This is an extremely broad topic. There is no one singular "AI". These mentions of "AI", each one, are often referring to entirely different things. For instance, CorridorKey is an "AI" based green screening technique that supports transparency and is a hell of a lot faster and more accurate than just about any other traditional tool out there. So is this the "bad" kind of "AI"? Or is this (and similar) just new tools in the tool box to speed of particularly boring and repetitive tasks nobody actually wanted to do? Masking a green screen is hardly on the "artistic" side of things, its just technical work destined for automation.

https://github.com/nikopuering...

1) nobody wants to read an "article" that has more ads than content when researching a topic (yes, ablock exists for us techies and I have it, but talking in the more broad common person sense).

2) nobody wants to read a forum thread with over 100 posts to try to find the one post with the real answer surrounded by a collection of bad answers and people saying "YUP, IT WORKS NOW" with zero context otherwise.

Make your content on-topic and valuable, and users will return. it is as simple as that. These are just two of the main things I see as devauluation of a lot of these web sites complaining about a drop in traffic. They're most certainly not the ONLY reasons, but it is two of the major points!

If you think that simply because there are 3 possible outcomes, that there is perfectly even distribution of said outcomes, I think you should go back and attend an entry into statistics course again.

Ignoring sales figures for a moment, I'm honestly perplexed why everyone is missing the glaringly obvious reason for the death of optical media

1) storage density.
2) storage bandwidth.

Nobody wants to wait 20+ minutes to load a single level in a game. Developers are making games larger and larger, well beyond the capacity of a BluRay disk. Optical density has not kept up with increasing storage demands. The technology simply isn't there today, and there is little incentive to continue to invest into higher density disks because of the cost reduction in online distribution.

Do I hate this reality? Sure. But do I understand the economics of it? Totally.

I'm confused by this. Apple ALREADY offers financing through their own Apple Credit program to purchase hardware. So, why then outsource what they already have in-house and working great where they get to keep everything themselves?

"OnStar brought in about $800 million in the second quarter"

Meta (Facebook/Instagram/Threads) and Alphabet (Google/YouTube) has machine learning to train models that then their ad networks use inference on to deliver you advertising across the entire internet.

This is one of those R&D cost things. Those companies entire bottom lines depend on how competitive their advertising networks are. They invest in AI/ML and get the return elsewhere invisibly.

Is it truly paying off at the scale they're investing in right now? Well, maybe... but it certainly is no where near the losses being advertised right now.

Drivers!? Who needs drivers? Just netcat the PDF output directly to the printer! Printers natively have their own printing drivers and capabilities built in, and will translate documents internally.

https://retrohacker.substack.c...

NT4 wasn't a consumer OS though. So of course they didn't care.

Qwen works. It does what I need. It does it well. I can run it locally on my own hardware.

FUCK YEAH I'm using a Chinese model because like hell am I paying per-token costs from the subscription model companies!

Anyone running non-Windows is most likely running an agent switcher because the number of bumfuck dumbass web sites still doing OS checks to see if a browser is compatible or not with their basic HTML bullshit. This is one of the absolute largest annoyances of the internet.

Apple MacBook Pro, M5 MAX, 128GB RAM, 2TB SSD

Unproven? This is the technology directly responsible for the onslaught of zero-day exploits in every major F/OSS (and even some closed source) ecosystems all over the internet. Go check the CVE news from the past couple months, its all Mythos up and down entirely. This is literally the tool cracking every major codebase on the internet, so calling it "unproven" is a bit of a stretch at this point.

Chances are this is a configuration issue.

For instance, I recently saw a case of this with LMStudio. It turns out that LMStudio doesn't properly handle K/V caching when paired with Claude Code. The solution? Literally the exact same model, but using Ollama as the server application instead. Problem instantly solved.

The other big reason for SSD thrashing is not having enough RAM to handle the model, context, and cache. If you do, then your disk shouldn't be touched at all for model work other than initially loading the model. These things are designed to entirely run from RAM because disk, even the fastest NVMe drives, are slow as shit in comparison. For reference, on my test laptop, it does ~15GB/sec from NVMe, but well over 600GB/sec from RAM. The NVMe speed overall is great for filling content into RAM but once that is done, it is never touched again (save for basic small text/log files)

So just run your own inference infrastructure then?

A traditional "developer" laptop generally costs around $2000-3000 for high end engineers. An extremely capable inference laptop, otherwise doing the rest of the same jobs, is around $5000-6000. One time cost. Everything runs locally, which also solves the "do I trust LLM vendor?" issue.