Yesterday I took part in a (compensated) focus group organized by Ronin International to discuss the creative process and preview Adobe Project Concept.
Prior to session start I was unaware of what material would be covered. I had no previous knowledge of Project Concept.
The first half of the virtual meeting was spent by the group leader performing casual interviews of me and three other creative professionals. Half of us were creative leaders, the other half were designers. Each of us shared our experiences in brainstorming, documenting and sharing ideas, and putting together mocks and early drafts for client review.
The second half introduced Adobe Project Concept. Adobe made the reel available for download so I have embedded it below:
Without being able to use the app, it is hard for me to give a granular analysis of the product. But I believe I learned enough in the focus group to make a short list of optimistic, and pessimistic points.
Optimism:
It’s an Adobe app that will be fully integrated with the usual suspects: Illustrator and Photoshop. It will also pair with Adobe Express.
As an Adobe product, it is reasonable to expect the interface will be familiar with habitual Adobe users.
Concept is built for collaboration. The details were murky, but it was implied that project files would be shared through Creative Cloud.
Based on the reel, and the info I received from the group leader, it looks like this app might be a godsend for those of us who are sick to death of paying for mockup PSDs on an as-needed basis, or risking god-knows-what on some susaf page with a .ru tag.
The reel demonstrates the software will be able to deliver high-quality photo comps. More on this in a bit.
Pessimism:
Without question, this app will be used to shoehorn more efficiencies into the early stages of a project. I doubt Concept will be included in any existing CC plans without a billing adjustment. This means the bean-counters will want to see a return. But how will that return be defined? Expect deadline compression for those of us in-house.
AI has a longstanding problem with racial bias. AI has issues with gender bias.AI has issues with age bias. AI has issues with beauty standards. Anyone who has seen their share of slop knows that AI def has a “type.” During the focus group, the leader seemed very enthused about the power of AI to break the boundaries of human imagination and digital editing skill. So far I haven’t seen much from Firefly or Midjourney to suggest that AI does more than generate a generic avatar for Black Woman, Asian Grandparent, or—FFS—Jesus Christ himself. To drill into this point, I challenge you to ask your AI of choice to produce an image of “an intelligent designer.” Strangely, when I put both Midjourney and GPT to this test, I always wind up with a young, kinda-cute-looking Asian male or his slightly quirkier (and bustier) female equivalent. This is a problem. It is even more of a problem when thought of in the context of my next point…
Concept looks poised to fill the role of social content design. Having led social for a very high-volume content producer myself, I can assure you that churning ever more imagery into the insatiable gape of social is a time-consuming and costly enterprise. But Concept, paired with the right “prompt engineer,” could generate months of posts in very little time. With Firefly at the ready, image sourcing would be “responsible.” Responsible or not, the demands of design for social virtually guarantee even more slop—and even more AI biases.
“Project Concept will also respect the work of artists by leveraging Content Credentials technology to properly recognize the source of images, respect generative AI usage and training preferences, and promote transparency for how content was created. Adobe’s Firefly generative AI models were also trained responsibly on content that Adobe has a license to train on.”
But there is another responsibility. It is one everyone must consider because the situation is increasingly grim. It is one that no one in tech wants to talk about. AI is exceptionally costly. Juicing the machine is equivalent to creating an entirely new, densely populated nation-state out of thin air. While Adobe assures us that the AI was trained only on content Adobe had the rights to (as if we had any other fucking choice, btw), Adobe does not have the rights to the global climate. So it is up to the end user to determine whether or not the cost is worth the loss. I’m pretty sure I know how that choice will be made. I think you know too.
Of note is the concern the group leader showed when I brought up issues included in the list above. This was peculiar to me, as I had assumed his role was non-partisan. In this I may have been mistaken.
But, like it or not, Concept is coming. It is coming with a lot of valid use cases and will streamline iteration. But it is also coming to drop a hell of a lot more shit like this on the web…
