John Maeda

24.8K posts

John Maeda banner
John Maeda

John Maeda

@johnmaeda

VP Eng, Microsoft CoreAI / How To Speak Machine (2025) https://t.co/tlzThtcVKS or (2019) https://t.co/eb6gj2wf1b

Redmond, WA Katılım Temmuz 2008
1.5K Takip Edilen358.6K Takipçiler
Sabitlenmiş Tweet
John Maeda
John Maeda@johnmaeda·
Metaverse in NYT (2007): “John Maeda isn't real, nor is the island. He is the Resident, and the island is the Metaverse in the virtual world of the Second Life online game. Other works exhibited are a beige Apple II computer, and a dozen or so iPod Nanos.” nytimes.com/2007/05/04/sty…
John Maeda tweet media
English
19
28
179
0
John Maeda
John Maeda@johnmaeda·
GENERATIVE UI IN THE TIME IT TAKES TO MAKE RAMEN 🍜: I was a bit startled by how easy it was to create a generative UI demo for the #DesignInTech Report 2026 designintech.report series of technical demonstrations. What does this mean? It means that we're in an accelerated transformation of what "design" will mean for the remainder of this decade. The good news is that if you already grok what you see here, you're definitely headed in the direction of a "🌳Tree-shaped" designer in tech. Keep learning how to speak machine! And don't forget to speak human in the process ... —JM --- Mini-demos are over here: dit-2026-app.john-04e.workers.dev 2026 Design in Tech Report: designintech.report
English
1
3
16
2.3K
John Maeda
John Maeda@johnmaeda·
THE 🌳 TREE-SHAPED DESIGNER: Correction to my earlier T-shaped designer diagram for the hashtag#DesignInTech Report 2026. Based on feedback, I agree that “T-shaped” turned out to be the wrong metaphor. It already means something established: broad across disciplines, deep in one. What I meant was something different. Instead, I think the future designer feels more 🌳 Tree-shaped: rooted in craft, growing deeper into stewardship, and able to branch at any stage toward amplified designer, amplified engineer, or both. When I tried to take my original hand drawn image to prompt it into the image inside my head, I couldn't get there. So I drew the new image by hand over 40+ minutes, and then used the new starting point as the context for image generation. And it turned out “good enough“ to move on from this idea space. In the process, I got to think deeper about the concept. I really like the concept of “System 3“** thinking as a way to describe this approach of solving problems. --- #DesignInTech Report 2026 “From UX to AX” is available March 18 SXSW at designintech.report --- **“Thinking—Fast, Slow, and Artificial: How AI is Reshaping Human Reasoning and the Rise of Cognitive Surrender” via UPenn/SSRN: lnkd.in/g5E98-XH
English
0
3
26
4.4K
John Maeda
John Maeda@johnmaeda·
TONY RUTH EQUITY SERIES: Back in 2019, I commissioned @lunchbreath to create a series of illustrations that fundamentally changed how I saw the differences between inequality, equality, equity, and justice. For a variety of reasons, those images were never able to be broadly distributed. As I’ve been preparing the 2026 hashtag#DesignInTech Report, I felt it was time to recraft them for a new era. I’m grateful to CEO Anna Veronika Dorogush, whose uniquely designer-centric approach to imaging and imagining made that possible. Because today is International Women’s Day, it feels like the right moment to revisit these ideas. They still are not understood clearly enough to move society in a better direction. And in the AI-fied era ahead, equity will only become more important. This is not a side issue. It is core to how we build, distribute, and benefit from the future. The recrafted images are now archived at designintech.report, and you can download them as a zip archive to share with colleagues and friends.
John Maeda tweet mediaJohn Maeda tweet mediaJohn Maeda tweet mediaJohn Maeda tweet media
English
1
1
11
2K
J. Paul Neeley
J. Paul Neeley@jape·
I really like this build John. I've always found the T model helpful and this defining of the depth I think is meaningful. I'm struggling a little with the horizontal. Hasn't this been more multiple disciplines in the past, know a little about everything. And are these descriptions another way of describing depth? Wondering if there is an adjustment... or I'm not quite understanding.
English
1
0
0
34
John Maeda
John Maeda@johnmaeda·
T-SHAPED DESIGNER MOD4AI: Like writing with my own mind, drawing with my own mind helps me learn an idea better. It’s way more effort when you compare it to what’s possible today with an image-generating AI model. But a learner’s goal isn’t a final output per se. It’s the process of discovery they earn along the way. I truly learned a lot by drawing the diagram 👇 below. That said, you might not gain much by trying to read it ... because learning doesn’t immediately transfer to another person. Putting in the work yourself is how you ... learn. The figure’s related to the new AI skilling framework EPIAS × SAE which seems to hold promise github.com/aji-ai/dit-202… as a combination of an early idea built by Monty Hammontree UXR team and the SAE’s autonomy framework from 2014. I’ve found the autonomy framework to be eminently useful across almost all disciplines because of it’s plain language approach to what it means when you’re either driving the car manually vs sitting in the back of the car while it drives you all the way to your destination. In pursuit of simplification, I’ve started to map EPIAS × SAE to the canonical ”T-shaped Designer” idea via the Stanford folks. But with a mod where I have narrowed the scope to the ongoing challenges of adopting AI in one of two flavors: 1/ augmentation, 2/ engineering. The two paths are conflated right now, so I’m using the top of the “T” to signify a detour instead of general good ole-fashioned breadth-in-anything-you-like. TL;DR: It’s okay if you’re a product designer who doesn’t want to push the boundaries of your coding skills in an IDE of your choice. Just go left (L1/L2) and pursue more autonomy. That’s within your grasp. For those who want to push production code and find their path as a “design engineer” be sure to turn right (L3/L4). BUT don’t forget that you’re not much of a designer if you haven‘t gone down the craft rabbit hole of what classical design has loooooong meant to civilization. Craft is the foundation of what “good“ has stood for, and even though there’s a lot of exponential change going on right now ... we human beings can’t change exponentially. And that’s a relief. —JM --- On March 18 at @sxsw I’ll be sharing my annual summary of everything that’s happening out there #DesignInTech schedule.sxsw.com/2026/events/PP…
John Maeda tweet media
English
2
2
35
4.5K
John Maeda
John Maeda@johnmaeda·
“My father was a blacksmith. He had his anvil and hammer. One morning he went to his shop and someone had replaced his anvil and hammer with a welder’s torch. I am fortunate to have worked in that period before the computer when we had to search for solutions with our own hands. When I did Updike’s covers, the computer as a graphic art tool was not even in existence. I didn’t just design the type for those book jackets; I drew it with my quill pen, using india inks and dyes. It’s tough for designers today that have to use the computer. The first thing I saw when I returned to Parsons after having taught there for 11 years was the students’ computers on their desks. The students weren’t there, just their computers!” –S. Neil Fujita (1921-2010) youtu.be/CpfhVmDNuho?t=… hellerbooks.com/pdfs/voice_s_n… research.library.kutztown.edu/cgi/viewconten…
YouTube video
YouTube
John Maeda tweet media
English
1
3
44
2.8K
John Maeda retweetledi
42courses
42courses@42Courses·
Book of the day 📘: How to Speak Machine by John Maeda As AI becomes part of everyday business and design, understanding how it works is no longer optional. In this accessible guide, John Maeda offers a simple framework to help you understand how machines “think” — and how to use AI thoughtfully, responsibly, and creatively. @johnmaeda amzn.eu/d/0cdY1VwP
42courses tweet media
English
0
3
14
1.3K
John Maeda
John Maeda@johnmaeda·
DESIGNING DESIGN: “Verbalizing design is another act of design.” —Kenya Hara (2011) There is a wonderful book on design from Lars Müller Publishers that is unfortunately no longer in print. It was written by one of Japan’s foremost designers, arguably at the peak of his career. Everyone knew about Hara-san in Japan, and this book in English put his flag everywhere as an understated, perfectly designed book. Instead of a Phaidon coffee table book with pictures pictures pictures. With all the blah-blah about “taste” out there #DesignInTech right now, Hara’s book stands out for its restraint (= “taste”) and its audacity (= “taste”) in the context of the non-digital domain just as computers began to emerge. To me, the work of designers Erik Spiekermann and Hartmut Esslinger carry a similar nuanced genius that is revealed thru directly experiencing their work, and sometimes in their writings. More often, the artifacts of their work spoke for them. Loudly. Similarly, it’s only by flipping through the pages of Hara’s book when the actual pages as text become secondary to the book-as-object-itself. Its weight. Its binding. Its printing. Its exceedingly high paper quality. This is the kind of design experience that many in the newer generation may not have had a chance to “feel” because frankly, great design objects are usually priced out of their market. While we hear about “world models” in the AI universe, it’s increasingly clear to me that for many centuries artisans, artists, and designers have held very clear models of our world that couldn’t easily be reduced to written language. And for that reason alone, the idea of automating “taste” with modern generative models will remain an elusive task. Long live design. —JM
John Maeda tweet media
English
11
68
642
26.4K