Learning the Language of Ai and Who to Follow - AIxC: 72
+ Image-to-3D Strategy for Paid Members
Welcome back. As always lots of change in this space in the matter of a week. This reminds me of the ever-present challenge we face here - that of keeping up with the latest, so that we can share our findings with you.
The majority of our projects come about via conversations with design leaders - CDOs, VPs of Design/Creative, Directors, Founders. They share the same challenges.
In today’s newsletter, I’ll focus on how we can all keep some semblance of comfort in this rapidly evolving business & design landscape.
Hector & the AIxC Team
01 / NEWS
Know Where To Look
Do you know what my number one source of Ai information is?
Linkedin.
My network is heavily design and Ai-focused and therefore so is my feed. I see what my network posts, but more importantly, Linkedin shows me what my network is liking and commenting on.
My advice:
1. Don’t follow the engagement-farmers. Sensational posts are rarely helpful.
2. Follow these people instead, which aside from great posts, post on a consistent basis:
Tatiana Tsiguleva - Lots of Midjourney explorations and clever implementations.
Santiago Valdarrama - Machine learning, but I pick up lots of parallels to my domain.
Allie Miller - Stays on top of the overarching Ai landscape. Lots of good content.
Gizem Akdag - Lots of great Midjourney SREF efforts and a great eye for visuals.
Luka Tisler - Latest from the Ai tech point of view for visual world.
Reuven Cohen - Focused on coding side, but some crazy experiments to expand thinking.
Matthieu Lorrain - Consistently posting latest news in Ai world.
Lori Mazor - Really dig her writing style. Simplifies Ai in unique ways.
Greg Aper - My Ai compadre. Leading-edge UX / Design efforts (not daily posts).
Ken Cheng - Nothing to do with Ai but makes me laugh every day.
There are lots more, but this is a good start to get a good well-rounded sampling of Ai news.
02 / THOUGHTS
Learn the Language, Not the Tools
Here’s something I see often.
If someone was stumped by a CAD model, they would know that it’s just a matter of spending more time with the CAD tool. They would understand that a more experienced CAD user could accomplish it easily.
The rules are simple with traditional tools. There are no shortcuts to proficiency. It takes hours of practice. You could tell who has put the time in and who hasn’t.
When it comes to Ai though, we don’t apply the same rules. We expect successful results with minimal hours of practice. At the most extreme, we discard the entire technology as useless because it doesn’t do what we want - after 1 week of testing it out.
It’s an interesting phenomenon.
I think there are multiple reasons for this:
Ai tools change at a significantly faster rate than previous tools. It feels like an unstable foundation to invest in.
We don’t trust Ai tools and Ai practitioners (see first advice above).
Cognitive hurdles. It’s hard to fully understand tools that sometimes contradict how traditional tools work.
Emotional hurdles. We don’t like change.
The good news is that Ai is like learning how to speak a new language. If you learn the language, you can utilize it to communicate with any Ai tool, regardless of new features and new launches.
In a future where Ai tools will be a staple in all of our toolkits, it’s of my opinion that learning the language of Ai is a great investment of our time.
Invest in learning the language, not the tools.
03 / THOUGHTS
Consistent Learning via R&D
We help brands with eLearning, Live Training and R&D Projects.
My favorite structure is when we work with brands on long-term retainers, with monthly R&D efforts. The result is the following:
Teams learn-by-doing in their line of work (solves cognitive and emotional hurdles)
Teams stay up-to-date with the latest features for an extended period of time (long enough to learn language).
Execution of education efforts on real projects for tangible results (quick wins while making forward progress on existing projects is a win-win)
Whether with an external source like ours, or whether done internally, I suggest a consistent R&D effort like this. It’s proven to be a successful path towards Ai fluency for us and our clients.
If you’re a design leader and want to chat, always happy to talk ai + design.
04 / WORKFLOWS
The Language of Ai
Speaking of learning the language of Ai, this language is best spoken with visuals.
Krea gives you the ability to add custom trained (on your images) models and reference images, controlling how much of each you want to influence your results.
The more you use Ai tools, the more you realize this approach of images and levers is consistent throughout tools.
Give this a shot to steer Krea’s Flux tool away from the averages.
05 / FOR PAID SUBSCRIBERS
Multiple Views to 3D
We’ve uploaded some new workflows to the AIxC R&D Lab. In these workflows we’re exploring:
Generating multiple views with Vizcom
Multiple views for training Ai tool models
Generating multiple views with trained models
Multiple views for image-to-3D
Sign up below for access to the AIxC R&D Lab board. Link and password behind the paywall.