Rapid Prototyping and Proof-of-Concepts


structure:

  • hook: pay attention to your ideas and dare to act on them
  • Do not miss the point
  • Test your ideas soon and fast
  • Do not start from scratch
  • Yes, you can use AI, nobody cares
  • Work on it
  • Share it, gather feedback, understand the problem
  • Iterate on it
  • Decide your next move

Ideas are avocados they go grey quickly you can lose momentum, over think an idea, and forget about it forever you should chase it shiny object syndrome can lead to beautiful things you’ll know when to activate tunnel vision chase them, act on them enough to document them and have a story about them this is inspired by tiny experiments life is too precious to not try things out or to try only one do not try to be a renaissance man do not try to be a master you cannot excel at anything in short period of time but you can still learn significantly more this way, instead of passively reading about something that’s why we value hands-on training, or labs in universities

“you may or may not subscribe to the idea that the human brain is merely receptor of information, and ideas travel from host to host, when they’re not acted upon, they get broadcast, but it at least motivates me to act quickly, to validate the idea that just popped into my head”


Do not miss the point - what are you doing this for?

  • are you curious about this technology?
  • are you trying scratch your own itch?
  • do you want to start a profitable business?
  • do you want to replace an existing workflow within your organization?

be silly

this applies to hobbies, side hustles, even your career

this should link to personal software


you should act on ideas quickly

You may agree that you are wrong about at least one of your beliefs. You obviously don’t know which one, though you know that you must be wrong in some way, and you can conclude you will find out eventually, and so when the time comes, you will learn from it, you will integrate it into your project/product/program, and you will repeat this process for every mistake/bug/error you encounter till you pivot to a different thing (life,company,etc) You may read this and think you can avoid this by preparing thoroughly for a while, until you feel ready, or until the project/product/program feels ready. Or you may read this and think you should just start that thing right now because, like I said, you are bound to make mistakes anyway.

My entire life, I have leaned into the idea of waiting to feel ready to actually start that thing, it’s pretty boring. It’s so easy to consume information and feel like I understand that thing that I’m supposedly learning about. Years can go by before you realize you still don’t know where to start. I am not advocating for being impulsive - you should avoid expensive mistakes. Do not yolo $10k in the stock market if you’re barely learning about stocks, do not start a business without the required permits, and do not deploy. Do not attempt to hack or steal personal data. etc, you get the point It’s also easy lose sight of the goal and start with easiest part of the project, or start with the branding, the logos, the overall vibe

close a deal before having the product (do not commit fraud). “the art of the deal”


Do not start from scratch

start with a brain dump

Start with templates. Automate the boring and repetitive. Don’t learn everything at once Define the metadata for an MVP. What information do you actually need to know? Do not add the nice-to-haves.

copy other people use low/no code tools use open source software - fork it, copy it, learn from it

do not reinvent the wheel the wheel is a primitive, a foundational component start with the wheel start with the engine start with the axis reinvent the shape of the system that results from putting the primitives together


Yes, you can use AI

Use AI: You cannot create a full featured product with AI. But you can prototype many components. Draft text, prototype images, videos, audio. Prototype a UI, a button, a banner, a table. AI is not the builder, you are AI is a component generator, and you must glue its products to make a monster

vibe code


Share it, gather feedback, understand the problem

I solved product-market fit with this weird trick

You are not your users, and that’s a problem. Talk to them, use your own product users do not care about your stack do not over engineer

Users only care about UX

  • This obviously includes UI + functionality. How many clicks do I need to get what I need? This usually follows CRUD. How many clicks do I need to create, or read (get), update/edit, or delete information. Reduce friction. This, actually, makes the app simpler to build

Work on it

You can actually start at any point with preparation without talking to users wit

Deploy the first draft. Take notes and gather feedback


Iterate on it

test it, share again, integrate the feedback again Iterate fast Remove/ fix the prototypes components one by one, as needed.


Decide your next move

do you care about this thing? do you want to keep going? is it still interesting? did you learn what you wanted to learn? do you want to make money from this? can you?


notes: maybe shift focus to AI? nahh who is my target audience here?

  • vibe coders
  • hobbyists
  • people feeling lost

Don’t communicate by sharing memory, share memory by communicating.
Rob Pike1

inspiration

  • this is largely inspired by the idea that we should follow our curiosity and venture into tiny experiments, we have high agency to pursue them, and they must be exciting

eldritch should use AWK to render html

Bibliography

https://www.highagency.com/ https://en.wikipedia.org/wiki/Prototype https://en.wikipedia.org/wiki/Proof_of_concept

Footnotes

  1. The above quote is excerpted from Rob Pike’s talk during Gopherfest, November 18, 2015.

You might also like: