That's precisely what he is saying, there is diminishing returns (or optimization left on the table).
replyI read it as it is impressive because smaller models 2.5T are squeezing similar returns as 10T models despite being 1/4th size not that there beyond 2T today the number or parameters do not have much meaning
replyor the latest qwen3.8 27B doing so well at ~1/100 the size of K3
replyWhat about general knowledge you can get out of it before hallucinations start?
replyStoring general knowledge in VRAM has always been a dumb idea in the first place.
replyQwen 3.8 27B beats Opus, Fable and GPT 5.6 by a comfortable margin on the AA-Omniscience Hallucination Rate benchmark.
replyIt did OK on schlongbench v1.0 (test of a specific niche word that doesn't make it into smaller LLMs) but it sure does love to count words
replyI do not rely on any LLM of any size for general knowledge baked into the weights, they all hallucinate and that is the wrong way to hold them imo
replyI think there is some merit in that smaller models cannot memorize so much of the training data, i.e. that they are less likely to do copyright infringement, and by analogy not having memorized SDK / API surfaces that have since changed from the training data
> I do not rely on any LLM of any size for general knowledge baked into the weights
replyYou have to rely on it to a certain level for agentic/coding work, presuming that's the general subject we're talking about here... For instance I recently encountered a project where it would have been a lot worse if the LLM didn't already know "what is" xterm.js and a bunch of its associated npm-related/node related software. If it was still smart but had to google and find results for everything it would have been a lot more time consuming and risked sending it down a wrong path.