4 ms·
I agree with some of the below comments that there is more hype than reason. For me the observation is from a different starting point: progress in mathematics
by mikorym 9y ago
I agree with some of the below comments that there is more hype than reason. For me the observation is from a different starting point: progress in mathematics is not increasing exponentially, and personally I feel that revisiting parts of mathematics is more enlightening than the current trends in AI.
NN's for example are mostly just an application of linear algebra where design determines the nodes and training the weights; and then, the "decision making" is done by a notion of product. Very useful, but not at all superintellegence.
- foodie_ 9y agoMy feelings are similar. Current AI seems to be the infinite monkey theorem put into practice. I'm not saying it doesn't work, it obviously is, but I wonder what the limitations of it are. When do we start running out of monkeys too the point that only nation states, and companies with the resources of one, are able to achieve the next level?
- randcraw 9y agoViewing math as a delimiter of intelligence only works if you can also factor the requisite skills of intelligence and know that math is sufficient to achieve it, which we can't. The grand AI question has always been, "How much cognitive power + information is needed to implement strong AI/AGI?" The power of linear algebra can't answer that since it measures only one of the two criteria, and only some aspects of cognition. The information threshold to achieve HAL 9000 remains a mystery, not to mention the necessary aspects of the facility to acquire it, namely learning.