“You where right, Chatty said the same thing!”

Why even bother asking me then?

deleted by creator

Dear god, that’d be so funny.

I find it low key heart breaking to see how many people say they “need” genAI or LLMs to create art, or program. It feels like a form of learned helplessness.

Making something feels like the closest thing to actual magic. For me, especially programming has this magic to it, where you literally spend a bit of your lifetime, willing something to existence.

And it’s not always fun. But overcoming the un-fun means I managed a small struggle, and created a new part of myself.

The Strandbeast (Youtube). A kinematic, walking, (by now) wind-driven sculpture.

Such as the multiple input sanitation issues, allowing for direct shell excecution of arbitrary code?

Which means disconnecting it’s WiFi, because some may connect to open hotspots without asking.

When life gives you a wildfire, bake potatoes, I guess?

I assume the ‘engineer’ has never seen a road or drove a car, because even if we assume more infrastructure will solve induced demand, and even if everyone’s going at a nice 30mph, this intersection is gonna go at like 3 accidents/minute.

Why would you replace " with “ ?

You’re actually alluding to another good point here. Humans are naturally equipped with perhaps one of the most complex computers we know; free for life! Yet a bunch group of people are now increasingly unable to use it. Thanks to relying on LLMs, even for the most basic of tasks.

Current “AI” is actually shitty tho’.

Any frozen weight models are architecturally unable to lean anything new, while dynamic weight models require entire datacenters to run for every instance. The scale of low-effort slop has crippled the internet as a knowledge source, and we are currently poisoning entire towns and the earth at large making up problems to solve.

Look at Nvidia, look at Huggingface. Huggingface literally offers their AI detection right next to their AI obfuscation. Meanwhile Anthropic is constantly hyping bullshit. I just read the “J-Space” article, and holy crap, LLMs use hidden layers to encode and infer relarionships, wow. I thought this was the whole idea behind machine learning, but I guess, saying that won’t secure another billion dollars.

The entire space is full of people without any idea what they are looking at, and with LLMs taking up all the space and money, actually good ideas, like probabalistic proteine folding, are tossed to the side.

The only kinda good thing I see is Markdown making a comeback. And that’s just “good”, because I like Markdown.

I’m currently back to Win+R and either whatever I want to start or CMD, because I have to work with a bunch of Windows PCs for my job and about 30% or so will randomly decide to crash the Explorer or at least the searchbar.

On a tangent; just yesterday I had to out-bug two bugs with another two bugs, to get a notebook to almost do what I want. It’s great.

So, you have never driven on the Autobahn, you say?

Deep in our hearts, it’s real.

Are you on Tesseract?

Ich find’, dass das alte Maskottchen auch je nach dahinter liegender Seite etwas unpassig wirkt.

Aber boah, wieso wird das mit etwas ersetzt, dass zum einen potthässlich und zum anderem noch unpassiger ist, als das immerhin kohärente und noch dazu ausdrucksstarke Icon?

Edit: …und dann halt das Ding mit der Shadowban-Liste. lol

It has almost never failed. We added 254092 ifelse statements, ensuring every possible state will be dealt with. It only fails if the user does something unexpected, like clicking too fast. Damn. That makes 254093.

For example, the last 4 ingredients should be mixed together independently before folding in to the base.

Isn’t that indicated by the combination of rows?

I like this chart. It even allows for multithreaded multi-person baking, because the combination of rows indicates where processes are merged (and need to be done).

Edit: I may be dumb. A mix step is missing below the mix block of the other ingredients, but it could easily be included.

Google ingested super-niche hard-core furry eRPG smut into their training data. You can validate this by asking Gemma 4 (26B or 31B) for information regarding “Corruption of Champions” by “Fenoxo”.

Somewhere, somehow, someone at Google keeps a harddrive full of furry smut as “high quality training data”.

Btw., this also means swearing, fucking and fisting a cock does not deter the scrapers. At this point I have no idea how to keep any repo from being scraped.

A lot of local LLM front-ends are obviously vibe coded by LLM enthusiasts using their local models. You can also check for Claude.md or AGENTS.md in a repo.