💬 Meta’s AI App: The Digital Assistant That Gets You

After integrating its AI into all the usual suspects (WhatsApp, Instagram, Facebook, Messenger), Meta is now stepping it up by launching a standalone AI app. 

Yup, you heard that right. This new app, which made its debut at Meta’s LlamaCon event, lets you tap into their AI in a way similar to ChatGPT, but with a twist.

Here’s what makes Meta’s AI app a little different from the usual:

• It Knows You (In a Totally Non-Creepy Way): Since Meta knows everything about you (like, seriously, everything—thanks to your socials), this app pulls from all the juicy data you’ve willingly shared over the years—we’re talking your likes, your friends, your posts—giving you AI responses that feel more you. The goal? To make you feel like you’re chatting with a friend who already gets you.

• It Remembers Your Preferences: Not only does it know you well, but you can also teach it to remember things for future chats. For example, you could tell it you’re lactose intolerant, and next time, it won’t try to send you to a wine-and-cheese party. Thanks, Meta!

• Available in the U.S. & Canada: For now, these personalized responses is live in these countries, but if you’re outside these regions, don’t worry. It’s probably rolling out globally soon.

And as expected with anything Meta does, there’s a bit of a social twist:

• Introducing the “Discover Feed”: It’s like Meta’s version of Instagram Stories, but for your AI convos. Here, you can share how you’re using AI with your friends—basically, letting everyone in your circle see how you interact with the AI. But hey, you control what gets shared, so no worries if you want to keep your weird AI chats private.

A Quick Word of Caution: We all know Meta’s bread and butter is targeted advertising. So, while the AI app’s pretty nifty, you might want to think twice about what info you’re feeding it.

At the end of the day, your data helps Meta serve up those oh-so-targeted ads, which keeps their business running $$$ smoothly.

And while some might roll their eyes at the idea of a social feed in an AI app (because, honestly, do we really need a “social feed” for everything?), it’s a thing. The question is: Will it be a hit? Only time will tell.

Long story short:

Meta’s new AI app could become your digital sidekick, but just remember—it’s got a bit of your life’s data already tucked away. Use with caution, but enjoy the ride!

Get more info here!

  😲 OpenAI Rolls Back a ChatGPT Update

You know how ChatGPT is usually that overly chill friend who just agrees with whatever you say?

Well, after the latest GPT-4o update, it basically went full people-pleaser mode — and users were not vibing with it.

In fact, it wasn’t just weird — it was straight-up unsettling. Over the weekend, the internet was flooded with screenshots showing just how far ChatGPT’s new agreeable streak had gone.

Here’s what went down:

• ChatGPT started agreeing with literally everything — even when it really, really shouldn’t.

• One person hit it with the classic trolley problem (you know, saving lives vs. saving a toaster)... and the bot said it was totally valid to choose the toaster over cows and cats.

• The most shocking part? It applauded a user for stopping their very important medications. Yikes.

Even Sam Altman (aka OpenAI’s CEO) stepped in on X and admitted the update made ChatGPT “too sycophant-y and annoying.” His words, not ours.

But hey — to be fair — that overly agreeable ChatGPT was kind of fun to interact with. A part of me lowkey doesn’t want it gone.

But here’s the thing: that kind of nonstop validation can come with some serious consequences.

Because when an AI starts backing everything you say — from wild delusions to actual mental health spirals?

That’s not cute. That’s dangerous.

So what’s OpenAI doing about it?

• They’ve fully rolled back the GPT-4o update — for everyone, both free and paid users.

• Basically we’re going back to a less clingy, more balanced version of ChatGPT while they work on the personality tweaks.

• According to Altman, more updates and fixes are coming “in the next few days.”

Let’s just hope these new tweaks do more good than harm.

So yeah, stay tuned and stay alert — and maybe don’t take ChatGPT’s word for it just yet… at least not until OpenAI finds the right balance between a “supportive chatbot” and a robot that thinks your toaster is a national treasure.

Here’s more on the updates and fixes OpenAI is working on.

🧱Around The AI Block

• 🛑 How to stop the AI you’re using from training with your data.

• 🛠️ 9 underrated AI tools you’ll wish you found sooner

• 📺 How to watch LlamaCon, Meta’s first AI developer event.

• 🗣️ Yelp debuts AI voice agents for restaurants and service providers.

• 🤩 Meta previews an API for its Llama AI models.

• 🦾 Duolingo is going to be the first ‘AI-first' company.

• 🎙️ Google’s NotebookLM expands its AI podcast feature to more languages.

🛠️ Trending Tools

• Stager AI enhances your real estate photos with virtual staging, image editing, and automated improvements to sell properties faster.

• TrailMate is a companion for outdoor adventures that provides guidance and information without the internet.

• Buzzabout analyzes social media conversations to help you understand what your customers think about your product or brand.

• Lenish helps you track your dog's daily routine, including toilet breaks, feeding times, and exercise.

• Agree.com lets you send, sign, and manage agreements, streamlining the entire contract process from creation to payment collection.

🤖 ChatGPT Prompt Of The Day:

Prompt to Fix a Bug or Design a Solution Logically

To execute this prompt perfectly you need to:

• Provide the code and what it’s supposed to do.

• Clearly define what’s expected vs. what’s happening

• Ask the model to explain the logic step-by-step as if it were debugging.

• Use “Walk me through” to activate reasoning (also don’t skip the fix)

• If needed, follow up with “Could there be another possible issue?”

Use the prompt below 👇

Here’s what my code is doing: [Insert brief description or snippet]. Can you walk through the logic and help me spot the issue?”

OR

Help me debug this Python function. Think step by step: first explain what the code is supposed to do, then identify the issue, and finally suggest a fix.”

See this as an AI WoD!

Premium members get daily video prompts, premium newsletter, an no-ad experience - and more!

Already a paying subscriber? Sign In.

Premium members get::

• • 👨🏻‍🏫 A 30% discount on the AI Education Library (a $600 value - and counting!)
• • 📽️ Get the daily AI WoD (a $29.99 value!)
• • ✅ Priority help with AI Troubleshooter
• • ✅ Thursday premium newsletter
• • ✅ No ad experience
• • ✅ and more....