admiralrohan
Joined 2 years ago
Joined 2 years ago
In future LLMs will make buying decisions and this strategy can become viable again.
From their docs "After using 10M input (cache miss) tokens of MiMo-V2.5-Pro, it is equivalent to consuming 3000M Credits, and you can still enjoy 1100M Credits of MiMo-V2.5". So it's around 12M input credit vs Earlier 60M tokens.
The question is how they are managing to do so? They are supposed to struggle due to chip sanctions.
Secondly, why now? The US companies were supposed to subsidize too but now they are unable to keep up. Everyone going to usage based pricing, so it's unsustainable for them. They are well funded too.
If there are genuine hardware breakthrough reducing compute needs then that is good for the whole world I believe.
One good thing is I have gained so much confidence on my work. I can't surely do worse work than those smart Google engineers.
I am working on an original theory to explain human behavior that would unify fragmented theories across psychology, mental health, productivity, economics, and beyond.
This would allow teaching critical thinking to the masses, ultimately leveraging it to position India as the next superpower.
Blog: https://www.moderncynicism.com/