| | | Before Google existed, Larry Page and Sergey Brin named their search engine BackRub. They eventually decided to rename it, considered "The Whatbox" at one point, and landed on Google only because a colleague accidentally typed it wrong while searching for "googol." | A $2 trillion company has a name that is technically a typo. They chose it over BackRub and The Whatbox, which somehow makes it feel even more accidental. | What’s in store: | Google has another banana for us. OpenAI has 99 problems, and they're all maths. Mindstream Picks: Googlebooks are bringing back Android’s classic navigation buttons. HubSpot’s AI CRM is free to use!
| Read Time: 5 minutes |
| |
|
| | | | | GOOGLE | | | Google has introduced Nano Banana 2.1, its latest AI image generation and editing model, built on Gemini 3.6 Flash. | According to Google's October report, the model scored higher than Nano Banana 2 and Nano Banana Pro across several image generation and editing tests. | It can create detailed infographics, generate readable text within images and edit photos while keeping characters, objects and styles consistent. | In brief: | Nano Banana 2.1 scored higher than previous models in Google's tests for image quality, infographic design and editing. Users can create and edit images using text prompts, reference pictures and sketches. Google's safety testing found similar or improved results compared with Gemini 3 Flash, although the model can still make mistakes.
| Peel the difference | The model can also work with text and image prompts containing up to one million tokens, allowing users to provide detailed instructions and multiple references. | It's available across Google's products, including Gemini, Google AI Studio, Search AI Mode, Google Ads and Flow. | However, Google says there's still room for improvement, particularly when generating small text, following complex instructions and getting factual details right. | The company also carried out safety testing before release. | You can really give it a whole book as a prompt. - MV | How are we feeling about Google's latest banana? | | Vote for live results + see results and opinions from yesterday at the bottom of the email. |
| |
|
| | | | | AI TOOLS | | | HubSpot Smart CRM unifies your data and tools, automates the busywork, and gives every team the context they need to take the next best action. | No more switching between systems or reconciling data across disconnected tools. The intelligent CRM connects your apps, syncs your data, and integrates with your existing tools, so your agents, workflows, and data all work from the same source of truth. | Every tool connected. No data left behind. | |
| |
|
| | | | A barista makes 3 types of coffee. An espresso takes 1 minute, a latte 3 minutes, and a cappuccino 4 minutes. In a 2-hour shift, she makes 30 espressos, 15 lattes, and 10 cappuccinos. How many minutes does she spend NOT making coffee? | Find the answer at the bottom of the email! |
| |
|
| | | | | Black holes are where physics starts arguing with itself. Brian Cox explains why these cosmic monsters sit right at the clash between quantum theory and general relativity, and why figuring out what happens to the stuff that falls inside could completely change how we understand space, time, and reality. | |
| |
|
| | | | | OPENAI | | | OpenAI has released a collection of new maths research produced by one of its most advanced internal AI models, with the results now available on GitHub for researchers to review. | The company worked with an independent advisory group at the Institute for Advanced Study to help decide how to share the findings. | Here’s what you should know: | Each result required an average of roughly three hours of ChatGPT Pro thinking power. OpenAI plans to fund workshops, conferences and research programmes to help mathematicians study AI-generated discoveries. The company is also working towards releasing the model behind the research, while continuing to test its abilities in maths and other scientific fields.
| The square root of showing off | Some of the mathematical proofs are also available in Lean, a programming language that lets computers check whether proofs are correct. | OpenAI is sharing more details about how the model reached its results, including summaries of its reasoning, the number of problems attempted and how much computing power was used. | I bet this model would've cried over getting 99% on a test. - MG |
| |
|
| | | | Health: New York has declared a state of emergency after recording more than 100 measles cases this year, expanding access to testing and MMR vaccinations to help contain the outbreak.
Gaming: The writer of Resident Evil Requiem has reportedly leaked details about the upcoming story DLC, hinting at a deeper exploration of one of the franchise’s major villains.
Movies: Aaron Sorkin’s The Social Reckoning has received mixed reviews ahead of its theatrical release, debuting with a 65% Rotten Tomatoes score compared with the original Social Network’s 96%. | Don’t Miss: Google’s new Googlebook laptops are bringing back Android’s classic triangle, circle, and square navigation buttons, now built into the F1, F2, and F3 keys. The shortcuts let users go back, return to the desktop, and switch between apps, making Android-style navigation easier on a laptop. |
| |
|
| | | | Image of the Day |  | Artwork submitted by Mindstream reader Rob: “Cartoon illustration of an underground dark rusted mega city” |
| Daily Image Prompt | Rough wax crayon drawing of a green Coca-Cola can |
|
| Submit your artwork to Mindstream → |
| |
|
| | | | “Would an invisible AI watermark change how you use ChatGPT?” | Yes: I’d edit more carefully - 63% ✅ | No: Watermark away, babes - 37% | Your Views: | “I do so anyway, but if not done, it could reduce confidence that the person hired to do the writing didn't check the AI's writing for accuracy or relevance. That's a problem.” - joseph | “I would use another system” - ascot | “I think this significantly reduces the utility of the LLM” - carats | Submit your opinions in our polls to be featured! |
| |
|
| Number Crunch Answer: 5 minutes! | #1226 gestures wildly |  | | | |
|