3 ms·
You just need to devise a system where the vast majority of universes give the right answer for this to work. Start by at least having a universe for each poss
by AnotherGoodName 2y ago
You just need to devise a system where the vast majority of universes give the right answer for this to work.
Start by at least having a universe for each possibility so that every code path is computed. Then add a mechanism to create many more universes when a correct result is hit.
So each wrong result has 1 universe and the correct result alone has 2^300 a universes. Run this. You now end up with the correct result 99.99999% of the time.
I’m not arguing on the above take or not fwiw, it’s just that it easy to see how this could happen in many worlds. Effectively error correction just becomes a mechanism to create more universes for correct answers than incorrect answers and it all works. In fact it’s very reasonable to think of quantum error correction this way (it really is a mechanism to favour correct answers being observed and in many worlds that means it creates more correct answer universes).
- skirmish 2y ago> add a mechanism to create many more universes when a correct result is hit Simple count has not much to do with the outcome observed, IMU each branch in a split is not equally likely. If you finely sub-divide a low-probability branch into 2^300 sub-branches, so what. Instead, you need to merge multiple high probability branches back into a single outcome that is observed "99.99999% of the time".
- AnotherGoodName 2y agoIn many worlds the probability observed can absolutely be thought of as the ratio of worlds with that outcome vs another. Any specific numbers here are just as an example.
- knome 2y agoThis feels like taking quantum bogo sort seriously.