coffeecoders

I agree. Specialized engineering is very different from software, you can't easily replace that expertise.

My question is more on software roles, those CRUD or React component roles, where the work can be done remotely. Instead of Americans filling those positions, companies may/will simply hire those engineers abroad.

But look at the other side. An engineer living in Seattle pays taxes and contributes to the economy. Seattle loses that benefits if that same job moves to India.

So, I am not sure if that housing argument alone makes that a better outcome for America (and workers).

I know this is a much popular opinion now. But I can't help and ask this again:

If a company hires an engineer in India, instead of Seattle, how does America (and workers) benefit? Software doesn't require factories, or supply chains, its just a hiring decision.

The question is whether these restrictions increase American employment or just accelerate the exodus of jobs while making the US a less attractive place for talented people.

What is interesting to me is that stripping the C2PA data is easy, but faking it is hard.

You can resave the file and the "made with Claude" signal disappears, but you cannot make a random file pass as Claude-made without Anthropic's signing key. So the useful guarantee is one-way. No signature means almost nothing.

One place where I find the Flash models surprisingly bad is Google Search's "AI Mode".

A recent example - I searched for how to unsubscribe from Pearson emails. Google Search "AI Mode" confidently gave me a sequence of steps along the lines of Settings > Profile > Email preferences > Unsubscribe.

Of course, I looked for an unsubscribe link before asking Google. None of those options existed. The correct answer was there is no way to unsubscribe through the account, so I just blockthe emails instead.

I've run into this pattern quite a few times. AI Mode seems to make up things all the time.

It only makes sense if we are measuring raw capability per dollar. Once you factor in noise, physical size, idle power, battery etc, I think the calculation changes.

For a home sever that's mostly idle, it is almost like commuting in a pickup truck.

Just add this to uBlock origin:

``` linkedin.com##main#workspace section:has(div[componentkey*="container-update-list_mainFeed-lazy"]) ```

I think the problem is concentration. AI now absorbs almost all of the oxygen from every category. A story about new chips becomes an AI story. Cloud pricing becomes an AI story. Even personal projects get pulled into the same orbit.

I need this but for dentists.

If I am archiving PBs of data for 10+ years, I don't want to rely on a WASM interpreter being available and performant in the future just to read a file. I want a dead-simple, heavily documented byte specification like Parquet.

Additionally, putting the decoding logic inside an WASM binary introduces an active execution layer into what should be a cold storage.

It's a selection bias issue. The categories you have listed are essentially web services wrapped in an app shell. Of course they need the internet. Consider these examples:

- Photo/Video editors - Snapsheed, Lightroom, Video trimmers etc.

- Document readers & scanners - PDF viewers, e-readers, OCR scanners

- Note taking - Obsidian

- File/Password managers - Authenticators etc.

- Single player games - Chess, puzzles etc.

- Audio/Video players - VLC players

We've just become conditioned to accept that every app needs to phone home for tracking and ad-delivery.

This is excellent. Seeing this makes me appreciate how much visual awareness tools like this are needed.

I built something similar, for the web. https://neberej.github.io/exposedbydefault/

Github: https://github.com/neberej/exposedbydefault

We (software engineers) get better outcomes from the same algorithms by improving data flow, constraints, instrumentation etc. (Better) prompting, retrieval, context engineering etc seem like the LLM equivalents.

The model weights haven't changed but the system is making more use of the capabilities already present in the model.

I've been spending my time away from screens, on woodworking, mostly small projects like plant pot holders.

One thing I've learned is how hard it is to get things perfectly aligned.

I've accumulated a bunch of old website archives over the years. The funny thing is the ugly HTML dumps have been more useful than the "perfect" archive.

It's one of the reasons I've become a bigger fan of RSS over time. A feed from 10-ish years ago is often more usable today than a carefully preserved (application) website.

I've shifted to a "slow code" approach with AI, treating it more like a design partner than a code generator.

I mostly do TDD with TypeScript. I write the test, write the code myself (sometimes with the help of LLM), and then hand it to the LLM. Instead of asking it to write things for me, I use it to find edge cases, check if it's leak-proof, and verify efficiency.

For architecture questions, I debate with it for a while. I almost never ask for code without conversing 4-5 times first to push back on its assumptions. It's the best rubber-ducking partner I've had.

Personal plug: I wrote more about why/how I use AI to write slow, better code on my blog: https://nabraj.com/blog/ai-write-slow-better-code

Me: Senior Full-Stack Engineer with 10+ years of experience building high-traffic web applications and scalable backend systems. I've worked exclusively with US-based companies.

My blog: nabraj(dot)com

Looking for: Senior full-stack, backend, or platform engineering roles.

Location: Vancouver, BC, Canada. Open to remote US opportunities.

Tech: TypeScript/JavaScript, Python, React, Next.js, Node.js, GraphQL, PostgreSQL, MongoDB, Kafka, RabbitMQ.

Email: ntemail@protonmail.com

That pelican looks like it lost 100k on NFTs and now runs a paid stock-trading group.

This is the reason. I have just been vibe-coding my way for a few months now, got almost all the tools (except Browser and Mail) that I use daily, designed by me (with the help of LLM).

For me, its those Who's hiring or Who wants to get hired posts. I used a throwaway email once and got emails about SEO and AI projects.

I don’t engage. I mark as spam, block the sender/domain, and move on.

I built a tiny Notepad clone in ~5 minutes using an LLM: open/save, plain text, no surprises.

Lately I've been doing the same for other small utilities. Roughly half the little tools I use are ones I generated and kept because they’re predictable and easy to audit.

The point isn't replacing built-ins; it's reducing dependence on shifting defaults. I want to care less about what the software/os vendor changes this time.

Slightly related, I have been writing all my local tools with the help of AIs.

https://news.ycombinator.com/item?id=47006108

Why not install just the Messenger app? I never install Facebook app, but I keep Messenger app to chat with my college peeps.

I am a fan of https://blogs.hn/. It is mostly HN-like content, but I visit it daily. I wish there was a "new" view though.

Agree on that. HN is a good example of resisting that pull. It hasn’t optimized for engagement loops.

Credit to @dang and the moderation team for keeping it that way.

It's been fantastic. The mouse puts my hand in a natural "handshake" position, which has cut down on the wrist strain I used to get after long hours of work or browsing.

I'd highly recommend giving it a shot.

For work: Vertical Ergonomic Optical Mouse, Anti fatigue mat

For home: Bidet

For personal: Kindle

I am a big fan of https://blogs.hn/ (I visit it daily), your directory looks similar.

You should automate this, maybe drive all of these content from a json file and accept PRs.

---

My blog: https://nabraj.com/

Most popular: Why is boarding a plane still a mess? (https://nabraj.com/blog/boarding-methods)

This suggests HN is functioning as designed. Votes signal agreement while comments surface disagreement.

Negative posts outperform because they create unfinished cognitive work. A clean, agreeable story closes the loop, a contested claim or engagement opens and follows the open loop.

Nice, Linear is a perfect target for this mindset.

The REST client went surprisingly smoothly once I committed to keeping it boring.

I'm building Mac apps in Xcode, and I keep multiple small apps in a single Xcode project with a few shared libraries (basic UI components, buttons, layout helpers, etc. to keep all my apps similar).

The REST client is literally just another target + essentially one extra file on top of that. No workspaces, no collections, no sync, no plugins. Just method, URL, headers, body, hit send, show response. Requests are saved and loaded as plain local JSON.

What surprised me is how little code is actually required once you strip away "product features".