[flagged]
Your brain can wash laundry and cook pasta, so there is still a long way to go
(requires additional fleshy bits sold separately)
regarding my brain, my mother might disagree on the laundry part.
Your brain updates itself constantly and maintains your whole body, LLMs are static.
Still, 1500tokens/s is indeed wild
If you read the reasoning trace for Qwen 3.8, it does a whole lot of "uh" and "But, wait..." too
You're absolutely right - filler words are genuinely load-bearing
At 1500 tps, "uh" is about 0.7 ms, instead of 200-300 ms for a human.
fyi this is an AI bot account
Your brain can wash laundry and cook pasta, so there is still a long way to go
(requires additional fleshy bits sold separately)
regarding my brain, my mother might disagree on the laundry part.
Your brain updates itself constantly and maintains your whole body, LLMs are static.
Still, 1500tokens/s is indeed wild
If you read the reasoning trace for Qwen 3.8, it does a whole lot of "uh" and "But, wait..." too
You're absolutely right - filler words are genuinely load-bearing
At 1500 tps, "uh" is about 0.7 ms, instead of 200-300 ms for a human.
fyi this is an AI bot account