings up a list of “sources”, or results.

However, Perplexity goes one step further. Perplexity actually understands the sources, and provides an answer, directly, in the form of a paragraph.

The user interface (UI) is beautifully designed, making interactions easy and intuitive.

If you need to, you can continue to ask follow-up questions, which it interprets and answers.

Note that even though Perplexity references sources, its interpretation in the answer might be wrong. For any critical information, it is always suggested to navigate to the sources to confirm the information.

Can’t ChatGPT also do this?

ChatGPT Plus can and often does browse the web for answers, although it may not always do so automatically. Sometimes, it will quickly answer from its own internal knowledge, getting the answer wrong. By prompting it to conduct a web search, ChatGPT then retrieves and references data directly from the internet, akin to Perplexity’s approach. To make sure it gets the most recent information, and reference sources, you can simply directly ask it to browse the web.

So, what’s the point? Why would you use these search engines instead of simply using Google?

The benefit here is that you can talk to the results. It’s basically like having an assistant, who has read all of the results, and you can now ask them for summaries, key take aways or follow-up questions, without having to read all of the content yourself, at virtually no cost, and in real-time.

The important consideration to keep in mind is that AI models like ChatGPT are prone to ‘hallucination’, making it crucial not to accept their responses without question. Certain prompting techniques can enhance ChatGPT’s accuracy. Perplexity, addresses this issue directly by providing sourced references with its answers. This is at the core of Perplexity, and is therefore currently free of charge, whereas ChatGPT only offers this to paid users.

Bonus: Learning anything complicated, simplified with AI.

There are sometimes interesting concepts that we feel are out of our depth. Well, with AI models, like ChatGPT, you can ask as many questions as you want, and also ask the model to provide an answer in a specific format. Whatever works best for you.

The easiest way I find is to ask it to explain a concept, like I’m a 5-year-old. Such a prompt results in a very simple answer, with metaphors and examples, make it very easy to digest.

Here’s an example with an explanation of a “jet engine”.

You can also continue to prompt it, to understand different aspects. It’s just like having a conversation with a teacher.