Guess you depend on how you measure. As "released product to the public" then probably yes. Cumulative training hours spent actually creating and adjusting the weights during training? Probably no.
Maybe, I do not know how the number of neurons in a child's brain and the connections compare to Claude and its training but I would think it is comparable. Also, while a child might sleep the brain does not just switch off, there is still stuff going on that adds to the child's development.
I am, here, not counting time for separate instances of Claude (so 10 instances running for a year is not 10 years). So, I think the 8yo is still older.
Full disclosure, I am not a neurologist or computer scientist (although I find both interesting). I would consider fair criticism of this fair and would even like to see what those in those fields would have to say.
And Claude is even younger.
Guess you depend on how you measure. As "released product to the public" then probably yes. Cumulative training hours spent actually creating and adjusting the weights during training? Probably no.
Maybe, I do not know how the number of neurons in a child's brain and the connections compare to Claude and its training but I would think it is comparable. Also, while a child might sleep the brain does not just switch off, there is still stuff going on that adds to the child's development.
I am, here, not counting time for separate instances of Claude (so 10 instances running for a year is not 10 years). So, I think the 8yo is still older.
Full disclosure, I am not a neurologist or computer scientist (although I find both interesting). I would consider fair criticism of this fair and would even like to see what those in those fields would have to say.
1 reply →