GLOSSARY: 🍌 Google Took it Personally 🍌

Lately, social media has been drowning in images that look real but are not.‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌

🍌 Google Took it Personally 🍌

THE BELAMY

Weekly Newsletter of AIM

Monday, Nov 24, 2025 | By Mohit Pandey

Now, subscribe to our Digital & Print Editions >

Lately, social media has been drowning in images that look real but are not. One moment, you're scrolling past a medieval bakery in Bengaluru with steam rising softly from a wooden oven, and in the next, you're staring at a black hole rendered like a studio portrait. A comic strip made entirely from O Henry's short story 'The Gift of the Magi', and images generated perfectly with just geo coordinates.

Every one of these visuals carries the same quiet stamp at the bottom corner: Nano Banana Pro. Google has just killed it this time.

It did not just ship an upgrade. It shipped a statement. And the timing is hard to ignore.

A few months ago, OpenAI CEO Sam Altman wrote that he could not stop thinking about the "aesthetic difference between OpenAI and Google". He paired it with a photo of OpenAI's launch in a warm living room, contrasted with the Google I/O stage that looks like geometry class at scale.

That post has resurfaced, now floating around with a bit of extra spice with what people have been calling, "And Google took that personally."

Look Where We Are

Google has released Gemini 3 and Nano Banana Pro in the same week. In a congratulatory message to Google on Gemini 3, Altman wrote, "Congrats to Google on Gemini 3! Looks like a great model." CEO Sundar Pichai responded with the folded hands emoji, and, Elon Musk, naturally, added his voice too.

But behind closed doors, things are tense.

Reports say Altman has warned OpenAI employees to brace for 'rough vibes' ahead. The memo mentions tough competition from Google, cooling investor sentiment and the heavy lift of running a research lab, an infrastructure company and a product company at once.

He says, "It s**ks that we have to do so many hard things at the same time—the best research lab, the best AI infrastructure company, and the best AI platform/product company—but such is our lot in life. And I wouldn't trade positions with any other company."

Altman has encouraged his own team in private, telling them to focus on superintelligence. Despite this, he says the company has enough strength to weather competition. Moreover, he acknowledged the progress of Anthropic, which recently raised $15 billion from Microsoft and NVIDIA.

The Slippery Banana

Nano Banana Pro is the model everyone is talking about at this moment. It sits on top of Gemini 3 Pro, and Google says it brings reasoning, real-world knowledge and more accurate visual output. Users can produce infographics, diagrams, recipes or snapshots using grounded information.

"Built on Gemini 3, it's really good at complex infographics—much like how engineers see the world," Pichai said.

The infographic below was created by a single prompt on Nano Banana Pro.

SAS Academy for Data & AI Excellence is Training India's Workforce for the GenAI Era

SAS, a global leader in analytics, is addressing the demand shortage of skilled engineers in India with SAS Academy for Data & AI Excellence, a curriculum designed to help learners move from foundational analytics to advanced AI and automation, guided by SAS-certified instructors through flexible weekend online programmes. Click here to learn more.

Google says the model can render long text inside images with clarity. It works across languages, blends multiple elements with improved consistency, can combine up to 14 images and still preserve the likeness of up to five people.

Creators already feel its weight. This comes after a long week of people criticising Google's model before its release. "I heard they let Gemini 3 play chess against itself during RL training, and both sides won," entrepreneur Yam Peleg said.

Even Andrej Karpathy joined in. "I heard Gemini 3 answers questions before you ask them. And that it can talk to your cat," he joked, poking fun at the growing hype.

But once the model dropped, reactions shifted dramatically.

Today, Karpathy is openly praising the models. "Gemini Nano Banana Pro can solve exam questions *in* the exam page image. With doodles, diagrams, all that," he wrote on X, adding a surprised emoji.

He called the Gemini 3 release a big step towards AGI, adding that the model is built to grasp depth and nuance while perceiving subtle clues in creative ideas. Meanwhile, DeepMind CEO Demis Hassabis and CTO Koray Kavukcuoglu declared, "We're beginning the Gemini 3 era."

>From day one, the model is embedded inside Search, Workspace, the Gemini app and developer tools. The numbers are loud. Gemini 3 Pro outperforms Gemini 2.5 Pro, OpenAI GPT 5.1 and Claude Sonnet 4.5 across LMArena, Humanity's Last Exam, GPQA Diamond and MathArena Apex.

Developers also get Google Antigravity, a new agent-first IDE built around Gemini 3. Agents can plan and execute complex tasks without supervision. Gemini 3 integrates with AI Studio, Vertex AI, Gemini CLI, Cursor, GitHub, JetBrains and Replit. However, online discussions suggest that it is just a fancy VS Code fork.

These multiple launches in a week have completely shifted the market, something Google desperately wanted.

Check out more details and reactions to Nano Banana Pro on AIM Network's Front Page >>

Google has rolled out this crazy image generation model inside the Gemini app under the 'Thinking' option for creating visuals.

Users part of the free tier get a small taste, while AI Plus, Pro and Ultra users unlock more power. It's also available in Search's AI Mode for Pro and Ultra users in the US. Professionals will start seeing it soon inside Ads, Workspace, Slides, Videos and Flow.

Google is also pushing for transparency as it accelerates. Every output continues to carry SynthID watermarking, and users can upload an image and ask Gemini if it was generated by Google AI. Free and Pro tier outputs also get a visible Gemini watermark.

Meanwhile, OpenAI has responded with GPT 5.1 Codex Max—a long-running coding model that can work for more than 24 hours without intervention. It uses fewer thinking tokens and can run large tasks through compaction.

OpenAI says its engineers ship 70% more pull requests since adopting Codex. The coding model is strong, the product roadmap is busy and the ecosystem is broad.

Yet, this week, the momentum, and the conversation, belongs to Google.

AIM Print November 2025

In the November 2025 edition of AIM Print, we map how India's tech landscape is shifting. For years, foreign companies shaped the country's digital habits. Now, the ground is moving under them. A new Swadeshi stack is taking shape through firms like MapMyIndia, Zoho and HCLSoftware, backed by a growing belief that India needs to build and control its own digital infrastructure. Click here to check it out.

[AIM Network Exclusive]

In this episode of Simulated Reality, Yadvendra Kshatri, co-founder and head of AI, Aidetic, shares the journey of building the company from early experimentation with video analytics and transformer models to delivering over 300 production-grade AI solutions.

[Developers' Meetup] Building Tomorrow: Where Cloud, AI & Data Converge

We just wrapped up the first developer's meetup in Bengaluru, and it was a blast!

Join us in Pune at the Dentsu Global Services Developers' Meet Up 2025 on December 6 as developers explore the powerful synergy that's reshaping innovation. Register here.