“We started with roughly, say, eight. We've generated $17 billion of profits on that original $8 billion. And we've paid out roughly $5 billion to the university”
“It's more like a massive 594 billion page library, but you only need to pull out 17 billion specific pages to get your answer.”
AI Brief · EP 683 : Thinking Machines Lab Inkling 2, Xiaomi Humanoids, MIT Memory Fix & Waymo Robotaxi Update
Audio streams from the publisher’s own feed. Verbatim does not host or copy it.
What was described
use a mixture-of-experts architecture that activates only 17 billion parameters out of 594 billion for each prompt
594 billion total parameters, 17 billion activated per prompt
Said nearby on the same subject
“A 26 billion parameter models can already orchestrate full replication chains ... able to create small $9 billion parameters models that established interface on eight of the nine hardware types”
“Hardware brought in a much bigger 307 billion in revenue but with a gross margin of only 37 it produced about 113 billion in gross profits”
“We mobilised $112 billion of private capital, $112 billion of private capital mobilisation. A few years ago, that figure stood at around $30 billion”
Moments like this, every Monday
The strongest new moments on your topics, in the speakers' own words. Your first brief is free, then USD {{ TIERS.starter.price }} a month for three topics.
No card to start. No password. Cancel any time.
Track a competitor?
Pro watches up to 25 names or phrases and emails you the morning after they are said.
See Pro, $29 a monthVerbatim quotes a short extract and links to the episode; the episode belongs to its publisher. Remove a show.