OpenAI founder feels awful about ‘significant issue’

🪲 ChatGPT Bug Exposed People’s Conversation History

🎞️ Gen-2 by Runway – huge step in AI with text-to-video

🖼️ Bing launches Image Creator – a text-to-image engine

🪰 Adobe Firefly launches its beta AI tool to quickly modify images

ChatGPT Bug Exposed People’s Conversation History (CNET)

What you need to know:

On Wednesday, OpenAI CEO Sam Altman said the “significant issue in ChatGPT” was the result of a “bug in an open source library”.

• The bug exposed titles of some users’ past conversations to other users.
• Open AI disabled ChatGPT to fix the issue on Monday.
• It impacted a small percentage of users whose chat histories from 1am to 10am PDT will not be accessible.

This serves as a reminder to all of us to be cautions when using AI tools, especially during beta/testing phases.

Gen-2 by Runway – The Next Step Forward for Generative AI

What you need to know:

• On Monday, AI startup Runway inched closer to text-to-video AI when they announced Gen-2.
• The technology aims to generate videos based on written descriptions.
• Runway already offers a web-based video editor with AI tools, and helped develop open-source text-to-image model Stable Diffusion.
• Gen-1 focsued on transforming existing video footage.
• Gen-2 focuses on generative videos from scratch (with some limitations).

Be sure to join Runway’s waitlist to gain early access.

Bing launches Bing Image Creator tool

Bing launched its Image Creator tool powered by OpenAI’s DALL-E model. It enables users to create images with text descriptions.

What you need to know:

• Bing Image Creator will be fully integrated, initially in creative mode, allowing users to type a description, context, and art style for their desired image.
• Bing will integrate its Image Creator into Microsoft’s Edge browser making it the first and only browser with an integrated AI-powered image generator.

You can try it here with a Microsoft account. Happy creating!

Adobe Firefly

On Tuesday, Adobe launched Firefly, a tool that will let users type commands to quickly modify images.

What you need to know:

• The first model focuses on image creation and text effects
• “Generate Variations” option creates different versions of an element
• Firefly can transform a scene with simple commands (e.g. summer to winter) — see their video below!
• Generative AI used to create paintbrushes based on existing image elements
• Firefly to launch as private beta
• Adobe aims to protect and benefit creators with Firefly

Cool AI stuff

Fliki Tweet-to-Video – Turn your Tweet into a video to share as TikTok, Instagram Reel, Youtube Short and more.

Reachout.ai – Create customized talking-head videos at scale.

Mails.ai – an automated AI-driven email platform automating email campaigns and more.

Have a cool AI tool to share? Send it to us on Instagram, Twitter, or email and we’ll consider it for an upcoming publication!

Featured AI art of the day

This image prompt was inspired by a prompt found on Midjourney’s Showcase. Check out some of the coolest imagery being created on the Internet now.

Prompt: quirky red-headed teen boy in the 70’s in his bedroom. There are old posters on the wall. He is working on his old computer and turning to look over his shoulder at the person with the camera taking the picture. There are old disks on the desk and various items and toys from the 70s (rubicks cube, etc). Reportage shot. Canon with Kodak film. Telephoto –v 5 –chaos 100 –v 5

Want your art featured? Dm it to us on Instagram, Twitter, or email and we’ll consider it for an upcoming blog!