3 ms·
> The problem here is that it's missing the "glue" to more real world applications. This is where more humdrum software engineering comes in. This is a bold st
by dasil003 4y ago
> The problem here is that it's missing the "glue" to more real world applications. This is where more humdrum software engineering comes in.
This is a bold statement. The world does not function based on "well-defined rewards". The concept of "common sense", which some consider table stakes for a human operating competently in our world, is mostly made up of things which are neither well-defined, nor allow a tremendous amount of training data. Current ML approaches require both.
- VHRanger 4y agoCorrect, but there are likely applications of AI in strategic reasoning that are used in the real world (outside the obvious finance bots) by engineering a "translation layer" from real world constraints to AI-compatible value functions and back. But in general, yes, this is why since 2013 we don't see AI making anywhere as massive strides everywhere as they do in boxed-in applications like games.
- roenxi 4y ago> The world does not function based on "well-defined rewards". Humans can be neatly reduced to a "procreate, and make sure your progeny procreate" value function. A lot of apparent counterexamples to that turn out to make sense when 2nd order effects on progeny are considered.
- FeepingCreature 4y agoHumans are adaptation-executors, not fitness-maximizers. Evolution has a procreation value function; humans have procreation-favoring traits.