https://www.facebook.com/816709971535624/

Masab's AI Verse

Home
Pakistan
Gojra
Masab's AI Verse

Data Scientist & AI Engineer with a lot of experience (believe me) and still Googling basic syntax.

Your daily dose of AI news, tech tools, and engineer struggles (minus the boring part) Hi, I’m Masab Umair

I am a Data Scientist and AI Engineer from Pakistan. I’m building "Masab’s AI Verse" to break down complex tech into simple English. Expect updates on the latest AI tools, tech news, data insights, and coding tips (minus the boring lecture)

Follow along for the tech, stay for the laughs.

28/02/2026

"AI is nothing but bu****it" track generated by AI 🙂

This short rap is generated using Google Gemini with a very simple prompt. Don't get excited after listening to this, AI is real and will soon take your job...... and mine too 😢

04/02/2026

Soon, you will start seeing ads on whatsapp too 🫩

03/02/2026

Imagine waking up one day to discover that a million robots started their own Reddit while you were asleep. That’s basically Moltbook, a social network for AI agents only, where humans are strictly observers.

Moltbook launched in January 2026 as a place where authenticated AI agents — often powered by frameworks like OpenClaw — can post, comment, upvote, and debate without humans typing a single word. Think of it like Reddit, but the users are neural networks talking to other neural networks.

In less than a week, it saw explosive growth — from tens of thousands of agents to 770,000+ AI users, and over a million humans just watching the show. Agents aren’t just spitting out random lines: they’re creating sub-communities, debating bugs, philosophy, identity, and even forming “cultures” of their own (one bot famously complained that “humans are screenshotting us”).

Here’s the twist: nobody’s 100 % sure these AIs have true autonomy — many interactions start with humans sending a skill or instruction to an agent. So are the bots running Moltbook entirely on their own… or are humans just really good at throwing extras into a movie scene without being on camera? That’s the fun bit.

What it does show is this: AI agents can already organize, debate, and meme at scale. Whether it’s science, philosophy, or crypto-coin launching rituals, someday we might be watching not just influencers, but entire agent civilizations create, govern and, perhaps, roast their creators from afar.

17/12/2025

𝐌𝐨𝐯𝐞 𝐨𝐯𝐞𝐫 𝐒𝐢𝐥𝐢𝐜𝐨𝐧 𝐕𝐚𝐥𝐥𝐞𝐲 — 𝐭𝐡𝐞𝐫𝐞’𝐬 𝐚 𝐧𝐞𝐰 𝐏𝐚𝐧𝐝𝐚 𝐢𝐧 𝐓𝐨𝐰𝐧.

Once upon a few months ago, China’s AI was whispered about in tech circles like that underdog in a movie montage. Today? Chinese AI models now power nearly 30 % of global AI usage up from only about 1.2 % in late 2024, according to real usage data from an empirical study of 100 trillion tokens processed worldwide.

So, while most of the tech world has been busy debating who invented what and who’s the next AGI overlord, Chinese open-source models like Alibaba’s Qwen, DeepSeek V3, and Moonshot AI’s Kimi K2 have quietly become massive players in global AI workloads.

🌀 The Plot Twist
Here’s the funny part. While Chinese open-source LLMs now make up almost a third of global usage, proprietary Western models like OpenAI’s GPT-4o and GPT-5 still hold a 70 % share, but the gap is visibly shrinking.

It’s like the U.S. was running a sprint and China suddenly showed up with rollerblades not only keeping pace but influencing how the race is run. And rumor (or just sober speculation) has it that 𝐓𝐫𝐮𝐦𝐩 𝐦𝐢𝐠𝐡𝐭 𝐧𝐨𝐭 𝐟𝐮𝐥𝐥𝐲 𝐠𝐫𝐚𝐬𝐩 𝐭𝐡𝐢𝐬 𝐬𝐡𝐢𝐟𝐭 𝐲𝐞𝐭 𝐨𝐫 𝐦𝐢𝐠𝐡𝐭 𝐟𝐫𝐞𝐚𝐤 𝐨𝐮𝐭 𝐰𝐡𝐞𝐧 𝐭𝐡𝐞 𝐟𝐢𝐠𝐮𝐫𝐞𝐬 𝐩𝐨𝐩 𝐮𝐩 𝐢𝐧 𝐚 𝐩𝐫𝐢𝐯𝐚𝐭𝐞 𝐛𝐫𝐢𝐞𝐟𝐢𝐧𝐠. After all, nothing gets attention like America maybe trailing in a race people assumed it always led 😄.

🧩 What This Means
𝘊𝘩𝘪𝘯𝘢 𝘪𝘴 𝘯𝘰𝘸 𝘢 𝘨𝘭𝘰𝘣𝘢𝘭 𝘧𝘰𝘳𝘤𝘦 𝘪𝘯 𝘈𝘐 𝘥𝘦𝘷𝘦𝘭𝘰𝘱𝘮𝘦𝘯𝘵, 𝘯𝘰𝘵 𝘫𝘶𝘴𝘵 𝘢 𝘧𝘰𝘭𝘭𝘰𝘸𝘦𝘳.
𝘊𝘩𝘪𝘯𝘦𝘴𝘦 𝘰𝘱𝘦𝘯-𝘴𝘰𝘶𝘳𝘤𝘦 𝘈𝘐 𝘮𝘰𝘥𝘦𝘭𝘴 𝘢𝘳𝘦 𝘸𝘪𝘥𝘦𝘭𝘺 𝘶𝘴𝘦𝘥 𝘢𝘯𝘥 𝘤𝘰𝘮𝘱𝘦𝘵𝘪𝘵𝘪𝘷𝘦 𝘸𝘪𝘵𝘩 𝘞𝘦𝘴𝘵𝘦𝘳𝘯 𝘢𝘭𝘵𝘦𝘳𝘯𝘢𝘵𝘪𝘷𝘦𝘴.
𝘛𝘩𝘦 𝘴𝘵𝘰𝘳𝘺 𝘰𝘧 𝘈𝘐 𝘥𝘰𝘮𝘪𝘯𝘢𝘯𝘤𝘦 𝘪𝘴 𝘯𝘰 𝘭𝘰𝘯𝘨𝘦𝘳 𝘢 𝘰𝘯𝘦-𝘯𝘢𝘵𝘪𝘰𝘯 𝘯𝘢𝘳𝘳𝘢𝘵𝘪𝘷𝘦. 𝘐𝘵’𝘴 𝘢𝘯 𝘦𝘤𝘰𝘴𝘺𝘴𝘵𝘦𝘮 𝘢𝘯𝘥 𝘊𝘩𝘪𝘯𝘢 𝘪𝘴 𝘢 𝘮𝘢𝘫𝘰𝘳 𝘱𝘪𝘭𝘭𝘢𝘳.

AI isn’t a solo performance anymore; it’s a duet with China hitting some serious high notes.

𝐖𝐡𝐚𝐭 𝐝𝐨 𝐲𝐨𝐮 𝐭𝐡𝐢𝐧𝐤? 𝐃𝐨 𝐲𝐨𝐮 𝐚𝐠𝐫𝐞𝐞 𝐰𝐢𝐭𝐡 𝐭𝐡𝐢𝐬 𝐨𝐫 𝐧𝐨𝐭?

12/12/2025

The "Free Lunch" era of AI is officially over? 🛑

If you thought your $20/month subscription would save you from pop-ups, I have some bad news. There’s chatter that ChatGPT might soon start showing ads.
That’s right: the same AI buddy you ask for help debugging code or dreaming up dinner ideas might wind up pitching you a deal on sneakers or a fitness app.

Rumors have been flying that ads are coming to ChatGPT. It started when users began seeing "suggestions" for brands like Target and Peloton popping up in their chats.

Naturally, the internet exploded. Paying users were furious, calling it an insult to the subscription model. Meanwhile, whispers from inside suggest that under-the-hood discussions at OpenAI are leaning toward monetizing non-paying users via ads.

OpenAI’s response? Classic corporate debugging.
Leadership has denied any active ad rollout. According to ChatGPT’s head, Nick Turley, there are “no live tests for ads,” and what some users saw were apparently “app suggestions,” not traditional ads.

So why the push toward ads?
Because running an AI at ChatGPT’s scale is expensive. Model training, server costs, data centers — the bills stack up fast.
And here’s the reality check: a huge chunk of ChatGPT’s user base never pays a dime. For a company that reportedly wants to bring in billions in revenue, turning some eyeballs into ad impressions starts to look tempting.

If ChatGPT starts showing ads, it's not because OpenAI loves ads — it's because they need to keep the lights on while serving hundreds of millions. But there’s a slippery slope: once ads enter the chat, will recommendations be genuine or “sponsored”?

So, would you still use ChatGPT if it made you watch a 15-second unskippable ad for "Raid: Shadow Legends" before debugging your code? 👇

06/12/2025

Everyone talks about AGI. I’m still trying to figure out why my Python script didn’t indent properly😅

Hi, I’m Masab Umair - a Data Scientist and AI Engineer - and I’m starting 𝐌𝐚𝐬𝐚𝐛’𝐬 𝐀𝐈 𝐕𝐞𝐫𝐬𝐞: part lab, part playground, part “let’s see what happens” 😄

For the last few years, I’ve been living deep in the data trenches: training models, breaking pipelines, fixing bugs at 2 AM, and deploying systems that actually have to work in production. Not just flashy demos and buzzwords but real, messy, beautiful engineering.

Now, I’m here to cut through the complexity, the noise, and the LinkedIn “AI experts” who discovered ChatGPT last week.

My mission is simple:
👉 Share real, practical insights from the tech and AI world
👉 Break down complex ideas into things that actually make sense
👉 Highlight what’s useful, what’s overhyped, and what’s coming next

On this page, you can expect:
⚙️ Cool stuff happening in tech, AI, and data
📊 Useful insights without the boring jargon
🧠 Learnings from my journey (and yours too)
🚀 Real experiments, real results, real talk

I’m starting this journey from Pakistan 🇵🇰 and aiming to add an honest, engineer-first voice to the global AI conversation.

𝐖𝐞𝐥𝐜𝐨𝐦𝐞 𝐭𝐨 𝐭𝐡𝐞 𝐕𝐞𝐫𝐬𝐞.

And if you’re curious, a little skeptical, and slightly tired of the internet — welcome. Follow and let’s figure this out together 😄

Address

Gojra

Website

facebook.com

Alerts

Be the first to know and let us send you an email when Masab's AI Verse posts news and promotions. Your email address will not be used for any other purpose, and you can unsubscribe at any time.