exploring shapes of thoughts: extracted my obsidian notes' embeddings and arranges them as a 3d network using 3 different topologies:
- centralized: one core idea connecting all
- decentralized: notes cluster into themed hubs
- distributed: edges labeled by llm describing how ideas connect
Strange times.
Europe is largely composed of museum ethnostates straining and failing to become pluralistic civic societies; the U.S. is a pluralistic immigrant civic state now straining and failing to become an ethnostate.
When Bitcoin wins...
0:00 It ends the Fed
02:10 Payments become packets
3:30 Keynesianism falls like Communism
4:20 Real estate falls in real terms
5:40 Fiat billionaires get flipped
6:20 Money printer runs out of toner
7:50 The battle begins...
10:00 ...and it's time to build
Two important principles that have a lot in common:
If you make users love you, investors will love you too.
If the people lead, the leaders will follow.
Figma co-founder and CEO Dylan Field (@zoink) on how AI will transform design and why craft and attention to detail will be the ultimate differentiators.
At AI Startup School in San Francisco.
01:38 – How Dylan and Evan Started Figma
02:58 – WebGL: From Games to Tools
04:10 – Buying Time to Build
05:48 – Staying Motivated Through Early Exploration
06:10 – Cold Emails, Early Feedback & First Users
08:24 – Launch Timing and Lessons on Shipping Early
10:08 – Culture of Constraints & Creative Problem-Solving
10:50 – Recognizing Product-Market Pull
13:32 – Why Design is the Differentiator in the AI Era
16:34 – Figma’s AI Product Launches & Expanding Beyond Design
18:56 – The Blurring of Design, Development, and Product
21:00 – The “MS-DOS Era” of AI Interfaces & Future Surfaces
23:26 – The Role of Design in AI Research & the Future of Designers
27:36 – Audience Q&A: AI Tools, Open Source, Principles, and Advice
0% AI is slow.
But 100% AI is slop.
So the optimal amount of AI is actually between 0-100%. The exact figure varies by situation, but just the idea that 0% and 100% are both suboptimal is useful.
It's the Laffer Curve, but for AI.
Beautiful design is now a commodity.
I've spent the last 24 hours with ChatGPT 4o images, and it's clear we've entered a new reality: "Execution is cheap, ideas are everything."
For decades, we were told the opposite. Everyone had ideas. Few could execute them well. The ability to turn a concept into reality separated the winners from the dreamers.
But in an AI world, it's completely flipped.
When anyone can execute at 90% perfection with the right prompts, the limiting factor becomes the quality of your ideas. The creative direction. The strategic insight. The unique perspective.
The most successful companies I'm seeing are shifting resources from production to ideation. Less time pushing pixels, more time exploring concepts.
They're running 20-30 creative directions where they used to do 2-3, because the cost of trying ideas has collapsed.
In a world where anyone can create a beautiful website, logo, or packaging, the winners are focusing on the things AI can't (yet) simulate:
I think it's authentic relationships, innovative products, and unique perspectives.
The real advantage is in knowing when to break the rules of good design in ways that resonate emotionally.
The human touch is becoming less about execution and more about strategic deviation from the optimized norm.
This is creating strange new dynamics in hiring too. When I started our design agency @meetLCA, we hired for world class technical skills - mastery of tools, execution ability.
But now we care more about hiring for conceptual ability and creative direction. People who consistently generate novel ideas rather than perfect executions. Obviously, top tech skills still matter, but way less.
As AI makes "good enough" design accessible to everyone, the market is splitting. At the low end, good enough is actually good enough.
But at the high end, there's a premium on the truly unexpected - the ideas an AI wouldn't generate because they break conventional patterns.
I think we're heading toward a bifurcated creative world: automated beauty for most purposes, with human creativity focused on creating the unexpected, the ideas and approaches an AI wouldn't think to try because they don't follow established patterns of "good design."
The challenge for most of us now isn't "how do we execute this idea?" but "which ideas are actually worth executing?"
Execution is cheap, ideas are everything.
Tremendous alpha in it.
You're an idea person now. We all are?