Google's controversial "Best Take" AI feature

PLUS: How to generate 3D scenes with text prompts


Good morning, human brains. Welcome back to your daily munch of AI news.

Here’s what’s on the menu today:

  • Is Google’s new AI ill or chill? 🤮 😎

    Its new features have tech reviews concerned about potential misuse.

  • Here’s why NASA-funded AI space lasers 🚀 💥

    To develop space lasers that divert space objects from collision paths.

  • Easily generate entire 3D scenes with text prompts 📝 🪩

    3D-GPT uses LLMs to create models and works with popular software.


Google Pixel’s new AI: ick or sick? 🤮 😎

Last week, we covered Google Batch Calibration. It’s a method to leverage an AI model’s pre-existing knowledge to learn new, unrelated tasks.

On Saturday, Google Pixel’s face-altering AI freaked people outand not in a hot way.

Here’s us playing with “Best Take:”

I have an iPhone like everyone else, what’s up with the Pixel?

Google added AI features to the Pixel and to the Google Photos library.

The “Best Take” feature scans photos to replace someone’s expression, like adding a smile from another picture.

Another feature is the “Magic Editor” which you can use to remove, shift, or resize elements in a photo.

Sounds good, why wouldn’t people like this?

Multiple tech reviewers expressed concerns about the potential misuse and ethical implications of these AI tools.

“Icky? A little bit.” - The Verge.

“Slightly creepy.” - Techradar.

“More ominously, it can pose serious threats to people’s (already fragile) trust of online content.” - CNET

Tinkering with “Magic Editor:”


Get a website designed by AI that doesn’t look like it was designed by AI. In 30 seconds or less.

Durable is the AI website builder and platform that makes running a business easier than having a job.

Build a fully functioning, beautiful website in 30 seconds.

Then, start attracting and booking more clients with integrated CRM, SEO, invoicing, and marketing tools.

Try it for free, no credit card is required.



They help machines learn from data, so they can make decisions or predictions on their own. It’s like a recipe that tells a computer exactly what steps to follow to solve a problem or complete a task.


NASA-funded AI space lasers? 🚀 💥

In August, we reported on HelioLinc3D discovering an asteroid. It’s the first “potentially hazardous“ asteroid discovered by AI.

A week later, we covered WJ-1A’s launch into space. It’s an AI-powered satellite with real-time observation and data processing capabilities.

More space stuff?

Yes. Earlier this month, NASA funded the development of AI-powered space lasers. It gave $200,000 over three years to the cause.

‘Murica 🇺🇸

Space lasers, really?

You bet, champ. Researchers propose using a network of lasers to divert space debris from potential collision paths with satellites, the International Space Station, and more.

How do they work?

The lasers will use AI to monitor space debris and decide which objects to target, ensuring they are redirected to safe orbits.

Who cares about space debris?

It’s actually a big problem. Around 34,600 pieces get tracked by Earth’s radar systems, but an estimated 130 million smaller fragments orbit the Earth undetected.

Last week, Amazon postponed its two prototype satellites to avoid a space object. India also adjusted its moon mission’s launch for similar reasons in July.

Nobody’s thought of space lasers before now?

We currently use ground-based lasers, but space-based ones are more accurate and efficient because they don’t have to navigate through Earth’s atmosphere.

Next? Sharks with laser beams attached to their heads. 🦈 🚀 


Generate full 3D scenes with text 📝 🪩

Last week, we reported on NVIDIA and Masterpiece Studio’s Masterpiece X. It’s a text-to-3D animation playground.

On Thursday, researchers published a paper on 3D-GPT. It’s a tool for using LLMs to create 3D models. It works with Blender, a popular 3D modeling software.

Is the metaverse making a comeback?

How does it work?

3D-GPT simplifies 3D modeling by breaking it into smaller tasks and deciding which part of the system should handle each task.

It has three main parts:

  • Task dispatch agent: Decides which part of the system should handle a task.

  • Conceptualization agent: Helps enhance and evolve initial scene descriptions.

  • Modeling agent: Works with 3D software to create the actual models.

So it’s better than Masterpiece X?

Eh, it depends. It can’t manage certain advanced features of 3D software, making tasks like adjusting tree branches or blending texture colors challenging.

Its algorithms also restrict its capabilities in some categories like creating hair, fur, and more.



Think Pieces

Want better LLM outputs? TRY SCREAMING IN ALL CAPS.

How does AI disrupt business models? Which business models will be helped by AI and which ones will be hindered?

Startup News

Apple added several AI-related job listings to its career page. This month, it added jobs for a conversation AI platform, AI developers, and more.

Microsoft launched its Early Access Program for Security Copilot. It aims to combat AI’s surge of cyber threats.


DreamSpace — a method for personalized VR indoor space design through text prompts.

SILC — a framework that enhances vision-language models (VLMs) for various computer vision tasks.

SALMONN — a multimodal LLMS that processes general audio inputs like speech, audio events, music, and more.


Think Diffusion — access your Stable Diffusion projects from any device.

Infr — record all your experiences in a database and interact with the data.

Voyager — Spotify tool on GitHub for approximate nearest-neighbor searches.

Conveyor — AI-powered security questionnaires for your business.


If you like Bot Eat Brain there’s a good chance you’ll like this newsletter too:

The AI RoadYour go-to source for the latest news and insights on artificial intelligence and its impact on the world around us.


Story time.

Ethan Mollick, a renowned Wharton University Professor, humorously mentioned he didn’t like people using DALL-E 3 to portray AI as a monstrous being.

He then said that he prefers it in the style of Byzantine mosaics.

A follower of his then presented an AI art generation combining both styles into this captivating, horrifying representation:

Tag us on Twitter @BotEatBrain for a chance to be featured here tomorrow.


Until next time 🤖😋🧠

What'd you think of today's newsletter?

Login or Subscribe to participate in polls.