Holy Cannoli
It's been a pretty big week for AI News. Three new tools and AI Movie tips.
AI Text Is Almost Here?
You can view it as one of the last sticking points for the AI naysayers, or the firewall that protected designers, but up until this week it was still correct to say “AI can’t do text”. Enter Ideogram, (it's pronounced eye-diogram) a high powered team of AI experts who have made more than just another image generator. While it still makes mistakes about 20% of the time, I was really impressed with how quickly it could pump out mockups of signs, posters, logos and other concepts. I could usually get a great proof of concept within a few re-rolls. Not good enough for a final draft, but an amazing step forward as a brainstorming tool. I’m really impressed and it’s a lot of fun. Read about it: https://ideogram.ai/launch
ModyFi has Text, and Layers and Masks and Blend Modes, Oh My!
With the exception of Photoshop Beta, most image generators are just that…and flat, without layers. ModyFi is a browser based image editor, like a mini Photoshop that allows you to quickly make a rough comp or collage of your idea by generating an image layer, uploading an image, or searching for a stock image. You can than delete backgrounds, add masks, add text or blend modes. Once you have a rough draft assembled, you can type type another text prompt to stylize the while thing and make additional variations. It’s a lot of fun! https://www.modyfi.com/. I’m just getting started and it’s really impressive.
Decoherence Has a Timeline and Layers, Superior to Kaiber
You’ve probably seen something made with Kaiber by now…that trippy, psychedelic look where people grow extra faces? It’s based on Stable Diffusion and there are just not enough controls or settings to get something different out of it. Decoherence, on the other hand, allows you to arrange clips and layers like a video editor, fine tune your prompts, and give you more control over your finished product. I’ve been using it everyday and loving it.
Board of Innovation
I got a lot of value out of the BOA livestream last week about how they’re running an AI Powered Innovation Sprint. You could watch the archive here, see a sample schedule here, or you can check out the list of tools here.
Chat GPT powered Brainstorming Within Figma?
You know how it is…you’re generating ideas with your team, but you’re too lazy to open up another browser tab and launch Chat GPT. Wouldn’t it be cool if you had an autonomous agent built into your FigJam file? Jambot has you covered and (once again) this link should help you look like the smartest person on your team: https://www.figma.com/blog/give-ideas-more-space-with-jambot/
Jambot brings the generative power of ChatGPT to your FigJam files, letting you ideate, summarize, and play.
From Idea to Video with Anthropic’s Claude
I’ve been doing a lot of experimental filmmaking this summer (for lack of a better term) and mixing and matching a lot of tools. I gained a lot of insight from this article where Claude wrote the prompt which was then generated in RunwayML.
The ChatGPT to Midjourney to RunwayML Workflow
As always, Chris Heidorn is ahead of the curve on this one an d provides and excellent walkthrough of how to make a cinematic movie trailer with ChatGPT, Midjourney, and RunwayML.
Hope your final weeks of summer are going well. Don’t let the heat or the smoke keep you inside for too long. As always, if you know someone who wants to learn more about AI and AI tools, please forward this newsletter.