Sony World Photography Awards. One photography award winning image ain’t a photo. The 1st prize winner in the Creative category of the Sony World Photography Awards apparently was not a photo but a stunning AI-generated image. Needless to say: lots of controversy. The creator, Boris Eldagsen, declined the prize: “AI images and photography should not compete with each other in an award like this. They are different entities. AI is not photography. Therefore I will not accept the award.” 

How did he create the image? “[through] a complex interplay of prompt engineering, inpainting, and outpainting that draws on my wealth of photographic knowledge. For me, working with AI image generators is a co-creation, in which I am the director.” 

His take on AI vs. photography: “Just as photography replaced painting in the reproduction of reality, AI will replace photography. Don’t be afraid of the future.” 

Join our Visual 1st Spotlight on May 17 to discuss the implications of generative AI for various parts of the photo & video industry!

May 17, 8:00 – 9:30 am PDT

(5:00 – 6:30 pm CEST)

Live-streamed

The Latest AI Tech Announcements

What do they mean for the

photo & video industry?

Early Bird: just $19 until May 3!

TIPA. Photo & Imaging Products Awards. The Technical Image Press Association (TIPA) announced the TIPA World Awards 2023 for the 40 Best Photo and Imaging Products of the year. Products include hardware, software and services. 

On a separate note, I’m honored to have been elected as a TIPA member 🎉. TIPA members include editors and representatives of photographic magazines and online publications from North America, Europe, Asia, and Australia, as well as individual imaging and technology experts.

Kodak Alaris. For sale. Britain’s Pension Protection Fund (PPF), owners of Kodak Alaris, is said to auction off Kodak Alaris. Kodak Alaris was spun off from Eastman Kodak in 2012 and is comprised of the Kodak photo kiosks and film division – under the Kodak Moments brand – and Alaris, which makes document scanners and software.

In its accounts for the year to March 31, 2022, Kodak Alaris reported a 29% increase in revenues to €130M, “although the scale of the increase was partly the result of its recovery from the COVID-19 pandemic.”

Adobe Firefly. AI video, here we come. Adobe is on a generative AI roll. Announced barely a month ago Firefly, a suite of generative AI models that can both create and transform visuals using text prompts in the same way that Dall-E and ChatGPT do, is now taking on video as well. 

Its new features are designed to help professional editors cut down on their video drudge work, such as boosting color levels, inserting placeholder images, adding effects, and autonomously recommending b-roll for a given project – simply by typing their ideas into Firefly's AI text prompt and letting the algorithm do its thing. This will include "text to color enhancements," a broad-ranging capability that can adjust the brightness and saturation levels, shift the time of day – even the time of year – using natural language prompts.

Picanova. Now aka: The Customization Group. Longtime Visual 1st sponsor Picanova has rebranded as The Customization Group. The rebranding to The Customization Group signifies the company’s evolution from a traditional photo product company to an industry leader offering comprehensive mass customization solutions, according to the company.

Bria. Ethically sourced AI. Bria, presenter at our Visual 1st Generative AI Spotlight earlier this year, announced plans for ethically sourced generative AI creation, including sharing revenues with content providers based on the contributions’ impact and allowing artists to set prices on a per-AI-training-run basis.

(more industry news below)

To whet your appetite, here is one example of a text-to-text generative AI creator, Google Bard, sharing his/her/their opinion on the key question that our Spotlight panel will address ☺

Not bad as an intro, but we look forward to diving much deeper into this question on May 17!

ON1. Upscaling + enhancing. ON1’s latest version of its Resize AI tool rescales images while also showing realistic details in the enhanced image. To mitigate noise and over-sharpening artifacts when enlarging photos, ON1 is also adding a component of its AI-powered NoNoise AI app to Resize AI’s upscaling workflow.

Midjourney. Flipping the formula. We all know: Midjourney creates images from text, but now it also creates text from images. Note this is different from good-old AI image recognition engines that provide tags for the objects they identify inside an image. 

Rather than providing a tag saying "woman" it text AI-triggered by an image might say: “a young woman wearing a black and white polka dot dress standing, in the style of hazy landscapes, pensive portraiture, sun-soaked colors, candid portraiture, tumblewave, troubadour style, close up.”

Use cases? Well, you could think of accessibility for the visual impaired, generating captions, creating better search engine indexing than from the images themselves, creating variations of the text which, in turn, then could generate variations of the original image. And so on.

EyeEm. RIP. EyeEm, the once fast-growing, AI-fueled, mobile-first photosharing site transformed into photo marketplace, filed for bankruptcy. Transforming a startup with great technology & community into a longterm profitable company turned out to be a whole different beast.

Amazon. Platform for AI language models. Generative AI tech requires massive computing power. So, it is no surprise that Amazon’s AWS is also entering the fray. Amazon launches Bedrock, an AI platform for AWS customers that offers a suite of (mostly) third-party generative AI tools that can build chatbots, generate and summarize text, and classify images based on prompts. Bedrock users can perform tasks by selecting from a range of machine learning models called “foundation models,” including Jurassic-2, Stability AI’s Stable Diffusion, and, also just announced, Amazon’s own Titan foundation model.

BeReal. Sizzled? When BeReal launched, it was a relief from all the inauthentic, self/stuff-promoting visuals Creators tend to post on Instagram, TikTok and the likes. Its main feature: prompting users at a different time each day to take shots with their front and rear phone cameras, and to share an unedited shot within 2 minutes. A runaway hit still 6 months ago, BeReal appears to have sizzled. Daily active usage has dropped 61% from its peak, from about 15M in October to less than 6M in March.

Are our (and our friends’) daily lives too boring to show authentically day after day? And why didn’t BeReal learn a lesson or two from Snapchat as to how to expand beyond a single use case when the app is still hot? 

Best,

Hans Hartman