Build first, plan second.
Unlock all features
FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.
Unlock all features
FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.
Unlock all features
FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.
Unlock all features
FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.
Unlock all features
FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.
Related videos
Anthropic fights back
Theo - t3․gg
92.6k views
This is bad...
Theo - t3․gg
104.3k views
Who Knows Wildes’ Best Almost Broke the Show 😭 | Second Things First
First Things First
18.3k views
I’m serious.
Theo - t3․gg
95.7k views
It's finally here.
Theo - t3․gg
152.2k views
Theo fell off
Theo - t3․gg
48.5k views
OpenAI: Trapped in 2nd place
Theo - t3․gg
59.0k views
I’m so tired.
Theo - t3․gg
65.4k views
GPT-5.1 is built for normies
Theo - t3․gg
67.0k views
Finally.
Theo - t3․gg
143.5k views
Top Comments (10)
Also shoutout to the twitch viewer that said his excalidraw diagrams are design docs
lmao this video is just theo renting about his colleagues
The correct approach is to make a proof of concept or a bunch of little tests for each thing you want to do. Then write the design doc using what you’ve learnt including figuring out the best way to do each thing. Then code the actual project. This way all the changes people would make to their projects if they got to build them a second time actually get implemented. In the example you give the people writing the doc should have been out talking to every relevant party and using the doc as a record of all the minutes from those discussions. The danger more often that not is that designers don’t communicate enough about what they’re doing and end up building something people don’t actually want. Docs are a method of communication.
buildthing vs planthing
Imo. it's f*d up that you have to hide your prototype from management (your own or of a customer company) if it's not ready to deploy. That's exactly my experience in the industry as well, but it's a pure management problem. The times I had positive experiences with PoCs and management, was when the managers have an engineering backgroud.
There is a contnous delivery video where he talks about the exact topic. he says "Everybody has a plan until they get punched in the mouth" and "Planning is absolutely essential but plans are useless" I agree with both statements.
I worked at a company where the PMs were never talking to the customers because "they know what the user wants", well we never got any adoption for what we built and it was so frustrating. I burnt out so fast, building stuff you know will never get used is so frustrating
Fully agree. Only way I have ever written a successful design doc is if we built a prototype version first. And the design doc was there mainly for regulatory reasons at that point.
I recently worked for 2 weeks on a webapp, i felt it was polished so i went to prod. And i found so many things i had to redo, it really speaks to how much the reality is different from planning.
A poc is good proof, a design doc allows you to explain specifics without building it and build a robust idea. Both are allowed to exist
Unlock the Data Inside
Turn Videos into Knowledge
- Get FREE 10/day: transcripts, summaries, chats
- Chat with videos, export text & PDF
- $1 free API credit for RAG, chatbots & research
Free forever plan • All features unlocked
Top Comments (10)
Also shoutout to the twitch viewer that said his excalidraw diagrams are design docs
lmao this video is just theo renting about his colleagues
The correct approach is to make a proof of concept or a bunch of little tests for each thing you want to do. Then write the design doc using what you’ve learnt including figuring out the best way to do each thing. Then code the actual project. This way all the changes people would make to their projects if they got to build them a second time actually get implemented. In the example you give the people writing the doc should have been out talking to every relevant party and using the doc as a record of all the minutes from those discussions. The danger more often that not is that designers don’t communicate enough about what they’re doing and end up building something people don’t actually want. Docs are a method of communication.
buildthing vs planthing
Imo. it's f*d up that you have to hide your prototype from management (your own or of a customer company) if it's not ready to deploy. That's exactly my experience in the industry as well, but it's a pure management problem. The times I had positive experiences with PoCs and management, was when the managers have an engineering backgroud.
There is a contnous delivery video where he talks about the exact topic. he says "Everybody has a plan until they get punched in the mouth" and "Planning is absolutely essential but plans are useless" I agree with both statements.
I worked at a company where the PMs were never talking to the customers because "they know what the user wants", well we never got any adoption for what we built and it was so frustrating. I burnt out so fast, building stuff you know will never get used is so frustrating
Fully agree. Only way I have ever written a successful design doc is if we built a prototype version first. And the design doc was there mainly for regulatory reasons at that point.
I recently worked for 2 weeks on a webapp, i felt it was polished so i went to prod. And i found so many things i had to redo, it really speaks to how much the reality is different from planning.
A poc is good proof, a design doc allows you to explain specifics without building it and build a robust idea. Both are allowed to exist