Hacker Newsnew | past | comments | ask | show | jobs | submit | lelele's commentslogin

There is no such thing as "desktop Linux". What we have instead is a large collection of distros, each with its own UX, unlike Windows or macOS which present a relatively unified platform.

I switched to Linux many years ago because a new laptop was unusably slow under the Windows Vista it came with, and I have not looked back since, yet I'd never recommend Linux to "the masses". Linux can work well for people who just browse the web and read email. Beyond that, the experience quickly becomes dependent on having a knowledgeable person nearby to help with choosing software and supported hardware or troubleshooting it.

To me, articles like this show how disconnected many technically inclined people are from average users' experience. Things like bloated software or aggressive advertising may be annoying to us, but to most users they are just part of using a computer.


- AI Chat in VS Code

- ChatGPT Web interface


Would you mind sharing your current workflow?

For example, do you begin with a rough design and refine it into concrete steps with the AI, or take another approach? Do you switch models based on task complexity to manage costs?


The Copilot extension in VS Code includes Opus as well. It costs three times as much as Claude, so I'd expect it to perform better or be able to handle more complex tasks, but if you're happy with Claude - I am too - more power to you.

s/Claude/Sonnet/

> I use Claude and ChatGPT

Both for code? For me, it's Claude only for code. ChatGPT is for general questions.


Yes, I use them in tandem. Generally Claude for coding and ChatGPT when I run out of tokens in Claude.

I also use ChatGPT to summarise my project. I ask it to generate mark down and PDFs, explaining the core functionality.


> One day when the well runs dry, you must be able to either pay the actual cost

What multiple of the current cost do you expect? Currently, GitHub Copilot and ChatGPT for Business cost $19/month and €29/month respectively. Even a 10×–20× increase will still be economically viable in a professional setting if the tools continue to save hours of work.


These tools (e.g. Chatgpt pro) lose money at $200/month

So expect, maybe, $1000 a month? Until your business is dependant on these LLMs. Then they can extract basically all your margin lol


At a $1,000/month price point, wouldn't the economics start favoring buying GPUs and running local LLMs? Even if they're weaker, local models can still cover enough use cases to justify the switch.

Claude max is like $100/mo, and if you’re a daily user you’re likely going to need max

Remember when Uber was cheap and showed you the surge pricing multiple?

Do you mean Windows 1x Pro/Enterprise?

Yes. Enterprise, Pro, and Home are the enshittified, retail editions. Enterprise just adds a few more features IIRC but still has ads. The other versions I mentioned above don't have any of that.

Enterprise is not retail and is usually done via volume licensing, but probably without any additional configuration it might have that stuff intact.

But you can use group policy etc. freely. I don't know how Win 11 is though


FWIW, ChatGPT advised against LTSC or Server editions for a dev workstation and recommended Enterprise, as you do. However, I can’t find Enterprise from a reputable EU vendor. Do you know of any? Is Enterprise available to end users?

It's not officially, but nothing prevents you from buying licenses.

I don't use Windows anymore but iirc the easiest way is to get the E3 or E5 licenses. The volume licensing is "Contact us" pricing

LTSC is also Enterprise, but it's meant for e.g. computers attached to an industrial machine/line that rarely gets updated and such. But it's used by many prosumers as a way to avoid bloat and e.g. keep Win10 for longer


I evaluated Kagi a couple of months ago, but its results were swamped with AI slop, too. These days, it's mostly DuckDuckGo for me.

DuckDuckGo has also been my goto for years, but it is also getting swamped with SEO-rigged spam sites.

2 days ago I was making multiple searches looking for the websites of specific hotels and spa facilities, and none of them showed up in the first 4 pages of results, even when searching by exact name.

Out of desperation I switched back to Google, and, surprisingly, it was willing to give me what I was looking for on the first page. (But not as first result, of course..)


> DuckDuckGo has also been my goto for years, but it is also getting swamped with SEO-rigged spam sites.

True, but DDG seems, at the very least, no more afflicted than other engines. If I'm supposed to pay for a search engine, I expect something better than DDG.


> With Sonnet, if you aren't clearing the context very often, you'll hit it within a few hours.

Do you mean that users should start a new chat for every new task, to save tokens? Thanks.


Short answer is yes. Not only is it more token-friendly and potentially lower latency, it also prevents weird context issues like forgetting Rules, compacting your conversation and missing relevant details, etc.


Yep. I have Claude snapshot to a markdown doc with key points and resume and iterate. Saves so many tokens.


Yes, it also helps keep it focused.


An interesting take on how to construct the most durable buildings: not with the strongest or most flexible concrete but with concrete that self-heals. I'm posting this here because perhaps such a lesson could be applied to other domains


And no iron rods.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: