3 ms·
If a "given capacity" is the ability to read books and answer questions about the content, then LLMs are more power efficient (4kW for 20 seconds beats 20W for
by p1esk 3y ago
If a "given capacity" is the ability to read books and answer questions about the content, then LLMs are more power efficient (4kW for 20 seconds beats 20W for 3 hours). That's on standard power hungry GPUs (8xA100 server consumes about 4kW). If we switch to analog deep learning accelerators, we gain even better power efficiency. There's simply no chance for brains to compete with electronic devices - as long as we match brain's "capacity".
- uoaei 3y agoOk so it's clear you are not familiar with the vocabulary of this domain. "Capacity" means something specific (if extremely hard to measure precisely) about the informational throughput of a system, not just "the ability to do high-level task X as judged by a naive human observer". I urge you to study more seriously about the things you seem so eager to speculate about before publishing underinformed opinions in public spaces.
- p1esk 3y agoWere you going to provide the definition? What is this “capacity” you are talking about?
- uoaei 3y agoYou have my permission to use your favorite search tool to answer that question.
- p1esk 3y agoTried googling for "ML model capacity" - only found informal handwaving. The closest to a formal definition is VC dimension: https://en.wikipedia.org/wiki/Vapnik%E2%80%93Chervonenkis_dimension https://en.wikipedia.org/wiki/Vapnik%E2%80%93Chervonenkis_di... Is that what you mean when you say "capacity"? Does not seem very relevant in the context of our discussion. If it's not what you have in mind, I'd appreciate a link to the Wikipedia article on "ML model capacity" or whatever specific term experts use to represent the concept.