So, cntrl-f search is broken because it's not loaded. I can't scroll down quickly because it does the constant stop-and-buffer routine.

This is just total ass because people have over-bloated the web. I don't need 20-50 MB pictures on a little screen. I don't need all the bloated java bullshit that companies, especially news media companies, are filling their pages with.

This is another symptom of shitty programmers using 100 different pre-made libraries all of which are shitty and bloated to begin with, along with oversize graphics and hundreds of links to third party ad servers all using bandwidth that's utterly unrelated to the actual content I want to read.

I mostly run application fullscreen and switch between them. The only exception is when I'm comparing the content of two windows (in which case I tile horizontally or vertically) and file selection (floating).

When an application uses the entire screen without the window decorations needed in a regular window manager, a screen's limited real estate is in fact better used in a tiled window manager.

Tiled windows don't solve a problem. They're just a different workflow. I've used both for decades and neither is inherently faster or better. It's just what you prefer.

At any rate, don't knock it till you try it.

on all my machines. Once you get past the tiled window manager paradigm - if you've never used one before - you realize how fast and seamless it is, and it truly is the least common denominator in terms of memory usage.

I left Mint (which is really a Ubuntu derivative) years ago, and now i3 / Sway let I have the same unified desktop on all my machines, fast or slow, new or old, and they all feel perfectly usable.

I highly recommend spending the time to create a i3 or Sway config file. It's well worth the effort and it's a one-off.

And if you just want to try i3 or Sway on your existing distro, install it and simply change the Window manager for your user in the display manager: it lives totally independently of whatever your currently use, so it's risk-free.

output is about 6 tokens/s with 16k context window i'm not having any issues since it went live this afternoon. it's not sparkling like opus 4.5/6 but gets the job done

i generally send it a voice note via telegram while driving and then check back in like 1-2 min, or it is sending me a reminder about something on our shared calendar. it's still faster than texting my buddy about making plans for this weekend or whatever.

At no point in the link you provided is any such claim made (that it is one of the most popular choices among players.) Would you like to play again?

I'm using a $200 used ~5 year old (from the ebay listing) HP EliteDesk 805 G6 DM Desktop Ryzen 5 PRO 4650GE 3.3GHz 32GB RAM 512GB SSD WiFi in cpu mode... you don't need a gpu to run single user local LLM... just a bunch of ram. This isn't 2022 anymore

It's about 5 tokens/second which is totally fine for an async assistant. 20 tokens/second is about the lower limit for usable in realtime. You can also set it up to use a smaller model for quick questions (what are the next 6 items on my calendar/to-do list?) and drop through to the bigger slower model for harder questions (can you add this feature to my internal ticketing system and redeploy?)

I ordered 64gb of ram about an hour ago and i'm planning on running either qwen 35B-A3B 8 bit or 122B-A10B 3 bit in fully offline mode.
 
  >the actual cost of 'running the AI.'
 
is a fixed $200 cost (ram upgrade) + electricity

I cancelled my subscription overnight, and I'm using the free credits they gave me to wrap up some things and transition away. I am not going to be locked into someone's walled garden again.

It said your choice was "unusual", not "poor". I would wager that AI has a better idea what actual choices most players make playing that particular online game than you do by basing your belief on the idea that your choice is the usual one.

The problem isn't the tools; it is the people. When AI is better human critical thinking and deductive reasoning capabilities will decline further, not magically return to a higher measure.

This, exactly. People are assuming that correlation equals causation here. IQ and critical thinking skills, as well as reading comprehension ability, have been on the decline for quite some time. AI is actually much better at getting things right than the average person who uses google to feed their confirmation bias and find echo chambers in which they can bask. I'm not going to post a bunch of links, because anyone can use the prompt "studies showing IQ and critical thinking in decline" in your "favorite" agentic AI. A hammer, the internet, and AI are three tools equally capable of constructive and destructive uses, and this study just shows that the same trend that started with technology in the hands of the incompetent is merely continuing as expected.

I don't know what AI system(s) you are using, or how long it has been since you tried, but Claude is extremely impressive and while it isn't perfect it is far more effective than google searching. I recently used it to troubleshoot a guitar problem I had with dead notes when fretting on the first fret and it was quite competent. I was able to interact with it and gain a solid understanding of what was causing the issue, including the physics behind it. Then it helped me find a great repair shop to replace the nut, after explaining the process in detail, including a temporary work around involving shimming the nut. As far as programming goes people need to read this before they keep spouting the old "AI can't work" lie.