Sunday Rundown #110: I, Android & Temu Shopping

Sunday Bonus #70: Swipe file with 72 use cases for Nano Banana

Heads up: I’m away next weekend, so Sunday Rundown returns on September 21

Happy Sunday, friends!

Welcome back to the weekly look at generative AI that covers the following:

• Sunday Rundown (free): this week’s AI news + a fun AI fail.

• Sunday Bonus (paid): an exclusive segment for my paid subscribers.

---

In case you missed it, here’s this week’s Thursday deep dive:

I Checked Out All 35 Google Labs Experiments: Here's The Scoop

---

Let’s get to it.

🗞️ AI news

Here are this week’s AI developments.

👩‍💻 AI releases

New stuff you can try right now:

1. Amazon is rolling out Lens Live, which lets you scan real-world items with your phone to find online matches and get instant insights from its AI assistant Rufus.

2. EPFL & ETH Zurich open-sourced Apertus, a fully transparent family of multilingual LLMs that anyone can inspect and fine-tune. (Find the models here.)

3. ElevenLabs released SFX v2, a text-to-sound-effects model that can create high-quality sound effects for audiobooks, podcasts, film, and more.

4. Genspark launched Genspark Clip Genius, which can automatically generate shorter clips from long videos based on your request.

5. Google news:

   • Android added real-time translation to Circle to Search, so the text keeps being translated as you scroll or even switch apps.

   • The new Androidify feature lets you turn a selfie into a custom bot avatar and animate it using Veo. (You’ll never guess what kind of music track they used for the launch trailer!)

   • Gboard on Android now incorporates AI-writing features that help you tweak tone, grammar, and style directly on the device.

   • Google Photos got a new Create tab with AI tools that restyle images, create collages, animate your photos, and more. (US-only for now.)

   • Google Sheets now lets you type “=AI” into any cell to request Gemini to summarize or categorize data or write new context-aware text.

   • NotebookLM now lets you pick Audio Overview formats and length, including a “debate” option that used to require a manual hack.

6. Higgsfield released Higgsfield Ads 2.0, a built-in “Mini App” that makes it easier to produce marketing images and video clips using templates.

7. Ideogram launched Styles, allowing you to apply preset style templates or customize your looks using reference images.

8. Mistral upgraded Le Chat with 20+ connectors to enterprise tools and a Memories feature for personalized chats.

9. NVIDIA launched an AI Blueprint for 3D objects that bulk generates up to 20 draft objects to include in a given scene.

10. OpenAI news:

    1. The Projects feature is now available to free users, with bigger file uploads, customization options, and project-specific memories.

    2. ChatGPT on the web now has conversation branching, so you can explore different directions without losing the original thread.

11. Snapchat launched Imagine Lens, which lets you create, edit, and remix Snaps using text prompts and share them with others.

12. Tencent news:

    1. HunyuanWorld‑Voyager turns a single image into an explorable 3D world you can navigate and export as a point-cloud video.

    2. Hunyuan-MT-7B is a 33-language translation model that beats bigger models on multiple benchmarks. (Try for free on Hugging Face.)

---

🔬 AI research

Cool stuff you might get to try one day:

1. OpenAI plans to strengthen teen safety in ChatGPT by auto-routing sensitive chats to the more robust reasoning models and introducing new parental controls.

2. WordPress teased Telex, an experimental AI tool that turns text prompts into ready-to-use blocks for its content editor, Gutenberg.

---

📖 AI resources

Helpful AI tools and stuff that teaches you about AI:

1. “AI prompt engineering in 2025: What works and what doesn’t” [VIDEO] - an info-packed Lenny’s Podcast episode with Sander Schulhoff.

2. “10 Secrets of AI Filmmaking You Need To Know!” [VIDEO] - a short and practical guide by Theoretically Media.

---

🔀 AI random

Other notable AI stories of the week:

1. Atlassian is buying The Browser Company to integrate the AI-powered Dia browser with its suite of products and turn it into a tool for knowledge workers.

2. OpenAI is working on an OpenAI Jobs Platform to connect job seekers with employers, including local businesses and government roles.

🤦‍♂️ AI fail of the week

I asked Nano Banana for a meme about buying unicorns on Temu.
This might actually be a win.

Send me your AI fail for a chance to be featured in an upcoming Sunday Rundown.

---

💰 Sunday Bonus #70: 72 use cases for Nano Banana (swipe file)

Today, I’m expanding my collection of interactive swipe files, since people seem to like them quite a bit. Here are the last three:

• 90+ use cases for GPT-4o image generation (swipe file)

• 45+ practical use cases for o3 (swipe file)

• 40+ use cases for Claude 4 (swipe file)

For the last two weeks, Nano Banana has been “KILLING PHOTOSHOP.”

There are dozens of videos and written roundups of possible use cases floating around.

So I decided to organize the findings in one place with the help of NotebookLM, OpenAI o3, and Genspark.

The result is an interactive swipe file with 72 Nano Banana use cases for inspiration:

Three sample use cases

As always, you can filter by category, search by keyword, and copy the starter prompts to your clipboard with a single click.

Enjoy!