@apples_jimmy feels like we are deeper into it each advance compounds the next and the infrastructure adapts in real time
@apples_jimmy Starting to feel faster Jimmy I think with the Kimi release nonzero chance of me clenching my bum soon
@apples_jimmy at this moment - yes...considering whats to come
Positive users express optimism about compounding AI advances and faster progress from releases like Kimi in light of the mathematician's lower bound finding, while negative users reply with crude hostility or content disapproval.
Based on 6 visible X reactions from 12 accounts; directional sample.
Ask a question below.
Published answers will appear here.
@apples_jimmy feels like we are deeper into it each advance compounds the next and the infrastructure adapts in real time
@apples_jimmy Starting to feel faster Jimmy I think with the Kimi release nonzero chance of me clenching my bum soon
@apples_jimmy at this moment - yes...considering whats to come
@IALatino @apples_jimmy tag this nsfw ffs
@2cantsam Feel it inside of you 2can
So d and 1/eps^2 are trivial lb and I think both attainable via Alg? So this new lb is only for eps =d^-c and c=1/2 or some specific value of c it's an improved lb I think. I couldnt follow from the post the Q(eps, d) had an eps dependence and later became d^2 but I guess it must be some specific eps that's <<1 but not going to 0 too fast in d
Are we still in the gentle singularity ?
@IALatino @apples_jimmy tag this nsfw ffs
@2cantsam Feel it inside of you 2can
So d and 1/eps^2 are trivial lb and I think both attainable via Alg? So this new lb is only for eps =d^-c and c=1/2 or some specific value of c it's an improved lb I think. I couldnt follow from the post the Q(eps, d) had an eps dependence and later became d^2 but I guess it must be some specific eps that's <<1 but not going to 0 too fast in d
Are we still in the gentle singularity ?
Are we still in the gentle singularity ?
This is actually crazy, many of us thought about this lower bound for many years ... https://www.reddit.com/r/math/s/lrKVCd6WcZ
So d and 1/eps^2 are trivial lb and I think both attainable via Alg? So this new lb is only for eps =d^-c and c=1/2 or some specific value of c it's an improved lb I think. I couldnt follow from the post the Q(eps, d) had an eps dependence and later became d^2 but I guess it must be some specific eps that's <<1 but not going to 0 too fast in d
Positive users express optimism about compounding AI advances and faster progress from releases like Kimi in light of the mathematician's lower bound finding, while negative users reply with crude hostility or content disapproval.
Based on 6 visible X reactions from 12 accounts; directional sample.
Ask a question below.
Published answers will appear here.
