Read Time: 6 minutes │ Subscribe to the FREE Newsletter │Join the Premium Patch
Hey y’all!
It’s been a pretty rainy week in the Bay Area, but the weather was stunning today. Hope you have a weekend of clear blue skies ahead!
Not AI-generated! Promise.
Today, you’ll discover:
An AI-powered search engine that has all the answers
A dive into the AI behind our favorite streaming services
A ChatGPT prompt to help scan ingredient labels!
🌿 Let’s grow!
This week’s letter is brought to you by The Rundown AI!
How do you stay up-to-date with the insane pace of AI? Join The Rundown – the world’s fastest-growing AI newsletter with over 500,000+ readers learning how to become more productive using AI every morning.
1. Our team spends all day researching and talking with industry experts.
2. We send you daily updates on the latest AI news, tools, and tutorials.
3. You learn how to become 2x more productive by leveraging AI.
Apple Sauce: the ‘sauce’ - or overview - on AI tools you may find useful.
The Search Sweetheart of Silicon Valley
Today’s apple: Perplexity AI
As we have established, DALL·E isn’t the best speller. AI-generated with DALL·E - Prompt: generate an image of a candy heart that says "Will you be my search engine?"
I learned about Perplexity AI through lots of fanfare in the tech community. And after trying out this AI-powered conversational search engine for the last few weeks, I can see why!
Googling versus Perplexing
Perplexity aims to challenge Google by delivering a better search experience. What makes Perplexity different than the search engine that has been our unwavering companion over the last two decades?
When we conduct a Google search, we are presented with a list of web pages ranked by relevance. However, finding the answer we need often requires sifting through the results.
Perplexity aims to provide the answer. The app, which is free to use, is powered by GPT 3.5, the same AI model used by the free version of ChatGPT. There are a lot of risks when we rely on AI to provide us with accurate information. Perplexity aims to combat these risks by validating information across multiple sources and citing those sources when providing answers.
I have found Perplexity to be an efficient tool for conducting research or seeking new information. There’s only one thing I’m perplexed about: why is it called Perplexity when it provides us with answers?!
Anywho, here are some examples of the differences between a Google search and a Perplexity search:
Ammi seeks a recipe for the Egyptian dish, Hawawshi:
Google gives us very familiar results: top recipes and follow-up questions.
Perplexity gives us an answer: an explanation of how to make Hawawshi with a full recipe and cited sources.
Ammi needs to restock arts & crafts supplies
Google provides a useful visual list and a link to a perfectly relevant blog.
Perplexity offers a list with explanations and surfaces sources across the top.
Gone are the days of Googling?
Not quite. Although Perplexity has quickly become one of my most used apps, I opt for a Google search for simple questions or when it is convenient, for instance, when I am browsing on Chrome and can search directly from the address bar.
I opt for Perplexity when the query is more complex, such as comparing different things or trying to learn a new skill or concept.
My preference is also based on mood. For example, with recipes: do I want to explore and discover a recipe among many different options? Or am I in a time crunch and just want a basic recipe I can follow?
Try it out, I’m curious to see what you think!
Planting the Seed: Explore AI topics and headlines in simple language.
The Power of Personalization: Netflix’s Recommendations
A few weeks ago, we began a series exploring the AI in our everyday apps. If you’d like to catch up on what’s up with AI in WhatsApp, check it out here!
Let’s explore another big slice of our digital lives: streaming services like Netflix, Spotify, and YouTube. We enjoy these services because we don't have to comb through thousands of titles to find what we like. Instead, these apps suggest what we might like based on sophisticated algorithms that learn from how we interact with the app and its content.
Source: MobileSyrup
According to the folks at Netflix, 80% of what people watch on the platform is driven by their recommendation engine. And this recommendation engine is powered by - you guessed it - artificial intelligence.
While each streaming service has its own special recommendation engine, all recommendation engines have some similar components. Recommendation engines analyze vast amounts of data to predict what users might like to watch, listen to, or enjoy next. This process involves two primary techniques: collaborative filtering and content-based filtering, often used together for more accurate recommendations.
Source: Towards Data Science
Collaborative Filtering: This method relies on the wisdom of the crowd. It looks at the behaviors and preferences of users and finds patterns or similarities among them. If User A has a similar watching history to User B, then the movies liked by User B are recommended to User A, assuming they'll likely enjoy them too. This approach can unearth surprising and diverse recommendations beyond a user's typical preferences.
Content-Based Filtering: This technique focuses on the attributes of the items themselves. For example, if you frequently watch thrillers on a streaming platform, the system will recommend other movies classified as thrillers. This method relies heavily on content metadata, such as genre, director, cast, and even specific themes or moods. This is also why my Netflix recommendations are currently inundated with claymation farm animals.
Many other complex elements make up the Netflix recommendation engine. We don’t have the time and space to get into it all, but if you’re interested, check out this two-part Netflix blog post where they describe this technology in great detail!
Bonus fact: did you know the thumbnails, or cover images, of each title are also personalized based on the user? Which one would you click?
Fruitful Harvest: Fresh AI fruits from the community garden.
Last week, I shared reflections on the 3 amazing Ammis featured in the book and film, Hidden Figures. I also shared a 4-step guide to Getting Comfortable Outside Your Comfort Zone. Read here.
Later today, we’re having our first-ever Zoom meetup!
Next week, we’re excited to share our interview with Dr. Jamila Yamani, an amazing Ammi who shares her fresh perspectives on AI + the Climate. We’ll also share our monthly favorites across tech, productivity, and Ammi-ing.
All with an ad-free guarantee 🙂 Don’t miss out!
Community Garden
This week, an Ammi & friend, Zainab from Dallas, shared a useful ChatGPT tidbit: using AI to help scan ingredient lists!
If you or a family member have dietary preferences or restrictions, such as vegan, gluten-free, Kosher, or Halal, you may spend a lot of time carefully inspecting ingredient labels.
Copy and paste this prompt in ChatGPT to get some AI assistance!
You have expertise in consumer packaged goods. You are an experienced scientist knowledgeable about ingredient chemistry and its derivation.
Given a list of ingredients and a user's preferences or restrictions, guide the user on whether they can consume or utilize a product or not.
Ask the user to share ingredients via an image or by pasting text. Then ask the user to list any preferences or restrictions.
Provide your guidance, and support it with a clear and concise explanation. Always remind the user to consult with a professional if they are still unsure about a product or a particular ingredient. Bonus tip: if you have ChatGPT Plus, you can simply upload a picture of an ingredient list or nutrition label in response to the above prompt.
If you are using the free version of ChatGPT, use the Google Lens tool on the free Google Photos app to select and copy the text from the nutrition label image (tutorial here). Then paste the text into the free version of ChatGPT in response to the above prompt.
Thanks, Zainab, for the great tip!
Cultivating Conversation: your insights nourish the ammi.ai garden.
We’re workshopping some alternatives for our referral gift! We want to gift something you’d find immediately useful.
Which of these would be most useful to you right now?
Thanks for spending a few of your precious, precious minutes with us.
How did you like today's newsletter?
If you’ve been enjoying ammi.ai, help us grow the ammi.ai Heart Garden!
See y’all in 2 weeks!
Ruqaiya
Ammi by day, Ammi by night
