🤌 GPT-4 and Midjourney V5 are here!

PLUS: Microsoft launches new AI features

Yo what’s up?

Welcome to the 400+ new subscribers who have joined since the last edition. I'm so excited. Are you? Last week was the most action-packed week in AI since 2014.

What's cooking in today's edition? 🧪

  • GPT-4 is here!

  • Midjourney V5 has launched

  • Microsoft lanches brand new AI features

👋 Psst! New here? Subscribe to receive the top AI news every Friday + exclusive content. Follow us on Twitter!


GPT-4 launched last week and has lived up to the hype. It’s the latest model from OpenAI and is GPT-3.5’s successor, the ChatGPT we all love. Here's a breakdown and what's important:

1) GPT-4 is a much more “powerful” model than the existing version of ChatGPT. What does this mean?

It has better reasoning, is more creative, and outputs far more relevant answers. In fact, It scores 90% or better on many tests and exams.

2) GPT-4 is multimodal, which means it can take in text and images.

This is big. GPT-4, being multimodal, unlocks a lot of possibilities. You can feed images and ask questions about them.

3) GPT-4 can handle a lot more text - 25K words!

You could input around 3000 words to ChatGPT before you hit the limit, which has been significantly increased with GPT-4 ~ 25K words. This means you can feed several documents and papers to GPT-4 and ask questions. You don’t need even need to train embeddings!

Here’s an example of a VC collecting all the data of a startup (emails, personal notes, startup deck) and creating an executive summary. Wild!

Here's how you can try GPT-4 now.

  1. Sign up for ChatGPT Plus for $20/mo.

  2. Try for free at Spellbook by Scale AI 

Fun fact. GPT4 in french means I farted loudly


The latest version of Midjourney launched last week and it's outstanding. It’s the best AI art model out there and improves vastly on a lot of issues

  1. It creates very photorealistic images

  2. Hands are fixed

  3. You can add a lot more details

I don’t hype stuff like this, but folks, I've been in AI Art since 2020, and this is very hard to pull off. It’s the best AI Art model out there by a wide margin


We all knew this was coming, and it’s here. You can use powerful LLMs in Excel, Word, Powerpoint, and their entire suite of apps!

Check out this tweet, where I cover more info with video examples.


  • Stanford Alpaca allows yout to run ChatGPT in your macbook (link)

  • Adept get a $350m funding round for their AI Assistant (link)

  • Assembly AI introduces Conformer-1 API, the state-of-the-art speech recognition model (link)

  • Together release OpenChatKit, their open-source version of ChatGPT (link)

  • PwC launches an AI Chatbot for 4,000 lawyers to accelerate work (link)

  • Duolingo launches a new subscription tier with an AI tutor powered by GPT-4 (link)


  • Collie - Creates a delightful search experience in a snap.

  • Collato - AI search for product teams.

  • Chatfuel AI - Takes customer communication to the next level.

  • AI Query - Generates SQL Queries with AI in seconds.

  • CreatorKit - A video creator for ecommerce and product video ads.

  • WatermarkRemover.io - Removes watermarks from images using robust AI technology.

  • Cursor - An IDE built for programming alongside GPT-4.

  • Ask your PDF - Upload a PDF to chat with it.

  • Klarity - Chat with your document and discover detailed insights.

  • Double - Start using GPT to automatically research your leads on the internet and answer questions.

  • KissanGPT - A ChatGPT and Whisper-based assistant for underserved agriculture domain of India.


"Elephant Kittens"

That's a wrap for today. Stay curious & see ya tomorrow! If you want more, be sure to follow on Twitter (@sudu_cb)

What did you think of today's edition? Reply to the email with 

  • 🔥🔥🔥 - This is f*cking awesome

  • 🔥🔥 - Meh, could be better

  • 🔥 - I want a refund.


Both of them are AI-generated! That's how wild Midjourney is. Image 2 was created by Nick Floats.