Yet we learning from 18 hours of ultra high resolution video per day (x2), along with 18 hours of reinforcement learning, along with 18 hours of audio and 18 hours of nerve stimulus data.
This is assuming we are learning nothing during sleep, which probably isn't true.
By the time a person is 21 years old, they have been trained on at least 1 petabyte of data.
By two years old, about 125TB of data. It makes LLMs look quite good in comparison.
This is assuming we are learning nothing during sleep, which probably isn't true.
By the time a person is 21 years old, they have been trained on at least 1 petabyte of data.
By two years old, about 125TB of data. It makes LLMs look quite good in comparison.