• Everyday AI
  • Posts
  • Perplexity Explained - 5 Game-Changing Facts for Your Workflow

Perplexity Explained - 5 Game-Changing Facts for Your Workflow

Musk backs AI California bill, Alexa AI coming in October, Anthropic releases system prompts and more!

Author

Everyday AI
August 27, 2024

Outsmart The Future

Today in Everyday AI
8 minute read

🎙 Daily Podcast Episode: Perplexity can be a great way to grow your business, IF you know how to use it properly. Here’s 5 things you need to know about it. Give it a listen.

🕵️‍♂️ Fresh Finds: Big tech backs AI California bill, Anthropic releases Artifacts to all, Google AI overview struggles and NVIDIA launches international tech. Read on for Fresh Finds.

🗞 Byte Sized Daily AI News: OpenAI demos ‘Strawberry’ to the feds, Elon Musk backs AI safety bill, Alexa AI to launch this October and Anthropic reveals AI system prompts. For that and more, read on for Byte Sized News.

🚀 AI In 5: We’re showing you 5 AI features in Microsoft Copilot! See it here

🧠 Learn & Leveraging AI: If you want to take full advantage of Perplexity’s capabilities, just follow these 5 things. Keep reading for that!

↩️ Don’t miss out: Did you miss our last newsletter? We talked about OpenAI supporting an AI content bill, Inflection's Pi adding usage caps and TikTok adding custom AI voices. Check it here!

Perplexity Explained - 5 Game-Changing Facts for Your Workflow ✋

Wanna know a lil secret?

Perplexity's a cheat code for growing your biz.

Chances are, though, you've got a few things wrong about this AI powerhouse and you're not using it to the fullest.

We dish the real 101 on Perplexity and the 5 things you need to know about it.

Join the conversation and ask Jordan questions on Perplexity here.

Also on the pod today:

• Free vs Paid Perplexity 🤔
• Use Cases of Perplexity 🧠
• Future Implications of Perplexity 💭

It’ll be worth your 57 minutes:

Listen on our site:

Click to listen

Subscribe and listen on your favorite podcast platform

Listen on:

Here’s our favorite AI finds from across the web:

New AI Tool Spotlight – 10xlaunch helps you get more users from the same website traffic, PackPack is an AI bookmark manager and Clockwise Prism is an AI intelligent calendar.

Trending in AI – OpenAI, Adobe and Microsoft all have announced their support for California’s AI bill on watermarks for AI content.

Google – Google’s AI overviews is struggling with its Hindi translation and needs some updates.

AI Model Updates — Anthropic announced that it’s releasing its popular Artifacts feature to all users, including free users.

NVIDIA – NVIDIA has launched NIM Microservices for Generative AI in Japan and Taiwan.

Big Tech - Big tech is protesting the EU’s new AI rules by withholding AI products from Europe.

AI in Society – Data center company Sustainable Metal Cloud explains how it can cut data center energy use by 50%.

1. Elon Musk Advocates for AI Safety Bill in California 📜

Elon Musk is making headlines by calling on California to pass the SB 1047 bill, which would mandate safety testing for AI technologies developed by tech companies.

A long-time proponent of AI regulation, Musk emphasizes the importance of ensuring that these innovations do not pose risks to the public. Simultaneously, Microsoft and OpenAI are backing another bill, AB 3211, which seeks to label AI-generated content to prevent misinformation.

2. Alexa's AI Upgrade Launches This October 🗣

Amazon is preparing to launch a new subscription service for Alexa in mid-October, introducing several innovative features. The upgrades include daily AI-generated news summaries and personalized user preferences designed to enhance user engagement. Additionally, the anticipated Project Metis, a potential competitor to ChatGPT, may debut alongside the subscription.

3. Anthropic Reveals AI System Prompts 🧑‍💻️

Anthropic has made headlines by disclosing the system prompts that govern their Claude AI models, offering an unprecedented look into how these systems operate. These guidelines include strict instructions such as “Claude cannot open URLs” and “always respond completely face blind,” showcasing the company’s commitment to ethical AI development.

This transparency not only sets a new standard for AI vendors but also encourages competitors to follow suit—potentially changing the landscape of AI accountability.

4. Experts Warn of Potential Dangers from Biological AI Models ☣️

Experts are ringing alarm bells on the rapidly advancing biological AI models, saying they could pave the way for engineered pandemics if not properly regulated. A recent paper from notable institutions emphasizes the need for mandatory oversight to mitigate risks associated with these powerful tools, which could accidentally create harmful pathogens.

With a history of biological weapon misuse, the stakes are high, and the authors stress that the time for action is NOW before we unleash something uncontrollable.

5. AI Revolutionizes Cellular Diagnosis 🧬

Researchers at the Centre for Genomic Regulation have introduced AINU, an advanced AI capable of detecting cancer and viral infections at an astonishing nanoscale. This innovative tool employs high-resolution images from a specialized microscopy method called STORM to identify subtle changes in cell nuclei, paving the way for faster and more accurate disease diagnosis.

By recognizing tiny alterations in DNA structures, AINU holds the potential to significantly enhance personalized treatment strategies.

6.OpenAI's Next-Gen AI Named Strawberry Set for Fall Release 🍓

In an exciting development for the AI community, Microsoft-backed OpenAI is gearing up to unveil its latest GPT update, Strawberry, this fall, according to a report from The Information.

This new update promises to tackle complex problems it's never seen before, a feat that current models struggle with, potentially revolutionizing tasks ranging from advanced math to marketing strategy development. Formerly known as Q*, Strawberry aims to enhance the capabilities of existing chatbots like ChatGPT-4o, making high-level research and problem-solving more accessible.

And here’s where things get interesting. The Information reports that OpenAI is actually using the ‘Strawberry’ model to train its next model, called Orion.

A version of Strawberry may be released by fall 2024, possibly within ChatGPT, and it has already been demonstrated to officials for its potential applications. In short, Orion is the main upcoming model, while Strawberry supports its development.

5 Microsoft Copilot AI Features You Must Try!

Microsoft updated its Edge browser and with it came some new and free Copilot features!

We’re showing you 5 AI features in Copilot and giving you use cases examples.

#4 is prolly our fave.

🦾How You Can Leverage:

Do you actually enjoy using the Internet nowadays?

TBH, we kinda hate it. 

Like we chatted about today, the rise of Internet-connected large language models like ChatGPT and Perplexity leads to an ugly downside — the internet starts stinking.

  • There’s more ads.

  • More noise.

  • And the rabbit hole searches of trying to find ONE simple piece of information are resulting in even more browser tabs open across ours screens

  • (Wait, what was I searching for again?)

Enter: Perplexity.

We’re biiiiiiiig fans of the Perplexity, the answers engine that can use different LLMs under the hood.

Today, we decided to do some live walkthroughs and demos. We ran some live Perplexity prompting with our audience (you can go check the results of our work here), went over the free vs. paid plan and more.

Annnnnd, you know we had to leave you with an easy-to-digest list of things you need to know.

So with that?

Here’s the 5 things you need to know about Perplexity. 👇

1 – Perplexity is NOT an LLM

Common misconception here y’all.

Perplexity is not a large language model. Rather, it’s an answers engine. In other words you:

  1. Choose a model:

  1. Shoot a lil query into the search box.

  2. Sit back and watch as Perplexity (literally) browses, reads and summarize like 20 websites to answer you questions:

Try this:
Best part about Perplexity?

The basics are free. Also, you get 5 free “pro” searches every four hours, which is the more powerful version of Perplexity.

If you’re not on the $20/mo plan, you don’t get all of the premium features. But you can also do like Woozy and have 9 different free accounts going at once until your computer goes boom.

2 – Perplexity lets you upload files 🗃️

If you’re looking for answers, you might as well throw your files in there too, right?

(PSA: Never upload proprietary, confidential, sensitive or classified information into LLMs unless the human in charge says it’s gucci.)

Oh, we’re back.

Being able to mix 20ish websites AND your own files at the same time is legit amazing. It’s probably a similar feeling people had when they listened to the radio for the first time.

Like. Whooooooooooa.

Try this:
Perplexity supports PDFs, text files, CSVs and other coding docs. Truth be told, though, Perplexity isn’t winning any awards in its abilities to handle long/big files.

But, if you just need to throw a few short PDFs into your long research mix, it’ll do the job.

(If you need to do some heavy doc lifting, go check out this video to get your AI fill with the free version of Copilot.)

3 – Perplexity is kinda in hot water 🔥

Are you still scratching your head that this is even possible?

So in a few Perplexity prompts, you can skip out on visiting like 100+ websites and just get the good stuff and have it do all your thinking/researching/ad blocking?

(We checked… in our live Perplexity prompting on today’s livestream, we crunched through more than 100+ webpages in a matter of minutes and grabbed only the important goodies.)

Like… what about all those blog post writers, media companies and others that created all the content to begin with?

Perplexity announced it wants to create a revenue-sharing deal with content publishers, but few details have emerged.

But what has emerged……

Try this:

You see how that could lead to some upset people? Some recent reports concluded that Perplexity might have used some deceptive practices to get data from some websites.

Essentially, it’s alleged that Perplexity uses third-party data scrapers to bypass content publishers who are trying to block Perplexity and other LLM makers from grabbing all of their perfectly primed prose.

Rut-roh.

4 – Perplexity is… going social? 🗣️

Is that a social media site?

Many have pointed out a unique feature that Perplexity rolled out earlier this year — its ‘Discover’ section.

It looks kinda like a social media feed. What gives?

Try this:

There’s something more than meets the surface here.

All of those pages we talked about in today’s show that you can make public? Those can end up in the Discover section with very social media-esque vibes.

Which brings us to…..

5 – Perplexity is gonna kill traditional search ☠️

Oh, now #4 makes a lotta sense, right?

Here’s the long story short on how we think Perplexity is using us all to do its dirty work — killing traditional search.

So, you can use Perplexity to (kinda) look at the Top 10 results for a particular topic and combine the best of it all into a Thread. Then, you can publish that Thread as a public page, like this one we did live on today’s show.

Then, those public pages get picked up by Google and (oftentimes) are already taking top rankings on traditional Google search. (Seriously…. go read about it.)

Try this:

Here’s what this could mean — AI search engines like Perplexity are (kinda) ‘borrowing’ from the articles they scrape, helping you rewrite them, and then encouraging you to publish them publicly.

ChatGPT recently released a similar feature.

So, we’ve seen that ChatGPT and other LLMs (Perplexity included) have picked up some major steam recently.

So, as more and more users start to use AI tools like ChatGPT and Perplexity to search, how are those companies hedging their bets?

They’re trying to make sure that whatever you DO search in traditional search engines, that you’ll end up finding AI-published recaps anyways. (Like Perplexity Pages.)

To recap

Perplexity is amazing.

Using an Internet-connected LLM or answers engine like Perplexity will end up saving you hours a week.

So is the takeover inevitable? (Or do you still enjoy having 27 tabs open and STILL searching for the info you need?)

Either the rise of answers engines will crush traditional search and change the way we all consume information online, or we may see so many lawsuits shut the whole thang down.

Either way, we’ll be here. Watching it all. Keeping you updated.

What do you think? Hit us with a reply and lettuce know.

(Or your fave LLM like Claude, Gemini, Copilot, etc)

Now This …

Let us know your thoughts!

Vote to see live results

Do you attend our livestreams?

Every weekday, we bring you fresh AI insights, exclusive interviews, and breaking news with our Everyday AI livestream.

Login or Subscribe to participate in polls.

Reply

or to participate.