HN user

zormino

162 karma
Posts0
Comments66
View on HN
No posts found.

I actually agree, Linux is well past the point a minimally tech competent person can use it fine, but it doesn't solve the fact that even if Linux was flawless, there is still a switching cost in time, relearning a new system, and worst (best?) of all of you decide you are willing to do all of that now you can get entirely lost in the weeds picking a distro. I used Linux all through university, then went back to windows out of convenience and needing to use it for work anyways.

Until one day I got so frustrated with constant settings resets, reboots at the worst times for software updates that fail, highjacking my pc after every update for a guided tour of the latest things Microsoft decided to break, and telemetry that can only be disabled with an obscure registry hack that changes every few months, I just couldn't anymore.

Linux has been good enough as a daily driver for a while now, but even with proton I don't know if the pull factors towards Linux will ever be strong enough for most people. For me though the push factors away from Microsoft had gotten so strong I couldn't take it anymore.

I really like the 3d printer analogy. You can still make some pretty cool stuff, you can make some pretty complex things if you carefully design the whole system and put in the effort to print each part individually, and quality depends both on how good of a 3d printer you have and on the proper use of it. "3d print me a new house" is still a pipe-dream: you'll get some miniature facsimile of a house, sure, but a proper house requires proper tools and expertise.

I also think some of this stems from the default 1m context window. Performance starts to degrade when context size increases, and each token over (i think the level is) 400k counts more towards your usage limit. Defaulting to 1m context size, if people arent carefully managing context (which they shouldnt ever have to in an ideal world), they would notice somewhat degraded performance and increased token usage regardless.

Because most people work for someone else and don't decide their own salaries. It's not doubling productivity, but even a 10-20% boost to productivity for a team of engineers means that, as a business, even $1k per month per seat is perfectly acceptable. For consumers and hobbyists that basically kills access.

I think removing Claude Code from the $20 tier is a terrible idea, I never would've gone from nothing right into the $100/200 tier. The $20 plan let me get my feet wet and see how good it could be, and in less than a week I was on the $100 plan.

I think they need to at least have a 1 month introductory rate for the max plan at $20, or devs that decide to try out agentic coding just won't go to Anthropic.

That leads to downstream impacts, like when a company is deciding which AI coding tools to provide and the feedback management hears everyone is already used to (e.x.) Codex, then Anthropic starts losing the enterprise side of things.

My biggest worry isn't that it will make me dumb (it won't), or that it will make me lazy (it will), but that people raised with it wont learn things in the first place. I'm split on if this is a real issue or an old man rants about slide rules and the decline of mental math kinda situation.

Claude Design 3 months ago

You mean you aren't still using Google Duo and Allo? Google Reader? Playing games on your Stadia? I'd be worried about really locking into a specific Anthropic product at this point other than Claude Code

Cursor 3 4 months ago

This is why I use Claude Code though, it pairs well with a regular old text editor (in my case Sublime). I've always had an editor and a terminal open, plugging an AI into my terminal has been a fantastic enhancement to my work without really anything else changing or giving up any control.

That's what you should be doing. Start from plain Claude, then add on to it for your specific use cases where needed. Skills are fantastic if used this way. The problem is people adding hundreds or thousands of skills that they download and will never use, but just bloat the entire system and drown out a useful system.

When I was in university, I had a math teacher that brought extra chalk to class everyday because if he saw you on your phone, he'd snap a piece off and throw it at you pretty damn hard. Maybe that wouldn't exactly work in a high school but damn if Dr Murphy didn't have me paying attention in that class.

Also in 1950 the population of the US was 151 million, today it's 341 million. it has more than doubled, and the amount of space has stayed the same. More people competing for the same amount of property will always lead to inflated housing costs beyond what inflation would predict.

NotebookLLM. Load in technical manuals and datasheets, then ask it questions and check the parts it references instead of searching tens of thousands of pages across dozens of diferrent documents. It's the most useful AI tool I've tried so far for embedded work.

I rarely hit the brake pedal in my car at all with one pedal driving. In fact I'd say 80% of drives I don't hit it at all, and the rest i might use it once or twice. I'm sure I'd still hit the brake in an emergency since it's so ingrained in me from driving normal cars for so long, but I'm not sure how long that will be true for, especially for new younger drivers that have really only driven EVs.

In a later tweet, he even mentions there were 6 other people and they built on top of existing work. Doesn't really sound like he's within a mile of solely creating anything, no matter what spin he likes to apply.