WH #9:🚨GPT-4 is solving blindness, AI is about to be in everyone's hands

This week's is a GOOD one, you wont't want to miss it

March 20, 2023

In this week’s update:

  • 🚨GPT-4 is out!

  • 📸Midjourney will kill social media

  • 👨‍💻Google + Microsoft is about to change how we all work forever

-Howard Huang

I need your help

This newsletter hasn’t added a new subscriber for the 3 straight weeks. If you know someone that might like what I have to say please forward it along and share it.

What I read this week

It’s been a whirlwind in the world of AI. My eyes are cross-eyed from all of the new products and services that is being built. We’ve opened pandora’s box and it’s never going back in. Here’s what happened just last week.

OpenAI

The rumors were true, GPT-4 launched on March 14th and is now available for ChatGPT Plus subscribers. Here is what it can do that it couldn’t do before:

  • A LOT more creative – can collaborate on creative and technical writing tasks, compose songs, screenplays, or learn a user’s writing style (the last one is HUGE for SEO content). How HUGE you ask? Well, huge enough that I wrote a blog about How I’m using GPT-4 for SEO Content Creation.

  • Understands images – GPT-4 accepts images as inputs and generate captions, classify, and conduct analyses. Um what? Take a photo of your pantry and let it decide on what dishes you can make. But that’s just small potatoes to what Be My Eyes is doing. It’s an app that allows the blind to use their smartphone camera to point to things, GPT-4 will provide instantaneous audible identification, interpretation, and conversational assistance. 🤯

  • Longer context – Handle over 25,000 words of text. Ever want cliff notes for a research paper? Just dump the entire thing and let GPT-4 summarize it for you.

  • GPT-4 outscores practically every standardized text. Isn’t this scary? Isn’t this version 1 of Artificial General Intelligence, where a machine is able to perform any intelligent task that a human being can do? 🤯

Midjourney

Midjourney is the best text to image AI generator (in my honest opinion) and on March 16th, they released version 5 which greatly improves on photo realism. There were criticisms on too many additional teeth, fingers, and feet, but it appears to be *mostly* fixed. Here are some images that I generated…just a friendly reminder, these are all AI generated and never existed before in the world until my prompt came along.

Are those photos not 🤯 incredible? Social media is in trouble. In a few years (perhaps much sooner) no one will be able to tell what is real and fake. Actually, it’s happening NOW! I started a new instagram account where I ask whether an image is real or AI and half the time people cannot tell!

Google vs Microsoft

These two tech giants are BOTH doing full AI integrations across their entire office work suites. Google and Microsoft together powers more than 95% of all office related software. How we all work at the office will be changed forever.

Google

On March 14th, Google announced on their blog that they are doing a full integration of AI across Google Workspace. You’ll get text to image generation, slides, gmail, google meetings, and etc.

Microsoft

On March 16th (2 days after Google😉), Microsoft announced on their blog that they are doing a full integration of AI across Microsoft 365 called “Microsoft 365 Copilot.” You’ll get text to image generations, powerpoint, outlook, etc etc…you know…everything Google’s doing except Microsoft. 🤣

What I watched this week

Microsoft’s announcement of Microsoft 365 Copilot

OpenAI’s Live Demo on GPT-4

What I did this week

Quote of the week

"We don't read and write poetry because it's cute. We read and write poetry because we are members of the human race. And the human race is filled with passion."

– John Keating, Dead Poets Society

How many 🤯 did I use this week? Leave a comment. Wrong answers only. 😊