jmathai

We haven’t fully proven that it is any different. Not at scale anyway. It took a decade for the seams of outsourcing to break.

But I have a hypothesis.

The quality of the output, when you don’t own the long term outcome or maintenance, is very poor.

This is not the case with AI in the same sense it is with human contractors.

I've seen some trim with channels for wire. They're not very common in the US - I've never found them at The Home Depot.

For my projects, I would have to replace the trim for the entire room and possibly adjacent rooms because the baseboard is the same throughout the house.

But if you're okay with or planning to replace all of your trim and could get it off-the-shelf then certainly that's a viable option.

For Claude Code, I now pay the $20/mo subscription for pro because I was spending more using it via API credits.

Even if I had to reload manually very often, I still would not enable auto reload. These APIs are crazy expensive and I'm not looking for a surprise bill.

Are you all enabling auto reload for personal projects?

I load $20 at a time and wait for it to break and add more.

Skills have frontmatter which includes a name and description. The description is what determines if the llm finds the skill useful for the task at hand.

If your agent isn’t being used, it’s not as simple as “agents aren’t getting called”. You have to figure out how to get the agent invoked.

3d printers are great.

I attempted to productize a super niche idea I had for myself that I have ended up using multiple times. It's a simpler method of running wires without having to cut/patch drywall and drill through studs.

It started with trying to create a router-like device equipped with a motor (which was version 1 [1]).

But then iterated it to be a plate which can be attached to existing routers. That resulted in a drastically improved user experience which went from doing it all inside and making a mess to doing it in your garage with more precision. I've used the final design on 4+ projects at home and it has saved me so much time [2].

[1] https://www.dropbox.com/scl/fi/i5p330nqo0cfrfx3gxf05/Wiresha...

[2] https://trywireshark.com (includes diagrams and videos of what I 3d printed)

Small? Gawd, I hate doing that for every single Zoom call I have to join.

Meh, fixed. Replaced with pastebin links in and kept the repo private.

Volume and statistical significance? I'm not sure what kind of narrative I would trust beyond the actual data.

It's the hard part of using LLMs and a mistake I think many people make. The only way to really understand or know is to have repeatable and consistent frameworks to validate your hypothesis (or in my case, have my hypothesis be proved wrong).

You can't get to 100% confidence with LLMs.

The most useful automation for me has been a few simple commands. Here are some examples I use for repos in GitHub to resolve issues and PRs.

/gh-issue [issue number]

/gh-pr [pr number]

Edit: replaced links to private github repo to pastebin.

https://pastebin.com/5Spg4jGu

https://pastebin.com/hubWMqGa

You may also be getting a worse result for higher cost.

For a medical use case, we tested multiple Anthropic and OpenAI models as well as MedGemma. Pleasantly surprised when the LLM as Judge scored gpt5-mini as the clear winner. I don't think I would have considered using it for the specific use cases - assuming higher reasoning was necessary.

Still waiting on human evaluation to confirm the LLM Judge was correct.

Try to turn off YouTube watch history.

You get no feed at all. It’s marvelous.

> when Obama deported 3M+ illegals, the sanctuary cities helped and the media didn't cover it like it was Nazi Germany

Perhaps it has to do with the method used?

I don't think most people are in favor of illegal immigration. That would be an extremist view. Most people are probably somewhere in the middle and I don't think that band is too wide.

I am super curious about this. I wonder baseline it needs to meet to pull me away from using ChatGPT or Claude.

My usage of it would be quite different than ChatGPT. I’d be much freer in what I ask it.

I think there’s a real opportunity for something like this. I would have thought Apple would have created it but they just announced they’ll use Gemini.

Awesome launch Moxie!

Congrats! It's not easy to build something people want and will pay for. It's even less easy to do it for 10+ years.

That's all I wanted to say - as much of a milestone as version 10 is - the past 9 were amazing as well.

I think it's worth understanding why. Because that's not everyone's experience and there's a chance you could make a change such that you find it extremely useful.

There's a lesser chance that you're working on a code base that Claude Code just isn't capable of helping with.

I can relate to the author. I have an actively used GitHub repository which doesn't need any more features. It's got 1.3k stars but the commit history has been sparse.

Even I look at the last commit date as a proxy for "is this maintained". I wish GitHub had something similar to "Archived" for projects like these.

Aren't those just the URLs in google search results if you copy from the results page instead of clicking through to the destination?

Thanks to LLMs, I have revived personal projects that became too time consuming and complex to familiarize myself with again.

While I enjoy the challenge of writing software, I more enjoy having the thing which does exactly what I want.

LLMs are amazing for this.

Very cool! I do this today by connecting my phone to my laptop and recording via Loom. It's a bit janky for sure and the video is off because I looking down at my phone and not the webcam [1].

Does Loom really not offer this capability as an iOS app? I never checked but that's crazy if they don't. Regardless, I think this solves a problem for sure! Congrats on building and launching.

[1] https://www.youtube.com/watch?v=11yHdB-wmgs

AI doesn't need or care about "high quality" code in the same ways we define it. It needs to understand the system so that it can evolve it to meet evolving requirements. It's not bound by tech debt in the same way humans are.

That being said, what will be critical is understanding business needs and being able to articulate them in a manner that computers (not humans) can translate into software.

More important than code quality is a joint understanding of the business problem and the technical solution for it. Today, that understanding is spread across multiple parties (eng, pm, etc).

Code quality can be poor as long as someone understands the tradeoffs for why it's poor.

I found this trick for store bought pizza dough as well. Instead of leaving out for 20 minutes, a warm oven helps it start rising a bit and results in a much better final product!

Those who are still employed by my local grocery store don't seem too motivated to do their jobs. It's an unfortunate but foreseeable scenario that expedites the problem where people prefer self-checkout.

I wish we'd pump the brakes on efficiency and profit.

45 years young.

The two times I did not have employer sponsored health care AND $0 income, the ACA plans were more expensive than getting a plan directly from the insurance company. It never costed me $27,000 though.

Most recently, in 2024, I had medical and dental coverage for my family of 6 for around $1,200/mo - roughly half what's quoted but again, family of 6. Wasn't the greatest insurance but it would have kept us from financial ruin in case of a catastrophic event which is all I expect from insurance anyways.

It's very unfortunate. As a parent, I feel like it requires regulation at the national level because I can't win against Meta (FB, Insta), Google (Youtube), Snapchat and TikTok.

That website reminds me of the one…where a guy sold pixels on a website for like a buck and people basically bought ads.

https://en.wikipedia.org/wiki/The_Million_Dollar_Homepage

heh...see my comment above about bypassing screen time limits on ios by simply deleting and reinstalling the app.

Found out last year that simply deleting an app and reinstalling it will reset time limits.