deleted by creator

how? when the linux kernel looks at you funny if you even mention kernel interface stability within a 100km radius

except for the no reclocking thing, which cripples them

“conditional on the context and the model’s learned parameters.” you seem to be under the wrong impression that “random dice roll” == “random dice roll from a uniform distribution”. I didn’t say that. If it outputs a probability distribution, which it does, then you sample it randomly according to that distribution, not a uniform one.

As for your last paragraph: I wasn’t, I didn’t do that, and if that’s all the system can do then people should stop claiming it is even remotely intelligent. Whatever the excuses, the systems aren’t (and won’t be getting) there. If you’re trying to get me to empathize with a couple of matrices, then you’re not going to succeed.

I don’t care if you get offended because someone else doesn’t like your line of work. I think what you do is actively harmful to humanity. I also dislike weapons manufacturers, how they feel about it is irrelevant. You’re no different

even when said “one program” is actually 69 (nice) different binaries

i installed mandrake in 2004. It came with a nice graphical installer.

extensions are just scripts that monke patch == there is no extension mechanism

they say they do is not the same as they do. Find me someone who isn’t them who says the same thing

no, the opposite. The problem with ai pull requests is that in most cases whoever submits them does not understand the code and expects someone else to review it for them (that’s if they are even aware of the concept of code reviews in the first place).

“why would that be anyone but the original author?”

That is what i was replying to, and I replied to the intended comment

in the case of ai generated code, that is almost always the case. People say “but I review all my pet neural network’s code!” but they don’t. If they did, the job would actuallydtake longer. Reading and understanding code takes longer than writing it.

it takes two instructions to materialize a constant in risc-v. X64 has LEA.

Risc-v is better!

not really. If the system outputs a probability distribution, then by definition, you’re picking somewhat randomly. So not really a simplification

if every single token is, at the end, chosen by random dice roll (and they are) then this is exactly what you’d expect.

but they rewrowe it in rust for safety!

everyone’s adopting it because they’re forced to. And shut up with your “but you can use X”, some distros literally plan to drop support for it entirely.

that’s the thing… wayland has repeatedly said they will not reach feature parity. So from the word “until” onwards cad be deleted, back to the older comment

“these new cars have a teeny tiny fuel tank with a tiny range! They used to have a bigger tank!”

“Drive an old car”

In this case the new car is objectively inferior, and I can’t buy a new old car anymore.

When something complains about very real problems due to missing functionality, the proper answer isn’t “fuck you, use the old stuff, or stop yearning for the functionality that te intentionally crippled”

I am not referring to it being a drop-in replacement. I’m referring to the fact that there are multiple supposedly-interoperable-but-not-really non-drop-in replacements is the problem. And it does affect the end user if devs find it difficult to adopt (as many do).

Wayland is designed for ease of development for wayland designers. “We’re just a protocol, the coding is left to anyone else” is the easiest way to write code. Because they decided not to write any at all.

well that’s the problem. “I don’t use it therefore it must be a bad idea”