EyeingAI Profile picture
Aug 13, 2024 12 tweets 4 min read Read on X
R.I.P Apple Intelligence!

Just minutes ago, the Made by Google event revealed the new Pixel 9 phones.

10 features you didn't think were possible: Image
1. The World’s First AI-Powered Camera Experience

Experience enhanced HDR+ with more accurate skin tones, textures, shadows, and more, capturing the world around you authentically.
2. Add Me

Add Me ensures no one is left out of a photo. Take one with the group, then swap in the photographer. Pixel seamlessly merges both photos.
3. Pixel Studio

Create anything you imagine with PixelStudio, a groundbreaking image generator powered by an on-device diffusion model. It’s your AI canvas.
4. Zoom Enhance

Get closer even after zooming in. AI enhances details and quality, making the photo look sharp.
5. Cutting-edge 8K video

Video Boost upgrades your video to 8K resolution, enhancing color, lighting, and stabilization for the best quality video on Pixel.
6. Move anything, anywhere.

Reposition people or objects within the frame to align your photos with your vision.
7. Never Forget a Thing With Pixel Screenshots

This new AI-powered feature helps you instantly find anything you've captured in a screenshot. From Wi-Fi passwords to must-try restaurants, it's all at your fingertips.
8. Call Assist

Let Pixel CallAssist track conversations, providing a private summary with easy access to phone numbers, times, and other important info.
9. Pixel Weather

Stay informed with super accurate forecasts and weather summaries. Customize your experience with interchangeable widgets for UV, wind, humidity, and air quality.
10. The First Phone to Launch with Gemini Nano & Multimodality.

This on-device AI model is three times more capable and sophisticated than the one launched last year. With it, Pixel 9 can generate an industry-leading mobile output rate of up to 45 tokens per second.
And that's a wrap!

Found the thread valuable?

→ Repost and Like the tweet below to share with others

→ Follow me @EyeingAI for interesting AI content every day

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with EyeingAI

EyeingAI Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @EyeingAI

Aug 19
WTF… While GPT‑5 disappointed everyone..

China quietly dropped the most advanced AI agent system we’ve seen yet.

– Fully autonomous
– 100+ agents in parallel
– Interrupt tasks mid-run 🔥
– Real-time memory + reasoning

Silicon Valley can't afford to ignore this: 🧵 Image
Okay, so let's be real.

GPT-5 was a letdown.

There was no "holy sh*t" leap.

And the AI game has shifted.

It’s no longer about “bigger models.” Image
What’s exploding instead?

A new AI race: not models… but agents.

And GenFlow 2.0 by Baidu Wenku might be the first real beast in the wild.

✅ Stop/tweak tasks mid-run
✅ 100+ agents in parallel
✅ Structured + unstructured data
✅ Markdown + charts
✅ Real-time edits in complex chains

Basically, what you wish ChatGPT could do.
Read 9 tweets
Aug 16
WTH... Everyone’s watching OpenAI, Anthropic, and Google.

Meanwhile, a quiet startup just built the operating system for your brain.

– reads & understands anything you give it
– creates exactly what you need
– wipes out most of the work you do

And nobody’s talking about it: 🧵
I’ve tried them all: chatbots, copilots, mind maps, note apps.

They nail one thing… then fall apart when you try to do it all.

@kuseHQ is different.

It's like an endless desk where you drop PDFs, videos, links, ideas & make sense of them instantly.

app.kuse.ai
One of the first things I tried in Kuse was dropping in a YouTube video.
(with the DeepSeek model).

From just one video, I made:

→ A full summary
→ A mind map
→ A LinkedIn post with key points
→ Crypto trends analysis charts
→ An Instagram carousel

WILD.
Read 10 tweets
Aug 13
WTF... This scraper works like a self-driving car.

You give it a destination… it plans the route, clicks through pages, cleans it up & delivers a perfect spreadsheet.

Goodbye manual scraping, Excel & data teams.

Tutorial + some crazy examples: 🧵
With @sheet0ai, you just tell it what you need→ it does the clicks, the cleanup, the structuring. Everything.

No coding.
No formulas.
No Chrome extensions.
No regex tutorials at 2AM.
No more endless copy-paste.

bit.ly/3HkWkHW
They call it an “L4 Data Agent”... borrowing from self-driving cars.

You give it a task and it scrapes, cleans and delivers perfect data.

In self-driving cars, Level 4 means you set the destination and the car drives itself.

So if Tesla built a scraper, this would be it. 🔥 Image
Read 10 tweets
Aug 4
Their early demo, Riffusion, went viral.

Millions of us heard AI‑made music for the first time.

Now…they’ve built the thing the music industry has feared for years.

– Any idea → full track in mins
– Edit & refine in real time
– All genres

This is bigger than you think 🧵
I didn’t try @producer_ai for another “AI music experiment.”

I wanted to see if Studio Mode could really replace sitting in a studio with a producer.

Could it take a half‑formed idea & turn it into a full track I’d actually keep?

First try, nailed it.

producer.ai
Instead of typing a prompt & hoping for the best…
you chat with it.

Tell it: “I want a moody synthwave track, slow tempo, with a pulsing bassline.”

It builds the base track.
Then you go back & forth refining it.

Record, edit, export. All in one place.
Read 10 tweets
Aug 1
WTF... First they said no to Elon Musk.

Then they built an AI that just shook the entire LLM industry.

– 27M parameters
– Trained on only 1000 examples
– Inspired by the human brain
– Beat Claude, o3-mini & Deepseek

You can’t afford to ignore this: 🧵 Image
It’s called the Hierarchical Reasoning Model.

HRM = 2 minds in sync

High-level → slow, strategic reasoning
Low-level → fast, detailed execution

It solves problems in one forward pass (no chain-of-thought needed).

Instead of guessing every step like CoT, HRM learns to reason internally.
Kind of like how your brain does it quietly before you speak.
Now the fun part: performance.

Despite being tiny (just 27M params), HRM beats some of the biggest models on hard reasoning tasks:

✅ ARC-AGI-1
✅ ARC-AGI-2
✅ Extreme Sudoku
✅ Maze navigation

With just 1000 training examples. No pretraining. Image
Read 9 tweets
Jul 25
It’s over for Gemini & ChatGPT.

The first AI with unlimited video memory just dropped. This beast can:

- Watch 100+ hour videos 👀
- Understand audio + visuals
- Remember every second
- Answer anything about it

Wild features + how to try: ↓ Image
Most AI can create videos.
But ask them what happened 5 mins ago? They’re clueless.

@memories_ai fixes that.

It watches, understands & remembers everything. It can process and remember hours to years of video without losing context.

Insane, isn't it?

memories.ai
1. Video Chat with Real Memory (Like NotebookLM, but for video)

Tools like Gemini forget what they saw.
You re-upload. Re-process. Repeat. Every time.

Memories ai skips all that.
Upload once → Ask anything, anytime.

Even across multiple videos. 🔥
Read 9 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us!

:(