FIRST OF ALL – INTRODUCTION

On Sunday, we gave you some “Chin-Chin”—a quick appetizer to tell you that Google had launched something new called Nano Banana Pro (powered by the Gemini 3 engine). We promised you that once we had time to really press the buttons and stress-test the engine, we would bring you the main course.

Well, the food is ready.

To call Nano Banana Pro just “another image generator” is like calling the Third Mainland Bridge in Lagos “just a road.” While technically true, it misses the magnitude of what is happening. Until now, AI tools like Midjourney or Stable Diffusion could be likened to talented artists who were also a bit drunk. Yes, they are creative, but they don’t always listen to instructions. You can ask for a “market woman holding a POS machine,” and they will give you a woman with seven fingers holding a calculator that is melting.

Nano Banana Pro is different. It is the first AI designer that thinks, drawing on Google’s Thinking process, before it draws. It builds a plan, which partly eliminates the risk of hallucination. It understands physics, it understands consistency, and for the first time, it understands that when we say “Lagos traffic,” we don’t just mean “many cars.” We mean real Lagos chaos: screaming conductors, tired passengers, impatient drivers, etc.

Let’s dig in.

KINI ANFAANI? WHAT DOES IT DO?

Let’s leave all the big English (technical jargon) in Silicon Valley. Let’s talk about what this tool can actually do for you when you are sitting in front of your laptop at 2 AM trying to beat a deadline.

1. It can “Reason” Through a Scene: This is the biggest flex. It understands cause and effect. If you ask for a rainy season scene, it knows the ground must be wet, reflections must be visible in puddles, and people should be huddled under shelter. Imagine asking for “A busy Danfo bus stop.”

• Other AI(s): Might put the yellow stripe on the wrong side of the bus or make the bus look like a London double-decker painted yellow.

• Nano Banana Pro: Reasons. It gets the boxy shape right. It understands the crowd density. It knows the conductor hangs out the door. It simulates the vibe.

2. Text Rendering (It Can Finally Spell): For years, AI has given us “gibberish text” that looks like alien hieroglyphics. Nano Banana Pro has cracked the code. It can render legible, correctly spelt text on signs, shirts, and billboards within the image.

   For example, let’s say we want to visualize a marketing campaign for us, Kini AI, before spending millions on real ad space. We want to see how our brand looks on a massive billboard at the Lekki-Ikoyi Link Bridge roundabout.

   Previously, we’d have to generate the image, then open Photoshop to manually type the text and try to match the perspective. Now, we simply have to ask Nano Banana. It renders “Learn AI in Simple Terms with Kini AI” perfectly on the billboard, with the correct font weight and shadow.

3. It can “Lock” Faces (Character Consistency): This is the holy grail for African storytellers and comic creators. You can generate a character, give them a name (let’s say “Korede“), and then put Korede in ten different situations without his face changing.

   • Example You are pitching a storyboard. You need Sola to be eating Amala in Panel 1, running for a bus in Panel 2, and delivering a speech in a boardroom in Panel 3. Nano Banana Pro ensures Oluwakorede has the same tribal marks, the same nose shape, and the same haircut in all three images. It is no longer “Random Black Man #1” and “Random Black Man #2.” It is Oluwakorede.

4. Grounded in Reality: Because it is connected to Google, it knows what real places look like. If you ask for “The National Theatre, Lagos,” it pulls from real data. It doesn’t invent a random building. This is crucial for architects and news reporters who need geographical accuracy.

5. The “12-in-1” Super-Composite: This one be like magic! It’s a huge feature that no other image generator AI has right now. Other AIs struggle if you give them two photos to combine. Nano Banana Pro allows you to upload up to 12 reference images and blend them into one coherent shot. In the example below, we uploaded a picture of Oluwakorede and some accessories, asking the AI to make him wear (or hold) them.

WHO YOU EPP? WHO’S IT FOR?

This is not just for you to read. You have to try it too. So, how might you?

1. The “I-Just-Dey-Start” UI Designer: You are mocking up a website for a Nigerian Real Estate client. You need photos of “Modern Duplexes in Lekki,” but stock photos are expensive, and free ones look like houses in California. You use Nano Banana Pro to generate specific, localized assets. “Modern white duplex, Lekki style, sunny afternoon, black gate, interlocked floor.” You get a placeholder that looks real enough to sell the design.

2. The Business Owner: You sell Ankara tote bags on Instagram. You have the product, but you can’t afford a model and a photographer every week. You can take a photo of your bag, upload it to Nano Banana Pro and prompt: “A stylish African lady walking down a street in Victoria Island, wearing this bag on her shoulder.”

3. The Event Planner (The Owambe Specialist): A bride wants an “Enchanted Garden” theme, but can’t visualize it. She is stressed. You prompt: “Wedding reception layout in a large marquee, purple and gold theme, hanging wisteria flowers from the ceiling, golden chivalry chairs, Nigerian wedding vibe.” You show her the picture on your phone. She screams in excitement. You get the deposit. Chikena (finish!)

4. The Fashion Designer: You have an idea for a new Agbada style, but you can’t draw well enough to show the tailor. You prompt: “Modern Agbada design, emerald green Aso-Oke, complex embroidery on the chest, worn by a tall man, fashion illustration style.” The tailor sees exactly where the embroidery should go. No more “I thought you said…” arguments.

To cut the short story shorter, as long as you have a job, Nano Banana Pro has something for you!

How Do You Enter? (Access & Pricing)

Oya o, I’m Ready! How do I start?

• The App: To access and use the Nano Banana Pro model, open the Gemini App on your phone or browser. Go to “Create Images” and ensure the toggle is set to “Gemini 3 (Thinking)”.

• Google AI Studio: You can also use the model in Google AI Studio, this is especially for developers.

• Google Workspace: Certain Google tools, like Docs and Slides, allow you to generate images within them. But this only works if you’re on the PRO version of Google’s Gemini model.

The Price Tag:

• Free Tier: Google gives you a “Daily Allowance” of about 20-50 generated images (depending on complexity) for free. After that, it slows down or kicks you to the older, dumber model.

• Google AI Premium: For about $20/month (approx. ₦30,000 – ₦35,000 depending on exchange rate), you get unlimited high-speed generations, access to the advanced “Thinking” mode, and integration into Google Docs/Slides.

• For Developers (API): If you are building an app, pricing is per image. A standard HD image costs roughly $0.04 (approx ₦60), while a complex “Thinking” image can cost up to $0.14.

PROs and CONs

Why You Should Consider Using Nano Banana Pro

• The “Thinking” Engine: It follows instructions better than any other tool. If you say “remove the background people,” it actually does it.

• Local Context King: Because it is connected to Google Search, it knows what the National Theatre looks like. It knows what Jollof Rice looks like.

• Text & Typography: Being able to put real dates, prices, and slogans in images is a massive productivity booster.

• Google Ecosystem: It plays nicely with Google Slides and Docs. You can generate images directly inside your presentation deck.

• Safety First: It has SynthID (invisible watermarking). If someone steals your AI image, you can prove it was generated by you/Google.

WAIT, Just Before You Jump In… (Shine Ya Eyes)

• The Speed Limit: Because it “thinks,” it is relatively slower. A complex render can take 30-60 seconds.

• Strict “Mummy” Filters: Google is very careful, as per usual. Nano Banana Pro will refuse to generate anything it considers “unsafe” or “copyrighted.” Sometimes it is oversensitive and blocks innocent prompts (like “shooting a movie scene“).

• The “Plastic” Look: Sometimes the images are too perfect. Skin looks like porcelain. You often need to add words like “film grain” or “raw photo” to the prompt to make it look human.

• Cost: The heavy reasoning burns data and credits (especially for API users). It is not light on your wallet if you are a heavy user.

KOKO OF THE MATTER (BOTTOM LINE)

For a long time, we treated AI like a magic trick. “Wow, look, it made a picture!” Nano Banana Pro is more than that. It is a productivity engine.

If you are a designer, do not fear that this will take your job. Fear the designer who uses this to do 3 days of work in 3 hours. This tool removes the grunt work: the tedious sketching, the endless searching for stock photos that don’t exist.

It frees you up to do the real work: The Creative Direction.

Stop fighting the current. The river flows one way. Grab a paddle (or a Nano Banana) and start rowing.

Till next time, stay curious. Stay ahead.

FIRST OF ALL – INTRODUCTION

Most of us reading this were born before Google search engine became the global number one. And we had barely just gained access to use mobile phones and internet when it happened. This means that most of our lives, every time we’ve had a question, Google has always been the first place we run to. Who is the president of Nigeria? What’s the capital of Uzbekistan? Top 5 thrift stores in Lagos? Just head to Google, type it in, and get your answer (albeit wrapped in hundreds of blue links).

That used to be the standard. You click, scroll, compare, and cross-check until you finally find something useful (if you don’t get tired first). Then came the conversational AI era. Now, you can just throw your question at a chatbot like ChatGPT and get a clean, structured response in seconds.

But that evolution came with its own trade-offs. These models, powerful as they are, were trained on static data. So if GPT-4 (ChatGPT) or Llama 4 (Meta AI) was last trained in 2023, it has no idea what happened in 2024. Ask it for the best thrift stores today, and it might suggest shops that shut down a year ago. Below, you will find a screenshot of me asking Meta AI who the last winner of the FA Community Shield is. It tells me — with receipts — that Manchester City is the last winner. However, just a couple of weeks ago, Crystal Palace won the Community Shield, but Meta doesn’t know this because it’s built on Llama 4 which was last updated in April, 2025.

This gap between static conversational AIs and the ever-evolving web was exactly what one man sought to fix. Aravind Srinivas had interned at Google DeepMind, where he got a front-row seat to how the world’s most powerful search engine works. Later, he worked full-time at OpenAI, learning how conversational AI could simplify access to information.

What happens when someone who understands both worlds— search engines and large language models— tries to build a better tool?

You get Perplexity AI.

Perplexity is a new kind of answer engine. It responds like ChatGPT, but behind the scenes, it’s scanning the internet in real-time. Instead of giving you outdated guesses or dumping a blizzard of blue links on you, it does the hard research—searching, sorting, verifying. Gbam! It responds with up-to-date answers tailored for your specific query.

While other AI models like ChatGPT and Gemini now offer real-time search and citation capabilities, Perplexity AI was the first to pioneer this innovation in answer engines. And it all started in 2022 when the CEO of Perplexity, Aravind, experienced a deeply personal frustration. He was searching for health insurance and instead of receiving direct, actionable answers, he was confronted with endless lists of blue links that required manual sifting through multiple websites (sound familiar?). This moment of frustration sparked the vision that would eventually become Perplexity AI.

KINI ANFAANI? WHAT DOES IT DO?

So, what exactly can Perplexity do? What makes it so special that we’re dedicating an entire week to spotlighting it?

As we already established, Perplexity is a real-time, web-connected answer engine. That means when you ask it a question, it doesn’t just pull from a limited, outdated dataset like many other AI tools. It goes out into the actual web, scans through live pages, pulls up-to-date information, and then delivers that info to you in clear, conversational language.

Here are some of its most impressive features:

1. Real-time Web Access
   Perplexity doesn’t just rely on what it was trained on. It actively searches the internet as you ask. So whether you’re looking for the thorough answer to your assignment, a breaking news update, or current market trends, it’s got your back.

2. Citations & Sources
   One thing people often struggle with when using AI tools is not knowing where the answers are coming from. Perplexity solves that. It shows citations for every response it gives, so you can see the exact websites it pulled info from.

   

   Citations from Kini AI’s research paper on Perplexity AI

   

   Web version of Perplexity’s citations

3. Deep Research Mode

   Students, see sharp update for here! Perplexity has a Deep Research mode that allows you to go deeper into a topic. When activated, it pulls more comprehensive insights, giving you a mini-report on your query. Before writing this article, we used this feature to do our own findings and it delivered a 19-page paper! This one no be bobo, see the paper here.

4. File Uploads & Document Queries
   You can upload documents, voice notes, images, etc. and ask questions about them. Perplexity will read through the doc and provide contextual answers, perfect for students, analysts, or legal teams. In the screenshot below, we uploaded the promotional poster containing the details of an essay competition. It analyzed the image, and produced our desired outcome.

5. Model Router
   Perplexity isn’t locked into just one AI model. With its Model Router, you can choose whether your query should be answered using GPT-4, Claude, or other top models, depending on what you want. This gives you richer responses, more flexibility, and access to diverse perspectives.

   When you activate the “Best” option, Perplexity chooses the most suitable model for your query. Otherwise, you get to decide.

   Note: This feature is available only to Pro users. Sign up for our workshop and get free access to a Perplexity Pro subscription!

6. Threaded Conversations
   You can have continuous, structured chats with Perplexity. It keeps the context of your previous questions, which makes it great for ongoing research, not just one-off queries.

7. Export & Share Features

   You can easily share your conversations with others via link (great for collaboration or transparency) and export your results as markdown or text. This is perfect for adding to reports, blog posts, or documents without having to copy and clean up manually.

8. Labs Feature
   This is one of Perplexity’s most exciting and underrated capabilities. The Labs section lets you run advanced tasks like multi-step research, writing assistance, data analysis, and even custom workflows. It’s perfect for people working on more technical or layered tasks who need the AI to go beyond a single query. You can instruct it to compare sources, summarize lengthy documents, or create detailed reports.

You’re probably wondering: “If Perplexity is this good, how come I’ve never heard of it?” Well, Perplexity is still a startup. They’re not yet as legendary or loud as Google and OpenAI. However, they’re growing fast. In fact, their strategy so far has been partnerships over publicity. Instead of splashing billboards everywhere, they’ve been working with institutions, communities, and creators to get their tool into the hands of people who will actually use it.

We at Kini AI are proud to be one of their official partners

And in keeping faith with our partnership, we are pleased to announce that we’re hosting a free hands-on workshop on September 30th to teach you how to use Perplexity like a pro. From mastering its Deep Research Mode and Labs feature to setting up personalized research threads, you’ll learn all the tips that’ll save you hours of research time and help you produce smarter work.

P.S – Attendees get free access to Perplexity Pro. There are limited slots available, so click on this link right now to register.

WHO THIS THING HELP? WHO SHOULD ACTUALLY USE IT?

1. Students & Researchers
   Especially final year students! Perplexity is your best friend. Instead of bouncing between tabs, skimming through long PDFs, or digging through outdated forums to find a reliable source, Perplexity should be your go-to tool. It will help you with real-time research, give citations, and save you hours.

2. Writers & Content Creators
   If you need accurate, timely information to back your writing, Perplexity helps you go beyond surface-level Google searches. It offers summaries, references, and follow-up questions that help you dive deeper into any topic.

3. Founders & Operators
   If you’re building something, or scaling, or trying to make a business case, Perplexity can help you benchmark competitors, understand market trends, or even gather strategic insight; all in one prompt.

4. Professionals in any field
   No matter the field you’re in or are trying to go into, if you need data, analysis, or expert summaries, Perplexity gives you solid starting points without all the clutter.

5. Curious minds

   Even if you’re not doing formal research, Perplexity is great for learning new things with context, depth, and clarity. You can ask it why, not just what, and get answers that actually make sense.

   If you fall into any of these categories, start exploring Perplexity on your:

   • Web Browser

   • Android device

   • iOS device

PROs and CONs

Why You Should Consider Perplexity

1. Real-Time Web Access
   Unlike many AI tools trained on static data, Perplexity browses the internet live so you’re getting fresh, up-to-date info.

2. Citations Built-In
   Every answer comes with sources, so you can trace where the information came from. No more guessing or wondering, “Is this even true?”

3. Model Router
   You can choose what model you want Perplexity to use: GPT-4, Claude, Mistral, etc. It’s like choosing the best brain for the job.

4. Deep Research with Follow-Ups
   Perplexity isn’t just good at giving answers, it’s good at asking the next best question. It recommends related queries that help you dig deeper.

5. Labs Feature
   This is where the tool gets more experimental and powerful. Labs lets you explore Perplexity’s full potential, like analyzing PDFs, building custom agents, or summarizing long docs.

6. Share & Export Options
   You can share your conversations or export them in multiple formats (Markdown, PDF, plain text). Useful for collaboration or documentation.

7. Speed & Simplicity
   It’s fast, intuitive, and clean, unlike the chaos of 15 tabs on your browser when you’re using Google.

WAIT, Just Before You Jump In… (Shine Ya Eyes)

1. Not 100% Hallucination-Free

   Like all AI tools, Perplexity can still make errors or misinterpret information. The citations help, but you still need to verify.

2. Limited Without Pro Subscription
   The free version is useful, but you won’t get access to multimodal model choices, Labs, or advanced capabilities without upgrading.

3. Sometimes Over-Simplified
   In trying to summarize or simplify, Perplexity may skip useful nuances, especially with complex or controversial topics.

4. No Personalization (Yet)
   Unlike some AI tools, Perplexity doesn’t yet learn your style, tone, or preferences over time.

Koko of The Matter – Bottom Line

If you Google a lot, then Perplexity is built for you.…and you should be Perplexity-ing a lot from now!

And if you want to get the best of it, join our free Perplexity workshop on September 30th — and get free access to a Perplexity Pro subscription. Limited slots available.

FIRST OF ALL – INTRODUCTION

No matter who we are or where we live, we’ll always have that one guy – or customer, or vendor, or someone sha – who promises more and delivers less. The ones that boast “I fit do am pass you!” but when it’s time to actually show workings, they fumble. Well, that’s been the story with AI image editors for years. We’ve had a plethora of them who talk big game but when you upload your selfie, suddenly you look like somebody else’s cousin from another continent.

But just a while ago, Google dropped something called Nano Banana that’s got Adobe executives probably having some very uncomfortable boardroom meetings. And before you ask, yes, “Nano Banana” is the real name.

Picture this scenario: You wan edit your photo for that quick Instagram post, but Photoshop dey do like say you wan learn rocket science. You download am, the thing heavy pass your laptop, tutorials long pass Nollywood movie, and by the time you finish learning, your data don finish and the moment wey you wan post don pass.

Enter Nano Banana: the AI that basically said “No wahala, just tell me wetin you want make I do am for you”. No layers, no masks, no “create new adjustment layer and blend with multiply mode.” You just talk to am like say you dey follow your guy talk: “Remove that person in the background” or “make this guy wear agbada” and the thing just delivers. That is exactly the vibe that Nano Banana brings to the table.

As mentioned earlier, Nano Banana is Google’s latest AI image editing powerhouse, built on the Gemini 2.5 Flash Image model. Despite its playful name (which the community initially coined as an internal placeholder),  this tool lets you edit photos using simple conversational prompts instead of wrestling with the technical jargons and complex Photoshop tools.

KINI ANFAANI? WHAT DOES IT DO?

1. Natural Language Editing: Upload any photo and describe what you want changed without sweating over the usual technical jargons. All you need is a simple “Make this guy wear agbada”. Nano Banana gets it.

1. Identity Preservation: Unlike other AI editors that turn your face into someone else’s after a few edits, Nano Banana keeps you looking like you throughout multiple rounds of changes.

2. Multi-Turn Refinement: Make one edit, then another, then another, like you’re having a conversation with your photo editor. The AI remembers your previous edits and builds on them.

3. 3D Figurine Creation: The viral trend that’s everywhere. With Nano Banana, you can turn any photo into a realistic 3D collectible figure complete with packaging.

5. Background Magic: With simple prompts, it can help you replace skies, swap locations, or completely transform environments while keeping everything else natural-looking.

WHO THIS THING HELP? (WHO SHOULD ACTUALLY USE IT?)

1. Content creators & influencers: Make consistent character shots for Instagram posts, product promos, or mini-series without a full photoshoot.

2. Small businesses & e-commerce sellers: Transform basic product shots into high-end lifestyle images without hiring photographers. It can also let customers see how clothing, accessories, or makeup looks on real people without expensive model shoots.

3. Designers & indie studios: Create quick mockups and concept visuals for client presentations without spending hours in traditional design software. You can also easily and quickly generate consistent brand mascots, characters, or visual elements that can be used across campaigns.

4. Casual users: Create professional-looking headshots and personal brand materials for LinkedIn, or other social media platforms. You can also transform ordinary selfies and photos – from that your phone wey the camera no too sharp – into share-worthy content that stands out in feeds.

The beauty of Nano Banana lies in its accessibility. You don’t need advanced technical skills or expensive software licenses to achieve professional-looking results. Nano Banana democratizes high-quality visual content creation for everyone, as long as you have access to the internet.

PROs and CONs

Why You Should Consider Nano Banana

1. Speed Over Everything: Most edits finish in 2-10 seconds versus the hours you’d spend watching YouTube tutorials and still not getting it right. It also makes your workflow faster, if you’re a professional.

2. Zero Learning Curve: If you can describe what you want, you can use Nano Banana. No need to understand all the terminologies of design and editing.

3. Insane character consistency: Nano Banana maintains facial features and object identity across multiple edits, a huge improvement over previous AI editors.

4. Multi-image workflows: It can handle storyboards or multiple product angles in one go.

5. Integration Paradise:  It lives inside Google Gemini, so there is no need to get separate apps or complicated workflows.

WAIT, Just Before You Jump In… (Shine Ya Eyes)

1. Deepfake risks & ethics: With great power, comes potentially  great misuse: the tool’s ability to keep identity consistent makes realistic fake images simpler to create. There are already worries and early controversies.

2. Cultural bias / source mix: Like many models, outputs can lean toward the visual styles or reference material the model saw during training; sometimes edits feel too “oyinbo” unless carefully prompted.

3. Quality Ceiling: Nano Banana’s resolution capacity is capped at 720p so while it is actually impressive, it still can’t match Photoshop’s precision for professional work. Fine details and complex scenes can also get muddy.

4. Watermark Reality: All images come with Google’s SynthID watermark, visible or invisible, marking them as AI-generated.

5. Control Constraints: You can’t “nudge left by 5 pixels” like in Photoshop. If the AI doesn’t get it right, you’re stuck with suggestions, not surgical precision.

Koko of The Matter – Bottom Line

Nano Banana isn’t going to replace Photoshop for professional photographers and designers (at least, not yet), but for the rest of us who just want our photos to look good without a computer science degree, it’s a game changer.

It’s democratizing image editing for the masses by making professional-looking results accessible to anyone, as long as they  can describe what they want. While Photoshop remains the tool for precision work, Nano Banana is the everyday helper that gets things done fast and well enough for social media, small business needs, and personal projects.

Try Nano Banana free in Google Gemini here. Just upload an image and start describing what you want to change.

Tool of The Week

Meta AI: The Antman of the AI Universe

My friends and I love to play the 20 Questions game where we split into two teams. One team picks a word, and the other tries to guess it by asking 20 yes-or-no questions. I’m sure you’re wondering what that has to do with Meta AI. Find out here

And there you have it! That’s all I can fit into today’s update. See you later this week. Peace! 🤓 

FIRST OF ALL – INTRODUCTION

My friends and I love to play the 20 Questions game where we split into two teams. One team picks a word, and the other tries to guess it by asking 20 yes-or-no questions. I’m sure you’re wondering what that has to do with Meta AI.

Well, you’ll find out soon. During one of these games, my team picked Antman as our word.

Now you’d think that once the guessing team figured out it was a fictional character…and then a superhero, they’d get the word in no time. They asked all the right questions: “Is he in the Marvel universe?” “Is he an Avenger?” But they somehow couldn’t land on Antman, despite him standing through every filter they applied. They yelled out every other Avenger’s name till they ran out of trials. Captain America. Iron Man. Even Vision. But not Antman.

That’s exactly what happens with Meta AI when you ask people about AI assistants. Last week at Kini AI, we held three workshops with three different groups, and our boss Rotimi Awaye asked all the attendants the same question: “What AI assistant do you use?” The answers rolled in and all we saw was ChatGPT, Perplexity, Claude, Gemini. Meta AI? Cricket sounds.

In the AI ecosystem, ChatGPT is Iron Man —flashy, brilliant, everyone’s favorite— and Perplexity is Batman with its detective skills and reliable intel, then Meta AI is Ant-Man. Nobody expects much from the guy who started as a petty criminal (hello, Cambridge Analytica scandal), and nobody’s calling him first-string Avengers material. Nobody calls on him to come save the world, but he’s surprisingly handy when you give him a decent shot. I mean, the Avengers literally couldn’t have pulled Endgame off without him. 

In this article, we’re going to talk about Meta AI: what it can do, how it works, who it’s useful for, the pros and cons, and why it might matter more than people give it credit for.

Meta AI, free of my Antman juxtapositions and ChatGPT comparisons, is just the conversational assistant from Meta (the folks behind Facebook, Instagram, and WhatsApp). It’s baked right into apps you already use daily, meaning you don’t need to go hunting for a separate site or download another “productivity app you’ll forget in two weeks.” It’s right there in your DMs or search bar.

• In WhatsApp, you can pull it up directly in chats or group conversations.

• On Instagram and Facebook, you’ll find it in the search bar and Messenger.

  

• It also exists as a web version at Meta AI.

  

KINI ANFAANI? WHAT DOES IT DO?

At its core, Meta AI is designed to function as a versatile digital assistant. Think of it as a tool that blends knowledge retrieval with creativity, helping users with everyday tasks like drafting emails, or brainstorming ideas, while also offering interactive features such as conversation, learning support, and entertainment.

It doesn’t have the depth of other top-tier models, but it is good enough for certain things. But sometimes “good enough” is exactly what people want. Not everyone needs a PhD-level response to “What should I make for dinner?” Sometimes you just want an AI that doesn’t make you download another app or remember another password. Here are a few things that Meta AI are, with good prompting of course, good at:

1. Quick Answers: Ask it trivia (“Who won the 2015 Champions League?”) or simple definitions, and it’ll respond almost instantly. It’s not as detailed as Perplexity or ChatGPT, but for fast facts, it gets the job done.

   

2. Creative Prompts & Writing Help: If you need a caption for your Instagram post or you need to draft a witty birthday message for that your guy, Meta AI will generate good enough ideas you could use on the spot.

   

3. Planning & Recommendations: It can suggest restaurants, create to-do lists, or even help plan that get-away something you have been thinking about. It will help to think of it as a basic life-organizer inside the apps you already check 100 times a day.

4. AI Image Generation: This is where Meta AI really surprises people. You can ask it to create simple images (e.g., “a motivational poster in the style of Japanese anime” or “a cat wearing sunglasses on the beach”), and it’ll deliver without you having to leave WhatsApp or Instagram. It’s not Midjourney level but it’s impressive for quick fun.

   Even more exciting: Meta AI and Midjourney have just entered a partnership, meaning Meta will start leveraging Midjourney’s advanced image technology in the future. So, expect the creative quality of Meta AI’s images to get a serious upgrade soon.

   

   Generate an image of a young Nigerian boy devouring a giant bowl of suya under Lagos street lights

Who This Thing Help? (Who Should Actually Use It?)

1. The Social Media Native: If you’re already deep in the Meta ecosystem and have made peace with the privacy trade-offs, Meta AI offers genuine convenience. You’re already giving them your data anyway, might as well get some AI assistance in return.

2. The Casual User Who Needs Quick Answers: For basic tasks, simple image generation, and quick information lookup, Meta AI gets the job done without the overhead of learning a new platform.

3. The Budget-Conscious Creator: If you need AI assistance for content creation but can’t justify another subscription, Meta AI provides decent tools at the low, low price of your personal data.

PROs AND CONs

Why You Should Consider Meta AI

1. The Price is Right: Meta AI is completely free, but as the old saying goes, if you’re not paying for the product, you are the product. Every query you make, every image you generate, every interaction feeds back into Meta’s vast data collection engine. But the plus side really is that unlike the big guns, it has no charges at all. It is absolutely free.

2. Speed: One thing Meta AI consistently gets right is speed. While tools like ChatGPT or Gemini sometimes take a few seconds to “think” before responding, Meta AI responds almost instantly. For everyday stuff like trivia and quick ideas, faster responses are often more valuable than deeper ones. And that’s exactly the lane Meta AI owns.

3. It Lives Where Your Life Happens: While other AI assistants exist in their own apps, Meta AI embedded itself into your existing digital routine. If you need help clearing up an argument in your group chat? It’s right there. Want to generate an image for your Instagram story? No app-switching required. Its genius is its super convenience.

WAIT, Just before you Jump in… (Shine Ya Eyes)

1. Hallucinations: Like all AI assistants built on large language models, Meta AI has a risk of “hallucinations” where the AI generates information that sounds factual but is actually made up or incorrect. This is a well-known issue with most conversational AI tools, and it happens because these systems predict text based on patterns rather than true understanding or real-time fact-checking. So as a user, you gats check am well!

2. Privacy Trade-Off: Meta AI’s biggest problem is creeping feeling that it knows way too much about you. The system is trained on massive amounts of user data from across Meta’s platforms, which means it potentially knows your relationship status, your political leanings, what you bought last week, and that 3am ofofo wey you dey do for IG. The company promises privacy-preserving AI, but historically speaking, (clears throat… seems like some bobo)

Koko of The Matter – Bottom Line

Meta AI may not be the showstopper in the AI scene, but it’s rapidly becoming the reliable sidekick for millions of everyday users. And it can be yours too. It doesn’t boast the flair or depth of other models, but it wins with speed, accessibility, and the simple fact that it lives right where you are. Whether you want quick answers, creative prompts, or just an extra bit of help organizing your day, Meta AI is proof that sometimes the underrated hero (like Antman) can also be the most useful and reliable one.

It’s been a little while, and we’re thrilled to be back in your inbox! We hope your year has been productive so far. We’ve been deep in the world of AI, exploring tools and strategies to help you work smarter, and we’re excited to start sharing valuable insights and opportunities with you again.

To kick things off, let’s talk about a common challenge: Are you drowning in digital notes, research papers, articles, and useful links? Struggling to connect the dots or find that crucial piece of information right when you need it most?

Imagine having a personal AI assistant that learns directly and only from your content, giving you answers you can actually trust.

That’s exactly what Google’s NotebookLM offers! It’s a powerful, yet surprisingly simple, AI research assistant designed to work with your knowledge base.

And to help you master it quickly, we’re hosting a live, hands-on Google NotebookLM Masterclass led by Rotimi Awaye (Perplexity AI Business Fellow) – Yours truly!.

Who Is This Masterclass For?

This session is perfect if you’re a:

• Student or Researcher: Drowning in papers or course materials? Learn to instantly query your sources, generate study guides, and synthesize information quickly.

• Writer or Content Creator: Facing writer’s block or need fresh angles? Use your own research to brainstorm ideas and generate outlines effortlessly.

• Professional or Entrepreneur: Need quick insights from reports, meeting notes, or project documents? Master NotebookLM to get up to speed fast.

• Anyone managing lots of digital information and looking for a smarter, more efficient way to learn, create, and work with your own content using AI.

When: THIS Saturday, May 10th, 2025

Time: 5:00 PM – 6:30 PM WAT (West Africa Time)

Where: Virtual

REGISTER HERE – https://selar.com/741kw6 

KINI BIG DEAL (Why does this matter)

The koko of the matter is that we’ve all used generic AI, but the real game-changer is having an AI that understands your specific context because it learns directly from your trusted documents. In an era where you need reliable information, mastering a tool like Google’s NotebookLM gives you relevant insights grounded in sources you control.

See you on Saturday!

The Claude Android app is now available, offering users a powerful new way to interact with artificial intelligence on their mobile devices. The app provides all the features of Claude on iOS and the web. (Read More)

Major Highlights

• The Claude Android app offers the same robust functionality as our iOS and web versions:

  • Cross-Platform Continuity: Seamlessly continue your conversations with Claude across web, iOS, and Android platforms.

  • Visual Analysis: Utilize Claude’s image processing capabilities by taking new pictures or uploading existing files for real-time analysis.

  • Multilingual Support: Benefit from instant language translation to enhance communication or interpret your surroundings while traveling.

  • Advanced Problem-Solving: Leverage Claude’s sophisticated reasoning to tackle complex tasks, from analyzing legal documents on-the-go to conducting market research for upcoming meetings.

KINI BIG DEAL (Why Does this matter)

A lot of Android users have been expecting this release, now we have a closer match to ChatGPT’s mobile app. The Claude mobile app offers several benefits to the average person, particularly its accessibility and helpful features for everyday tasks. Whether for professional tasks or personal projects, Claude is your capable companion:

• Draft business proposals between meetings

• Translate menus and signs while exploring new countries

• Brainstorm gift ideas during shopping trips

• Compose speeches or presentations while waiting for your flight

Download it here for free – Claude Android app

AI Education

Useful AI Tools and Resources

Prompt of the Day

And there you have it! That’s all I can fit into today’s update. See you later this week. Peace! 🤓 

Google is making a major push for its AI-powered research and writing assistant, NotebookLM. After a limited initial launch, NotebookLM is now going global, available in over 200 countries and territories. This global expansion is accompanied by a significant upgrade powered by Google’s latest AI technology, Gemini 1.5 Pro.

The core purpose of NotebookLM remains the same: to assist users in navigating complex information, making connections between ideas, and accelerating the writing process. It achieves this by analyzing uploaded source materials – research notes, interview transcripts, corporate documents – and becoming an “expert” on the specific content.

The upgrade brings a wealth of new functionalities to the table. First, NotebookLM expands its source material support beyond Google Docs, PDFs, and text files. Users can now upload Google Slides presentations and web URLs for analysis. This allows for a more comprehensive research workflow, incorporating visual aids and online resources seamlessly.

Secondly, NotebookLM now offers inline citations within its responses. These citations directly link back to the relevant passages in your uploaded sources. This feature serves two purposes: it allows for easy fact-checking of the AI’s output, and it provides users with a quick way to delve deeper into specific points within the original material.

Thirdly, a new “Notebook Guide” feature provides a high-level overview of your uploaded sources. This guide can convert the information into different formats, such as FAQs, briefing documents, or study guides. This functionality is particularly helpful for users who need to quickly grasp the key points of complex source materials.

Perhaps the most impressive upgrade lies in NotebookLM’s newfound ability to analyze visual information. Thanks to Gemini 1.5 Pro’s multimodal capabilities, NotebookLM can now answer questions about images, charts, and diagrams within your source materials. This extends its reach beyond purely textual analysis, offering a more comprehensive understanding of the information being presented. Additionally, NotebookLM can even include citations to relevant images within its responses, further strengthening the validity of its answers.

The video above provides an informative overview of how to effectively use the new tool. You can also read more about it here.

KINI BIG DEAL (Why Does this matter)

Remember NotebookLM, that research tool they were testing? Well, it just got a major upgrade and went global!

This isn’t your average research assistant. NotebookLM is now powered by Google’s top-of-the-line AI, Gemini 1.5 Pro (remember that one you had to pay for?). This means research and projects just got a whole lot easier. No more endless searches or clunky note-taking. NotebookLM can streamline your workflow and help you focus on the good stuff – actually making cool discoveries!

So, ditch the research blues and check out NotebookLM. It’s like having a personal AI research buddy in your pocket (well, virtually!).

OpenAI has unveiled its latest groundbreaking AI model, GPT-4o, during its Spring Update event. This next-generation language model promises to revolutionize how we interact with machines and harness the power of artificial intelligence. With its remarkable capabilities across text, vision, and audio, GPT-4o is poised to reshape the way we approach problem-solving, creativity, and knowledge acquisition. Watch full updates in this video.

Major Highlights

• GPT-4 Intelligence for All: OpenAI’s commitment to making advanced AI tools accessible to everyone is at the core of this release. GPT-4o brings the intelligence of GPT-4 to all users, including those on the free tier, making cutting-edge language capabilities available to a broader audience.

• Natural Language Interaction: GPT-4o introduces a paradigm shift in human-machine interaction. With its ability to reason across voice, text, and vision, the model offers a seamless and natural conversational experience, eliminating the need for multiple models and reducing latency.

• Desktop Availability: To enhance accessibility and user experience, OpenAI introduced a desktop version of ChatGPT powered by GPT-4o. This desktop app (seems like it’s only available on the MAC) offers a streamlined and intuitive interface, allowing users to collaborate with the AI assistant seamlessly within their workflow. The refreshed UI prioritizes natural interactions, ChatGPT can see and hear what is going on on your computer, enabling users to focus on their tasks while leveraging the full capabilities of GPT-4o.

• Enhanced Productivity and Learning: OpenAI envisions GPT-4o as a powerful tool for enhancing productivity and facilitating learning. Users can leverage the model’s advanced capabilities to create content, solve problems, and gain insights across various domains.

• Developer Ecosystem: The release of GPT-4o extends beyond the consumer market. OpenAI is empowering developers to build cutting-edge AI applications by providing access to the model through APIs, enabling the deployment of GPT-4o at scale.

KINI BIG DEAL (Why Does this matter)

GPT-4 performance is now your fingertips for FREE!

The widespread availability of GPT-4o has far-reaching implications for individuals, businesses, and society as a whole. It empowers users to leverage the power of AI in their daily lives, from personal productivity to educational endeavors.

This just opened up a new spectrum of possibilities… Watch this movie trailer to get the full jist!