The AI Breakfast: Last Week's Summarized Updates
Image made by me via Mid Journey V4

The AI Breakfast: Last Week's Summarized Updates


Ground-zero for a bunch of massive AI releases.

This is all you need to read to get up to speed on AI

  1. ? Midjourney V5
  2. ? GPT4
  3. ? Microsoft Co-Pilot in 365
  4. ? Google Workspace AI
  5. ? Anthropic Claude




1. → Midjourney V5

GPT 3 is cool but you know what's more cooler.

  • AI finally generating hands perfectly. Midjourney v5 can do this now.
  • Plus the photorealism is so much better, it almost feels like real.


Example below: Midjourney presents historic sneaker collections:

No alt text provided for this image
Midjourney presents historic sneaker collections: Queen Elizabeth

Queen Elizabeth



No alt text provided for this image
Midjourney presents historic sneaker collections: Mansa Musa

Mansa Musa




No alt text provided for this image
Midjourney presents historic sneaker collections: Ghengis Khan

Ghengis Khan




No alt text provided for this image
Midjourney presents historic sneaker collections: Da Vinci

Da Vinci






2. → GPT4

No alt text provided for this image
GPT-4 outperforms the pervious SOTA AI model and humans in many standard exams.


GPT-4 is a large multimodal model that can accept image and text inputs and emits text outputs.

  • Data Analysis on the fly.
  • Provide the image/text input or both and GPT-4 can answer all your questions just like an human analyst.
  • Research Made easy with OCR and Summarization on the fly
  • Just provide the PDF and ask your question. GPT-4 will provide you with pin-pointed answers!
  • Unlike any AI model, follows human instructions very closely!!
  • Full control over customizing user experience to take it to the next level
  • GPT-4 can understand humour and sarcasm which so far has been unique to humans.
  • Now you can ask the AI if your joke is funny or not


3 things GPT4 can do which blew my mind:

1. Recreate the "Breakout" nostalgic game in less than 60 seconds!

#1 try. No debugging.


2. Code apps and sites in minutes.

Prompt: hey gpt4, make me an iPhone app that recommends 5 new movies every day + trailers + where to watch.


3. Can understand images better than humans.

Here's how I gave GPT-4 a photo of a refrigerator and asked it to come up with food recipes in under 60 seconds.

No alt text provided for this image
how I gave GPT-4 a photo of a refrigerator


No alt text provided for this image
asked it to come up with food recipes in under 60 seconds.





3. → Microsoft Co-Pilot in 365

"we're giving people more agency and making technology more accessible with advanced AI and the most universal user interface: natural language." - Satya Nadella


"Copilot in Microsoft 365 is a big deal" - Greg Brockman

Microsoft copilot: gpt4 in powerpoint, excel, word, outlook, teams, more.

This adds instant competitive pressure to at least a hundred ai startups:



Copilot, make Microsoft Word actually function and not a complete waste of space."


The biggest difference for startups with these new LLM's vs. previous ML improvements is the models here are SO easy to use that they likely benefit incumbents (who have existing distribution) vs. new entrants.

Anyone with existing customer eyeballs can/will be able to integrate this technology with some simple API calls + a UI wrapper for their existing products.


Feels like a big win for the big guys.





4. → Google Workspace AI

It quickly got overshadowed by the GPT4 release, but Google basically released GenAI into all there workspace products. Google just launched Generative AI across ALL of Google Workspace -- Gmail, Docs, Sheets, Slides, Images -- EVERYTHING.

Google hits back with a full integration of AI across Google Workspace.

Includes generating images, presentations, emails, docs and more all within Google Workspace.

According to Google, here are just a few of the MANY things you will be able to do with their new AI:


  1. Draft, reply, summarize, and prioritize your Gmail
  2. Brainstorm, proofread, write, and rewrite in Docs.
  3. Bring your creative vision to life with auto-generated images, audio, and video in Slides Generate new backgrounds and capture notes in Meet
  4. Enable workflows for getting things done in Chat
  5. Go from raw data to insights and analysis via auto completion, formula generation, and contextual categorization in Sheets.

This is a BIG response to ChatGPT and Microsoft -- one I've been waiting to see for weeks. But it seems like it was worth the wait.


Will you be using Google's new AI tools in your workflows?





5. → Anthropic Claude - launches to its ChatGPT's rival, 'Claude'

Anthropic has opened up access to Claude, their AI assistant, to power businesses at scale.

Claude is based on Anthropic's research into training helpful, honest, and harmless AI systems and is capable of a wide variety of conversational and text processing tasks.


Users describe Claude’s answers as detailed and easily understood, and they like that exchanges feel like natural conversation,” says Autumn Besselman from Quora.

“Claude feels more conversational than ChatGPT,” says a happy Poe user, while another says, “I find Claude to be more interactive and creative in its storytelling.”

Early customers report that Claude is much less likely to produce harmful outputs, easier to converse with, and more steerable. Claude is being used by partners such as Quora, DuckDuckGo, and AssemblyAI to power their products and services.

You can read more in our full post here: Anthropic Claude

要查看或添加评论,请登录

Brian Shem ??的更多文章

社区洞察

其他会员也浏览了