16 years designing products.
Now I'm rebuilding my career in public.
Tiny audience.
Zero product sales.
What should I actually build in Q3/Q4 2026?
1. Fix my distribution (Framer templates aren't selling)
2. Build a social growth tool for designers
3. Fix the next bottleneck: AI-generated prototypes that developers won't touch
Reply with 1, 2 or 3 + why.
Or tell me I'm cooked. Both are welcome.
Garden room setup, optimised for light (and still iterating).
Moved the desk closer to the window. Natural light has been the quiet productivity win. Everything else stayed the same — just better energy and focus throughout the day. Been bouncing layout and setup ideas with ChatGPT.
Still refining what stays on the desk and what doesn’t. The Framer notebook is permanently on the desk. Feels right. Anyone else designing their workspace around light first?
Codex → Framer experiment.
I've been exploring whether AI can help prototype interactions, not just generate layouts.
This cursor-reactive animation started in Codex before moving into Framer.
It's definitely not perfect, but it's already fast enough for exploring ideas before refining them manually.
Feels like a glimpse of a different design workflow.
What interaction should I experiment with next?
Codex → Framer experiment.
I've been exploring whether AI can help prototype interactions, not just generate layouts.
This cursor-reactive animation started in Codex before moving into Framer.
It's definitely not perfect, but it's already fast enough for exploring ideas before refining them manually.
Feels like a glimpse of a different design workflow.
What interaction should I experiment with next?
Codex just made my Framer hero breathe.
Quick experiment:
Shared my Framer project URL with Codex and asked for a subtle pulse-ring animation that feels alive. What came back is a clean, looping breath effect — soft scale + opacity that never feels mechanical.
Dropped it straight into the hero. One component, no messy overrides. It’s the kind of micro-interaction that makes the whole page feel intentional.
Anyone else running similar experiments?
Just how easy is it to create variants in Framer?
Same hero section.
Only the background changes.
Dark → Color → Light
Done in under a minute.
No complex setup.
4 hero concepts testing how to visually guide the eye to the CTA.
1. Dot
2. Portrait gaze
3. Gradient beam
4. Sparkle shader
Which one wins
Testing lighting + Framer particle shaders on this hero.
Which version hits harder?
I'm curious...
Would you use a curated gallery of design posts from smaller creators on x that significantly outperformed their audience size?
I feel there are a lot of patterns hiding in those posts.
One concept. 10 visual directions.
Exploring how different backgrounds, composition and typography completely change the feel of the same hero section.
Which one stands out?
One concept. 10 visual directions.
Exploring how different backgrounds, composition and typography completely change the feel of the same hero section.
Which one stands out?
Text indent is one of those overlooked techniques that instantly makes long-form content feel more editorial and refined.
Built this as a Framer component.
1 hero. 5 different backgrounds.
Photo, Color, Illustration, Abstract, Gradient.
Which one wins?
One hero. Four layouts.
Framer's Grid makes it easy to customize a hero without rebuilding it.
Same content. Four different layouts.
Which one gets your vote?
One hero. Four fonts.
Four completely different first impressions.
Which one would you hire?
4 hero layouts, easy to change.
Using the grid can help make a lot of different layouts.
Keeping it simple and easy to change.
Testing a stripped-back hero with a clearer value proposition and a playful cursor image trail.
Would love to know what you think. 👀
One face, made entirely of letters. then it morphs into another.
It's a poster for today's World Cup Final, built in framer as a single canvas component, no plugin.
Live here: https://xxvi.framer.website/
Taste got it 90% there in an afternoon with Claude. the last 10% is timing, and it still drifts — but i'd rather show the crazy version now than the perfect one never.
Lots of layers to this:
- Blend modes
- Radial gradient
- Signatures from wikipedia
- Lamine yamal image from youtube thumbnail https://youtu.be/-gGggH5YJQw?si=0oxae0cJcNlbcMc3
- Lionel messi generated in midjourney
- Ascii art code component made with Claude
Built a World Cup final poster from a 4x4 grid, no images, just frames. Still figuring out the background. Ideas?
Some custom cursors are decoration.
These three react to what they're on.
16 years designing products.
Now I'm rebuilding my career in public.
Tiny audience.
Zero product sales.
What should I actually build in Q3/Q4 2026?
1. Fix my distribution (Framer templates aren't selling)
2. Build a social growth tool for designers
3. Fix the next bottleneck: AI-generated prototypes that developers won't touch
Reply with 1, 2 or 3 + why.
Or tell me I'm cooked. Both are welcome.
Garden room setup, optimised for light (and still iterating).
Moved the desk closer to the window. Natural light has been the quiet productivity win. Everything else stayed the same — just better energy and focus throughout the day. Been bouncing layout and setup ideas with ChatGPT.
Still refining what stays on the desk and what doesn’t. The Framer notebook is permanently on the desk. Feels right. Anyone else designing their workspace around light first?
Codex → Framer experiment.
I've been exploring whether AI can help prototype interactions, not just generate layouts.
This cursor-reactive animation started in Codex before moving into Framer.
It's definitely not perfect, but it's already fast enough for exploring ideas before refining them manually.
Feels like a glimpse of a different design workflow.
What interaction should I experiment with next?
Codex → Framer experiment.
I've been exploring whether AI can help prototype interactions, not just generate layouts.
This cursor-reactive animation started in Codex before moving into Framer.
It's definitely not perfect, but it's already fast enough for exploring ideas before refining them manually.
Feels like a glimpse of a different design workflow.
What interaction should I experiment with next?
Codex just made my Framer hero breathe.
Quick experiment:
Shared my Framer project URL with Codex and asked for a subtle pulse-ring animation that feels alive. What came back is a clean, looping breath effect — soft scale + opacity that never feels mechanical.
Dropped it straight into the hero. One component, no messy overrides. It’s the kind of micro-interaction that makes the whole page feel intentional.
Anyone else running similar experiments?
Just how easy is it to create variants in Framer?
Same hero section.
Only the background changes.
Dark → Color → Light
Done in under a minute.
No complex setup.
4 hero concepts testing how to visually guide the eye to the CTA.
1. Dot
2. Portrait gaze
3. Gradient beam
4. Sparkle shader
Which one wins
Testing lighting + Framer particle shaders on this hero.
Which version hits harder?
I'm curious...
Would you use a curated gallery of design posts from smaller creators on x that significantly outperformed their audience size?
I feel there are a lot of patterns hiding in those posts.
One concept. 10 visual directions.
Exploring how different backgrounds, composition and typography completely change the feel of the same hero section.
Which one stands out?
One concept. 10 visual directions.
Exploring how different backgrounds, composition and typography completely change the feel of the same hero section.
Which one stands out?
Text indent is one of those overlooked techniques that instantly makes long-form content feel more editorial and refined.
Built this as a Framer component.
1 hero. 5 different backgrounds.
Photo, Color, Illustration, Abstract, Gradient.
Which one wins?
One hero. Four layouts.
Framer's Grid makes it easy to customize a hero without rebuilding it.
Same content. Four different layouts.
Which one gets your vote?
One hero. Four fonts.
Four completely different first impressions.
Which one would you hire?
4 hero layouts, easy to change.
Using the grid can help make a lot of different layouts.
Keeping it simple and easy to change.
Testing a stripped-back hero with a clearer value proposition and a playful cursor image trail.
Would love to know what you think. 👀
One face, made entirely of letters. then it morphs into another.
It's a poster for today's World Cup Final, built in framer as a single canvas component, no plugin.
Live here: https://xxvi.framer.website/
Taste got it 90% there in an afternoon with Claude. the last 10% is timing, and it still drifts — but i'd rather show the crazy version now than the perfect one never.
Lots of layers to this:
- Blend modes
- Radial gradient
- Signatures from wikipedia
- Lamine yamal image from youtube thumbnail https://youtu.be/-gGggH5YJQw?si=0oxae0cJcNlbcMc3
- Lionel messi generated in midjourney
- Ascii art code component made with Claude
Built a World Cup final poster from a 4x4 grid, no images, just frames. Still figuring out the background. Ideas?
Some custom cursors are decoration.
These three react to what they're on.
16 years designing products.
Now I'm rebuilding my career in public.
Tiny audience.
Zero product sales.
What should I actually build in Q3/Q4 2026?
1. Fix my distribution (Framer templates aren't selling)
2. Build a social growth tool for designers
3. Fix the next bottleneck: AI-generated prototypes that developers won't touch
Reply with 1, 2 or 3 + why.
Or tell me I'm cooked. Both are welcome.
Garden room setup, optimised for light (and still iterating).
Moved the desk closer to the window. Natural light has been the quiet productivity win. Everything else stayed the same — just better energy and focus throughout the day. Been bouncing layout and setup ideas with ChatGPT.
Still refining what stays on the desk and what doesn’t. The Framer notebook is permanently on the desk. Feels right. Anyone else designing their workspace around light first?
Codex → Framer experiment.
I've been exploring whether AI can help prototype interactions, not just generate layouts.
This cursor-reactive animation started in Codex before moving into Framer.
It's definitely not perfect, but it's already fast enough for exploring ideas before refining them manually.
Feels like a glimpse of a different design workflow.
What interaction should I experiment with next?
Codex → Framer experiment.
I've been exploring whether AI can help prototype interactions, not just generate layouts.
This cursor-reactive animation started in Codex before moving into Framer.
It's definitely not perfect, but it's already fast enough for exploring ideas before refining them manually.
Feels like a glimpse of a different design workflow.
What interaction should I experiment with next?
Codex just made my Framer hero breathe.
Quick experiment:
Shared my Framer project URL with Codex and asked for a subtle pulse-ring animation that feels alive. What came back is a clean, looping breath effect — soft scale + opacity that never feels mechanical.
Dropped it straight into the hero. One component, no messy overrides. It’s the kind of micro-interaction that makes the whole page feel intentional.
Anyone else running similar experiments?
Just how easy is it to create variants in Framer?
Same hero section.
Only the background changes.
Dark → Color → Light
Done in under a minute.
No complex setup.
4 hero concepts testing how to visually guide the eye to the CTA.
1. Dot
2. Portrait gaze
3. Gradient beam
4. Sparkle shader
Which one wins
Testing lighting + Framer particle shaders on this hero.
Which version hits harder?
I'm curious...
Would you use a curated gallery of design posts from smaller creators on x that significantly outperformed their audience size?
I feel there are a lot of patterns hiding in those posts.
One concept. 10 visual directions.
Exploring how different backgrounds, composition and typography completely change the feel of the same hero section.
Which one stands out?
One concept. 10 visual directions.
Exploring how different backgrounds, composition and typography completely change the feel of the same hero section.
Which one stands out?
Text indent is one of those overlooked techniques that instantly makes long-form content feel more editorial and refined.
Built this as a Framer component.
1 hero. 5 different backgrounds.
Photo, Color, Illustration, Abstract, Gradient.
Which one wins?
One hero. Four layouts.
Framer's Grid makes it easy to customize a hero without rebuilding it.
Same content. Four different layouts.
Which one gets your vote?
One hero. Four fonts.
Four completely different first impressions.
Which one would you hire?
4 hero layouts, easy to change.
Using the grid can help make a lot of different layouts.
Keeping it simple and easy to change.
Testing a stripped-back hero with a clearer value proposition and a playful cursor image trail.
Would love to know what you think. 👀
One face, made entirely of letters. then it morphs into another.
It's a poster for today's World Cup Final, built in framer as a single canvas component, no plugin.
Live here: https://xxvi.framer.website/
Taste got it 90% there in an afternoon with Claude. the last 10% is timing, and it still drifts — but i'd rather show the crazy version now than the perfect one never.
Lots of layers to this:
- Blend modes
- Radial gradient
- Signatures from wikipedia
- Lamine yamal image from youtube thumbnail https://youtu.be/-gGggH5YJQw?si=0oxae0cJcNlbcMc3
- Lionel messi generated in midjourney
- Ascii art code component made with Claude
Built a World Cup final poster from a 4x4 grid, no images, just frames. Still figuring out the background. Ideas?
Some custom cursors are decoration.
These three react to what they're on.
