What is Perplexity AI?
How will ChatGPT change Search Engines?
In my review of the literature (ahem, clickbait) on how scared Google should be about OpenAI’s limited demo of ChatGPT (GPT-3.5), the most credible ones I found is how Neeva are using LLMs and this Perplexity AI search engine tool.
I think while ChatGPT won’t challenge Google, it does make us think differently about the interface and capabilities of Conversational A.I., fact-checking and truth-verification going forwards in A.I. alignment.
I don’t think ChatGPT is a serious threat to Google or a serious upgrade for Microsoft Bing, anytime soon, but I could be wrong!
Artificial intelligence may change the search landscape in a way that doesn't benefit Google. Although I’m pretty sure Google is motivated to benefit itself with the best R&D around A.I. including whatever GPT-4 allows. I don’t even think OpenAI’s foundational models will predominate in 2023 or perhaps by 2024.
So what is Perplexity AI then? Let’s take a quick look at the Beta.
This project has gotten some serious endorsements thus far:
I find the combination of search engines and LLMs exceedingly interesting, something I noticed Neeva claims to be doing. If I’m to be honest though, I’m not 100% comfortable with the Google roots of Neeva’s founders.
Given how we have been played by privacy-centric Search engines before.
Perplexity A.I. do mention:
This is a demo inspired by OpenAI WebGPT, not a commercial product. We use large language models (OpenAI API) and search engines (Microsoft Bing API). Accuracy is limited by search results and AI capabilities. Do not enter personal information. May generate offensive or dangerous content. Perplexity is not liable for content generated.
One of the main things we need from a ChatGPT type agent is actually showing citations and sourcing.
My guess is Perplexity A.I. is a relatively new project.
But this is the right track.
Easy to compare with Google, etc…
Please check it out and support the project, follow them on Twitter here. Or click on the image below.
Thanks for reading!