Because they're smart to know it makes them about 2.5x more likely to be killed if something happens involving someone else with a gun.

Sat phone, though- not a bad idea.

"Humans are good at picking out patterns."
...
Demonstrates way that humans are not good at picking out patterns.

Always cracks me up. False positives are still failures.
Human history doesn't give me a reason to think we're particularly reliable pattern recognition networks at all.
I do agree that in our heads, we have some tight neural circuits that are quite good at pattern recognition, though. They're just not attached to our conscious thoughts.

Apple Silicon can't realistically "replace" a discrete. Rather, they're... different.
The compute performance of Apple Silicon is vastly inferior to a mid-range discrete. Its bandwidth isn't great in comparison, either.
So, in terms of GB-of-VRAM-to-GB-of-VRAM, Apple Silicon is worst than any discrete you're likely to have for ML purposes.
However, they've got something you can't get on a discrete- 128GB of VRAM in a laptop, and 512GB of VRAM in a desktop.
This changes the equation, because it means your Apple Silicon (with enough RAM) can simply run models that the discrete just can't*

So in terms of "being able to run a local model of size X", where X>32GB for a top-end NVIDIA, or far less for a mid-grade, Apple Silicon is competing with datacenter cards, and clusters, at least as far as models it can actually run**
As for the NPU, they're useless for bandwidth-bound tasks (which means LLMs), and for non-bandwidth-bound tasks, they generally perform worse than the GPU, though much more efficiently. They also have the drawback of usually requiring interaction with weird frameworks (while compute shaders are generally well understood)
I would not say its GPU is designed for client-side processing, as unless the model is big- any discrete will do the job drastically better.
That being said, I have an M4 Max with 128GB that I purchased specifically for local agentic LLM testing and development.

* Not strictly true, but effectively true since the performance is generally as bad as doing it on your CPU alone.
** Technically, AMD has the AI Max 395+, which kind of competes with an M4 Pro, but not an M5 Pro, M4 Max, and absolutely not an M3 Ultra.

Prove?
Transistors aren't vacuum tubes- it's folly to think you could implement a computer on them. Ask drinky- he can prove it.

Seriously, and again, you're too fucking stupid to have this conversation.

I can do that if you like, but you and Gweihir would have to stop rubbing each other off in your little echo chamber of stupidity.

SRAM has never been built at this scale, afaik. Cerebras was ahead of the curve here, building wafer scale SRAMs years ago. The penalties of DRAM (even with HBM) are now so severe that everyone is taking the gloves off and building mighty SRAMs. This has always been possible in theory, but the high cost never justified it.

The impact on semiconductor fab demand is significant. SRAM cells are larger than DRAM bits: more silicon die area for the same amount of RAM.

Also, the training vs. inference split Google is baking into actual hardware is a big deal: it's the reality that training and inference are very distinct things asserting itself, which has been obvious to anyone that hasn't been drinking excessive NVidia cool-aid: there is a future where costly, general purpose GPU-like devices aren't actually necessary for operating LLMs.

No.

Improvements have been iterative. They haven't just now reached a magical threshold where that opinion is now wrong. It's been wrong for a while.

And vulnerabilities that no formal verification in the universe will find, but any LLM in the world will immediately.

Bold claim.
Bold, and potentially wrong.

False.

Can't tell if gaslighting, or honestly stupid.

It's never a waste of time to confront a partisan shit-for-brains.

Wrong answer, shit-for-brains.

Ah, there it is.

From my perspective the problem is those fascists, and the illiberal shit-for-brains like yourself that don't realize just how little sunlight exists between you and them.

No, it's not.

Inference, but yes.

It's also what your brain does.

The flaw is in your statement.

Is a provably false statement.
Humorously enough, because- as you mentioned elsewhere- of incompleteness.
You also then said,

Which is also trivial to prove false.

You should probably step away. You have made yourself look incredibly ignorant.

Dumbshit beating on a strawman, just another day ending in Y. I love it.

Drinky, you are way too fucking stupid to be having this conversation.

No, it isn't.

Correct.

Wait- wasn't it you who said "The only tool that is able to find all bugs in a piece of software is formal verification.".

lol- you fucking idiot.

Aye- the statement only becomes problematic when you assert something that includes 3, like "find all bugs in any program..."
The halting problem is not a problem for constrained sets of programs.

Agreed.