Leonardo's Live Canvas is a Game Changer
Mind blown
So I realize that using the term “game changer” over and over defeats the purpose of using a superlative, but we are living in unprecedented times. I spend a couple hours per day, every day, reading articles and watching videos about this stuff, and I naively assumed that I was on top of the emerging trends. And then I learned about Leonardo.ai’s “Live Canvas” and my brain almost exploded. This is a real game changer. Last week I reviewed Stable Diffusion Turbo and that feels like a toy in comparison. Just check out my screen capture below. You can play with it here without spending any money. This just in, Krea.ai has a similar tool, but they charge $30/month and I’m still on the waitlist. So that’s three tools in 1 week with instant AI image generation.
Detroit Talents: The AI Artist Agency (from Paris)
Occasionally, I still get negative comments on social media from people who claim that AI is bad because it steals from “real artists”. Well, you can be outraged all you want, but apparently “AI Artist” is a thing now, and there is a talent agency in Paris who represents them. Apparently, people using the latest technology to create cool images has a long history and lot of momentum. I remember when people were outraged by Photoshop in 1990 and Digital Cameras in 2000. The typical payout from stock photo sites is $0.87 per image and the typical project an Fiverr is about $50 per image. I’m guessing the payout here is a little higher. Hat tip to SCCA student Brooklyn Schugar for sharing this resource.
Detroit is an Ai-Artists agency based in Paris. We represent a curated selection of international talents, primarily working in the fields of communication, press, and culture. Our team is at the forefront of the latest technological advancements and industry trends. You can rely on our expertise to create visuals that meet the highest standards and stand out from the competition.
Google Offers 12 Days of Generative AI Training
Yes, here I am with yet another mention of free AI training from one of the big corporate monoliths. Are you unemployed, under employed, or pretty certain you’re going to be “made redundant” (as the Brits say) in the next wave of layoffs? Maybe som of these free online classes from Google will float your boat.
COLE: A Graphic Design Tools for the Masses
While it’s still a research paper and not a product yet, COLE is an exciting AI Design tool that can create simple, pattern based layouts with individual layers for text and image. Similar to Adobe Express and Canva, it means anyone can be a designer…and less work for entry level designers. You can read about it here.
Google Quietly Delays Project Gemini
You probably don’t watch these things that closely, but Google’s big answer to OpenAI that was supposed to remove the embarrassing sting left by Bard has been delayed. I could only find paywall reports, so you’ll have to settle for this:
The Information’s report noted that a factor in the delay was the uncertainty of whether Gemini could equal or surpass OpenAI’s most advanced LLM in GPT-4. Given much of the publicity around Gemini were the claims that it would be superior to the Microsoft-funded GPT-4, that standard possibly not being met suggested Google wanted more time to refine the product.
Update: Business Insider has some details of what Gemini will be able to do.
If I'm trying to bake a cake, I’ll ask, “Draw me 3 pictures of the steps to how to ice a three-layer cake”, and Gemini will actually create those images. These are completely novel pictures. These are not pictures from the internet. It's able to speak in imagery with humans now, not just text.
Why it Matters: Google Gemini is going to be a real Game Changer
Further Experiments with Runway’s Motion Brush
I watched some amazing AI generated films over the weekend, along with some helpful tutorials. To be honest, I still don’t know how they get such amazing results. I do know that almost everyone runs their footage through the Topaz Gigapixel upscaler, and several people mentioned generating a prompt 3, 5, or 10 times and then cherry-picking the best result. Maybe it just takes lots of practice, but here’s my timestamp of what I was capable of on 12/03/23. I left in all the glitches and awkward moments.
That’s it for today. I hope your holiday season is going well and you’re getting a lot out of this newsletter. When you’re at the company holiday party or celebrating with family and everyone compliments you on how smart you sound, remember to tell them your secret.