Hacker News new | ask | show | jobs
by lo0dot0 373 days ago
The answers can be recorded and reviewed. The other points are true, or is there a way to make outcomes deterministic, when compared to previous versions while allowing to add more knowledge in newer versions?
1 comments

It's possible to make any model deterministic. Used to be just to save the seed, but I'm not sure it still is now that everything is distributed. Maybe a little more effort.
determinism isn’t really enough, we want “predictable”. Most of these AI wavefunctions are “chaotic” - tiny changes in state can cause wildly divergent outcomes
A part of my question that you didn't go into was, can new knowledge be added in a new version without making the answers with knowledge learned in previous versions non-deterministic?
that’s not really how training works.

changing the input (data) means you get a different output (model).

source data has nothing to do with model determinism.

as an end-user of AI products, your perspective might be that the models are non-deterministic, but really it’s just different models returning different results … because they are different models.

“end-user non-determinism” is only really solved by repeatedly using the same version of a trained model (like a normal software dependency), potentially needing a bunch of work to upgrade the (model) dependency version later on.

This requires an exact lock-down of things like the hardware and driver version, doesn't it? Is that sustainable?
It shouldn't. It didn't used to, at least.
But that won't survive an upgrade, will it?