Rowan Cheung Profile picture
Dec 6 10 tweets 4 min read Twitter logo Read on Twitter
Google just revealed Gemini and will directly integrate the AI into Google apps.

The GPT-4 competitor comes in 3 models — Ultra, Pro, and Nano.

Here's a thread of EVERYTHING you need to know: Image
Gemini is multimodal and can recognize images and speak in real-time.

With a score of 90%, Gemini Ultra is the FIRST AI model to outperform human experts on the MMLU benchmark.

This demo is incredible.
Gemini has next-generation capabilities such as sophisticated reasoning, multimodality, and advanced coding.

The model is also advanced in math and coding, as compared to ChatGPT (GPT-4), which cannot perform math.

Check out this demo of them solving physics.
Gemini has an incredible understanding of science.

It can find and extract research across 1000's of research papers.

Because Gemini is multimodal, it can not only understand text but also graphs through images!
Gemini comes in three sizes — Ultra for complex tasks, Pro for scaling across a range of tasks, and Nano for efficient on-device tasks.

-Pro will be in Google products through Bard starting today.
-Ultra will be rolling out early next year.
-Nano will be available on Pixel. Image
Gemini Ultra’s performance beats current state-of-the-art results in 30 of 32 benchmarks used in LLM research & development.
Image
Image
Gemini Pro will be available for free in Bard and across Google apps today.

In six out of eight benchmarks, Gemini Pro outperformed GPT-3.5, making it 'the most powerful free chatbot on the market today'. Image
Gemini Nano now powers on-device generative AI features for Pixel 8 Pro.

New features include:
-Summarize in Recorder
-Smart Reply in Gboard
-Cutting-edge video
-Enhanced photography and image editing
I shared all the info on Gemini in my newsletter this morning.

Click here to join 400k+ readers, and you'll never miss a thing in AI ever again: therundown.ai/subscribe
Thanks to @GoogleDeepMind for an invitation to the early press conference invite, allowing me to share the news live.

I do these rundowns daily, follow me @rowancheung
for more.

If you found this helpful, spare me a like/retweet to support my content 👇

• • •

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

Keep Current with Rowan Cheung

Rowan Cheung 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 @rowancheung

Dec 7
Today's AI developments were WILD.

-Google announces Gemini
-Meta reveals 20 new AI features
-Meta reveals AI image generator
-McDonald's new AI chatbot 'Ask Pickles'
-Alibaba's video AI scrapes TikTok data
-AI helps decode a new whale language

Here's what you need to know:
Google DeepMind revealed ChatGPT's biggest competitor, Google Gemini.

With a score of 90%, Gemini Ultra is the FIRST AI model to outperform human experts on the MMLU benchmark.
Gemini comes in 3 models — Ultra, Pro, and Nano.

Gemini Pro will be integrated into Bard today, while Gemini Ultra (the best model that beats GPT-4 in 30/32 benchmarks) will be avail early next year.

The thread below has everything you need to know:
Read 11 tweets
Dec 6
AI NEWS: Meta and IBM just teamed up to form the 'AI Alliance'.

Plus, huge developments from Microsoft, Elon Musk, xAI, Google DeepMind, Looking Glass, Pika Labs, Runway, Getty Images, AstraZeneca, Mistral, and 10 new AI tools.

Here's everything you need to know:
Meta, IBM, and over 50 other companies, universities, and organizations just announced the formation of the ‘AI Alliance’.

The mission of the alliance is to advance open and responsible AI innovation.
Microsoft just revealed powerful upgrades to Copilot.

The upgrades include a GPT-4 Turbo, improved DALLE-E image generation, and new features like Code Interpreter.

Future capabilities include multimodal Search to, Deep Search, and math analysis.
Read 13 tweets
Dec 4
Nature just interviewed leading researchers on how they use ChatGPT.

The article is a fascinating piece on how to use AI in your work, responsibly.

Here's how top academics are using ChatGPT in their research:
Mushtaq Bilal (@MushtaqBilalPhD) focuses on using ChatGPT to create structure rather than content.

Mushtaq uses it as a brainstorming partner and a tool for outlining research papers. Image
@MushtaqBilalPhD Ethan Mollick (@emollick) believes AI will transform teaching and emphasizes the need for ethical experimentation with AI in education.

Notably, he also incorporates AI into every teaching assignment and lesson.

Extremely jealous of his students! Image
Read 8 tweets
Dec 3
AI video is inching closer to Hollywood-level disruption.

This week, we saw HUGE upgrades to text-to-video, real-time generation, photo enhancements, in-video editing, and more.

Here's everything you need to know:
1. Pika Labs announced Pika 1.0.

Not only can you create videos from text, but you can expand video canvases (0:34) and edit video in real-time like clothing (0:41).
2. Stability AI released SDXL Turbo — a new model for REAL-TIME text-to-image generation.

The model creates images that update in real-time as users type prompts.
Read 10 tweets
Dec 1
Microsoft launched a free course for beginners to learn AI.

The 12-week, 24 lesson curriculum is all publicly available on GitHub!

Here's what you'll learn & links to the lessons: Image
Before we get into it, I'm also giving away my AI tool tutorial guide free right now to new newsletter readers.

25+ AI workflow tutorials, 1000+ best ChatGPT prompts, 100+ AI tools, and more

Click to subscribe and get it sent to your inbox: therundown.ai/subscribe
Image
1. Introduction and History of AI

Link: github.com/microsoft/AI-F…
Image
Read 13 tweets
Dec 1
It's official: Sam Altman just confirmed the new Q* model breakthrough was a leak.

Plus, huge developments in AI from Elon Musk, Microsoft, the UK, Ideogram, Stanford/USC, Mastercard, and 9 new AI tools.

Here's the rundown of everything you need to know:
In an exclusive interview with The Verge, Sam Altman confirmed Q* was leaked (meaning it's real).

Sam quotes: "No particular comment on that unfortunate leak." Image
OpenAI also published a letter announcing that Sam Altman is officially back as CEO alongside a new board.

Noticeably absent is Ilya Sutskever — with Altman saying they're ‘discussing how he can continue his work.’

Will we ever learn what Ilya saw?
Read 10 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 on Twitter!

:(