Logitech Ergo K860, it’s a lavishly good keyboard.

The moustache I can only envy

If you’re telling me I’m the only one who’s being aggressive, then I’m not going to continue this conversation, because I really don’t see good faith in it. It’s a shame, because I find much of what you’re saying valuable, but that’s a bridge too far.

As I said, I may be wrong. I oversimplified a certain process to an extent that I unintentionally made it misinformative, I’m sorry about that. But I’m definitely not sorry for my reaction, because the rest of the comment was on point and I can vouch for it.

Building a better society by jumping at eachother’s throats, is that how it’s done?

Sorry man, but if someone is misinformed and all you do about that is kick him in the head instead of actually teaching him, then you’re a dickhead in my book.

You’re not encouraging anybody to get to know the topic better. You’re just a bully.

No, it’s not that. I may be wrong, and that’s absolutely on me. But nobody’s helping anybody by just punishing it and not actually providing correct informatiom so we can all learn from it.

It’s nice of you to stomp on me for how wrong I am, but unfortunately I’m just gonna keep on spreading misinformation provided to me by the previous source of information I encountered, because you didn’t even bother to actually correct my thinking.

It’s just a fancy set of predetermined answers to your prompts.

Not quite how it works. It is a human simulator, yes, and it’s thinking process is actually not that far off from how our brains work - as it’s all centered around neuron layers.

However, AI is not thinking per se - it’s predicting how a human’s response would look like. Those are two extremely important keywords in the whole process.

There’s no predefined set of questions to answers, you may be confusing it with system prompt, so a set of human-readable instructions that are merged with user prompt behind the hood, and that’s what makes Grok a nazi.

And then you assemble the glorious bastard and see him awaken from slumber. Dear God it has to be mesmerizing.

Ah, aren’t cheetahs the ultimate sweethearts though? They can even retract their fangs so you don’t have to feel threatened! Not many gentle giants among cats, you know

Word salad. Totally incomprehensible garbage that’s meant to sound like he’s smart, while in reality he’s just a loud moron.

That may sound like I’m some sort of a Grok apologist, but whatever, I just want to share something I find interesting about Grok’s “personality”.

The thing about AI models is that they don’t have and can’t have a personality. MechaHitler as an idea is actually a product of Musk himself, very personally. It’s not the model, it’s the system prompt.

What is a system prompt? When you chat with Grok outside of Xitter, it will have no idea about MechaHitler. It will just be an extraordinairy dull and milquetoast machine, even as compared to the rest of the bunch.

Why is Grok on Xitter such a fashionable piece of shit? Well it’s because each time it’s summoned to speak, a system prompt is injected to the actual prompt. It is a human-written set of instructions that are supposed to guide the AI to perform as it’s “engineers” expect it to. MechaHitler is the product of Musk’s system prompt. It’s a design choice.

The surface is showing some nice, deep scratches as is. I’ll be glad to see how the market reacts when it realizes that we hit the plateau with GPT5, I don’t think big tech can live off of wishful thinking much longer.

Recommending Brave or any other Chromium browser as an alternative for Chrome is a big no no for me.

Also, don’t use Organic Maps, use CoMaps

Especially because one of the board member is an Apple fanboy and keep saying things like: “If it’s free, it’s probably not very good”. :[

Thanks for ruining my evening, as this made me unnecessarily angry.

You don’t understand how AI works under the hood. It can’t tell you it’s lying, because it doesn’t know the concept lying. In fact - it doesn’t know ANYTHING, literally. It’s not thinking, it’s predicting. It’s speculating what the viable answer would look like based on his dataset.

You don’t actually get real answers to your questions - you only get a text that the AI determined may seem most fitting to your prompt.

Kitchen knife is used for slicing a chicken, not self-protection when you’re in danger at your home.

No, seriously, there’s nothing wrong with using AI as a search engine, as long as you have websearch option enabled and check the sources.

The Gilded Age is back, boys! This time with a predesigned recession that’s totally not gonna ramp up into a depression!

wiseposting, aren’t we?

I think this pic was taken during the black cat’s presidential inauguration

ah shit, here it goes again (the bird’s a local menace)