Except even moreso with language models. Here it is not just a similar technology, but the exact same vehicle.

Translation: The way scaling laws and gradient descent work for language models means that there is NO functional difference between a machine that does your homework and is designed to endlessly praise you, and a model specifically designed to identify viruses or medications.

deleted by creator

That was written when less than 10% of the population was literate because of how expensive and difficult it was to learn. There was no publicly funded literary education, and public opinion on literacy was at best contentious.

It’s not a similar conclusion. You are literally parroting the same talking points. They are just as fundamentally misguided and divorced from reality now as they were 2000 years ago.

Removed by mod

…said many things to Theuth in praise or blame of the various arts, which it would take too long to repeat; but when they came to the letters, “This invention, O king,” said Theuth, “will make the Egyptians wiser and will improve their memories; for it is an elixir of memory and wisdom that I have discovered.”

But Thamus replied, “Most ingenious Theuth, one man has the ability to beget arts, but the ability to judge of their usefulness or harmfulness to their users belongs to another; and now you, who are the father of letters, have been led by your affection to ascribe to them a power the opposite of that which they really possess.

“For this invention will produce forgetfulness in the minds of those who learn to use it, because they will not practice their memory. Their trust in writing, produced by external characters which are no part of themselves, will discourage the use of their own memory within them. You have invented an elixir not of memory, but of reminding; and you offer your pupils the appearance of wisdom, not true wisdom, for they will read many things without instruction and will therefore seem to know many things, when they are for the most part ignorant and hard to get along with, since they are not wise, but only appear wise.” (Phaedrus 274c-275b)

Why don’t you try getting a new argument, yours is a bit dated.

ollama

Shittiest local inference engine. Use llama.cpp instead.

qwen 3:4b-instruct

Old version 3 model from Alibaba. 4 billion parameters total. It will struggle with anything other than natural language generation.

on 6vCPU and no GPU

Running on a 6 vcore VM with no GPU pass through.

TTFT

Time.To.First.Token. How long it takes the model to start responding after you hit enter. Dependent on not only processing speed but also context length.

prompt is uncached

Cold start = no cache = slow. But faster when mid conversation and the key-value cache is hot.

100tps

100 tokens per second. 1 token is roughly 1/4 word.

Not to mention that it’s training and not inference where most of the compute is going anyway.

Friends don’t let friends use Ollama.

Which would have you right at the top of the watchlist. Gotta keep an eye on this guy; you never know what he’s gonna do.