built a mascot for a voice product
statics in krea 1 / flux
editing in nano banana pro
animation in seedance 1.5 pro and kling 2.5
voice-to-text solution by @bbssppllvv
we built a website for our design studio with head tracking magic (!)
vibe-coding stack: codex (gpt-5.3/5.4), claude code (opus 4.6), cursor.
tech: three.js + custom webgl/sdf rendering, mediapipe face mesh, matter.js (physics overlays), p5.js
play with it in the comments
i found a github repo that lets you spin up an ai agency with ai employees
engineers, designers, growth marketers, product managers
each role runs as its own agent and they coordinate to ship ideas
10k+ stars in under 7 days
1. engineering (7 agents)
frontend, backend, mobile, ai, devops, prototyping, senior development
2. design (7)
ui/ux, research, architecture, branding, visual storytelling, image generation
3. marketing (8)
growth hacking, content, twitter, tiktok, instagram, reddit, app store
4. product (3)
sprint prioritization, trend research, feedback synthesis
5. project management (5)
production, coordination, operations, experimentation
6. testing (7)
qa, performance analysis, api testing, quality verification
7. support (6)
customer service, analytics, finance, legal, executive reporting
8. spatial computing (6)
xr, visionos, webxr, metal, vision pro
9. specialized (6)
multi agent orchestration, data analytics, sales, distribution
what i like about this approach is the framing
instead of one big ai agent trying to do everything, you structure it more like a company. specialized agents, clear responsibilities, workflows between them
im curious to see what this actually feels like in practice and if its any good (do your own research)
https://t.co/plSvZIaDpr
but as always will share what i learn in public and on @startupideaspod
one thing is for certain and it reminds me
the future belongs to those who tinker with software like this
AI agents read markdown better than they read your mind
Built an ascii wireframe editor. Draw a page in 30 seconds, copy/paste into Claude Code and get a full working page back
i like saving things as screenshots. it’s the most universal button across any app. it helps me save anything instantly.
what if i could drop a screenshot into a collection, ask an ai about it, find the product, set a reminder, and more
screenshot → do something (ver. 2.0)
we built a small camera tool
for dynamic backgrounds:
disgram.
(1) a short clip is used as a glsl distortion texture and applied to (2) another video in real time, with swipe control for intensity. (3) the result is rendered as mp4 and saved on the device
I’m telling you, the tools we have right now will all feel outdated within a few months. You’ll be building your own tools. The future designer has their own unique set of tools that they’ve built themselves.
this why Hollywood is terrified of AI
this new AI Relight tool on Higgsfield can simulate film studio and change the source light direction, colour, exposure and even softness
AI is taking over film studios
here's how to use it:
Prompt engineering is dead.
Anthropic just published their internal playbook on what actually matters: XML-structured prompting.
Only 2% of users know this exists.
Here's what changed:
AI prompting isn't about being clever with words.
It's about understanding how the model was trained and structuring inputs to match that architecture.
Claude was built for structured reasoning.
Most users are still having unstructured conversations.
5/5 Resolution
This is why empathy matters.
If you want intelligence at scale to work with you,
you must communicate as if misunderstanding is the default —
and clarity is your responsibility.
If your prompts “don’t work”,
it’s rarely the model.
Vibecoding exposes a harder truth:
most people struggle to express intent clearly.
This is a communication problem, not an AI one.
🧵↓
4/5 What skill is actually required
Prompting is not a trick or a hack.
It’s applied communication.
It requires:
– structured language
– explicit constraints
– conscious removal of assumptions