camwest

Yeah, I am very open to skills and REST APIs, but REST APIs have an issue where the token gets injected, and so it's not really a good auth story for everyone. I think that skills have a feedback problem: how do you refine your skill that you've shipped to your customers? The suggestion you are giving isn't great for something like the ChatGPT or Claude iOS app. Where would you even store the token in that case, for example?

https://cameronwestland.com

I wrote this piece to share my view about what cursor should do to manage some of the negative sentiment they’re getting online. What do you folks think?

For me it's not about difficulty, it's about friction. BJ Fogg's behavioral model B=MAP says Behavior = Motivation × Ability × Prompt. When you increase ability (lower friction), you get more behavior for the same motivation.

AI lowers my writing friction. Same motivation, more output. I write more, I get feedback faster, I iterate more. That's the value for me.

Prior version was flagged. https://news.ycombinator.com/item?id=46325286

(No AI slop edition. See https://x.com/cjwestland/status/2002049554133229924?s=46 for context)

Same, logged an issue: https://github.com/zed-industries/claude-code-acp/issues/18

Agreed. I'm happy they're training on my data.

My reasoning: I use AI for development work (Claude Code), and better models = fewer wasted tokens = less compute = less environmental impact. This isn't a privacy issue for work context.

I regularly run concurrent AI tasks for planning, coding, testing - easily hundreds of requests per session. If training on that interaction data helps future models be more efficient and accurate, everyone wins.

The real problem isn't privacy invasion - it's AI velocity dumping cognitive tax on human reviewers. I'd rather have models that learned from real usage patterns and got better at being precise on the first try, instead of confidently verbose slop that wastes reviewer time.

Never? No. Way less likely? Yes!

In dev we do 100 consistency checks and get green. In CI we do 10.

This feels similar to a GitHub issue.

1. Editable description 2. Comments

A couple of example hooks: https://cameronwestland.com/building-my-first-claude-code-ho...

I'm happy to see Claude Code reaching parity with Cursor for linting/type checking after edits.

I agree it's not about the 10x engineers or the greenfield. I think YC's selection process is still focused on finding distinguished individuals, but within two specific constraints.

I've been using Figma to build my decks for a while now, and I bet a lot of other folks have too. Figma Slides looks like it takes what we've been doing and makes it way smoother.

Nice job on your deck. Startup seems cool.

Makes sense.

Congrats on launch! It’s inspiring to see you building in the open.

Why build Nest on the server when you already have Next for the client?

I’m curious how this is different from Zapier which has added a conversational layer (https://zapier.com/ai).

In my experience a lot of people don’t know how to write with a level of specificity needed to map to pre defined triggers.

For example: instead of saying spreadsheet they may say “roadmap” or instead of saying “Notion database” they may say “bug tracker”

This stuff fails in Zapier.

Any chance you handle those cases better?

Good point. The Chat GPT example was really only meant as a statsig test in a fun way.

You're totally correct in saying that forgetting the effects on other parts of the business is where a lot of the analysis needs to be directed.

Yeah exactly, it was some sort of seasonality effect that, or just plain randomness.

Agreed it's super hard. I think that's why you need to be careful about looking at data until you've got your hypothesis. Too much data and you'll overfit or introduce bias. I think of it more as a way to update your gut with an updated map of the state.

No definitely not. Kafka is for LinkedIn scale. Postgres is for everyone else.

Quick anecdote. I remember a friend who joined Shopify relatively early. He took ALL their data from many different services (analytics, Kafka etc) and loaded into Postgres on his laptop so he could query it efficiently.

It's amazing how over-engineered some startups are.

I think fun is the key difference as well! Even when it takes longer it’s a lot more engaging.

I think you've hit the nail on the head. Balancing is crucial, and taking calculated risks can help set you apart. It's all about skating to where the puck is going to be, as you said.

Great question.

Thanks for sharing the link to the post on unleashing creativity—I hope others find it helpful as well.

I completely agree that any form of harassment is unacceptable. As I've shared in another recent post, I'm committed to addressing my past mistakes, learning from them, and ensuring a safe and inclusive environment in my future workplaces.

I'm glad you found my other blog post on accountability and growth.

As you can see, I'm working on taking responsibility for my actions and working on self-improvement. I appreciate you taking the time to explore my writing.

Hi! I’m the tech lead on AutoCAD web. The engine is C++ which is compiled to Web Assembly, our core team doesn’t actually work in JavaScript day to day.

The browsers are doing a really good job making this a possibility today.

I have a talk about these ideas in the Context of a React Application last year at React.js Conf - https://youtu.be/395ou6k6C6k. It might be worth a watch if you’re interested in how to build applications like this on your own.

Yeah I think I read it here: http://shop.oreilly.com/product/0636920001416.do

Don't quote me on this but I'm pretty sure gmail uses it. I'd say that's a pretty big endorsement.

I wish Amazon used this for their sale of the Classic NES (https://techcrunch.com/2016/11/11/nintendowned-amazon-sells-...) flash sale.