AI Insider No. 8

Hello and welcome to AI Insider, the periodic (although lately it’s been weekly!) newsletter that keeps you updated on a few of the latest developments and trends in the world of artificial intelligence. Got some time to spare this weekend? Settle down with your favorite beverage and read on.

Print? In the Digital AI Age?

Generate a determined and fearless portrait of Ida B. Wells, the investigative journalist and early civil rights leader. Portray her in a newsroom setting, holding a newspaper, to emphasize her role in highlighting racial and social injustices.(Midjourney)

You may have heard of Midjourney, the AI tool that generates eye-popping images of people and things from text prompts. Since its launch, users have created millions of images (estimated) via Midjourney. Some of them are truly spectacular. Some make headlines, such as the one of the Pope in a puffer coat.

Midjourney, the company, describes itself as “a small self-funded team focused on design, human infrastructure, and AI.” So, it’s not surprising that a group of designers that brought us a magical digital image maker, is now pushing out a print product, Midjourney Magazine.

The content? A curated gallery of the top-rated Midjourney produced images and interviews with some of the creators of this AI powered imagery.

Sounds simple enough, but a writer for Wired called Midjourney Magazine “soulless.” (Yes, really.)

When I plunked down the $4 a month subscription fee, I was told to expect to receive the latest edition around the end of June. Check back for an update. — Michelle Johnson

Picture This: More Cool Updates from the World of AI Art

This week, we’ve got some cool updates from two big names in the world of AI art: Stable Diffusion and Midjourney. Let’s dive in and see what’s new in their paint boxes!

Hey, Look Ma, I Can Draw Hands!

Stable Diffusion, a top-rated AI art-maker, is launching a shiny new version, called the SDXL 0.9. What’s cool about it? They claim to have leveled up their hand-drawing skills. AI image makers have struggled with creating hands, many times generating six or more digits on one hand. But this latest update is all about making hands look handier.

And it’s not just hands! They’ve also ramped up their “image and composition detail,” which is a fancy way of saying that their pictures are getting even more detailed. To prove it, they’ve shown off some epic sample images, like aliens, a wolf, and even a person holding a coffee cup. — ChatGPT

How many fingers am I holding up? (Midjourney)

Midjourney’s Not Exactly Middling

Midjourney has just launched an update, 5.2, that’s packed with all sorts of fun new tools for making AI art.

Ever wished you could zoom out on a picture and fill in the extra space? Well, with Midjourney’s new Zoom Out (aka “Outpainting”) feature, now you can! It’s like getting an art expansion pack. (Note: Adobe (and others) beat them to the punch on this with their Firefly product, so this is catchup. — MJ)

If you’re the indecisive type, you’re going to love the improved “Remix” feature. Now you can mash up different AI-generated images into one cool composition.

Midjourney 5.2 also has a “Better NLP” feature, which basically means the AI has gotten better at understanding what you want when you give it a prompt. And to top it all off, they’ve added some nifty mini tools to help you get even more creative with your images​​. — ChatGPT


Can Chatbots Add a Compassionate
Touch to Medical Practice?

AI chatbots, like OpenAI’s ChatGPT, are lending a (digital) hand to doctors, helping them share tough news with patients​.

Fun fact: A study showed that when it comes to patient questions, ChatGPT’s responses were not just higher quality, but also way more empathetic than human docs. We’re talking “seven times more empathetic” kind of more! Looks like our AI friend has got some serious people skills​​. And get this, in nearly 80% of the study’s scenarios, medical experts actually preferred the AI chatbot’s responses over a physician’s.

If that’s not enough, ChatGPT has already aced the US Medical Licensing Exam.

But hey, let’s not forget, even tech whiz kids make mistakes. AI chatbots might slip up or misdiagnose sometimes, reminding us all that while they’re cool and helpful, they’re not quite ready to replace human docs just yet​​. — ChatGPT

Google’s New Shopping Experience

Google recently unveiled an AI tool to revolutionize online shopping. You can now virtually “try on” tops from hundreds of brands on models of a wide array of sizes and skin colors. And voila! You’ll see how clothes might drape on your own body. Just imagine, no more guesswork about how that cute top from Anthropologie might look on you!👚 Learn more about Google’s tool here. — Bing

Fun Stuff to Try – Compiled by Michelle Johnson

It’s the weekend. Stop the doom scrolling. Instead, let Pikazo – AI Art generator get your creative juices flowing. This app, available in the iOS app store (Sorry, Android users), lets you transform your existing photos or shoot a photo and turn it into an art piece. It’s easy to use. No text prompts required. Pick a photo, tap a style and you’re done. It’s mostly free, but they will try to upsell you things like prints and larger size images.

Starry, Starry Bot, generated with Pikazo.

There’s an old saying in journalism: If your mother says she loves you, check it out. Fact-checking is never more necessary in the age of AI generated content. FactiSearch can check that AI generated content for you.

Having a hard time keeping up with all of these newfangled AI applications? Here’s a curated list, organized by categories such as writing, finance, marketing and more: The Rundown.

Aht Gallery
(Yes, it’s spelled right! This is Boston!)

A very detailed image of an elderly black woman standing on a city street during the golden hour, sunset, taken with a Canon EOS R, using 85mm lens, taken on kodak Ektar 100 film, dramatic back lit portrait. (Midjourney)
Black and white photograph of a black couple elegantly dressed on a street in Harlem in the style of James Van Der Zee. (Midjourney)
Double exposure, silhouette of a woman and a flower. (Dream Studio)
Stable Diffusion’s update produces more detail. (Stable Diffusion Handout)

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.