DeepMind AlphaFold
-
EdLee
- Honinbo
- Posts: 8859
- Joined: Sat Apr 24, 2010 6:49 pm
- GD Posts: 312
- Location: Santa Barbara, CA
- Has thanked: 349 times
- Been thanked: 2070 times
-
SoDesuNe
- Gosei
- Posts: 1810
- Joined: Wed Apr 21, 2010 1:57 am
- Rank: KGS 1-dan
- GD Posts: 0
- Has thanked: 490 times
- Been thanked: 365 times
Re: DeepMind AlphaFold
So by taking part twice in a competition over four years they produced an algorithm "scoring" on average over 90% (basically solving the problem) whereas in the last eight installments (over the last 16 years!) the competition plateaued at around 60% each time... And even more mind-bogglingly in the 22 years before DeepMind took part in the competition other participants only managed to better the score by around 15%... in 22 years!
Simply astonishing. Kurzweil's law in practice.
And a limited AlphaFold-1-version (CASP13-version) was made available on GitHub for everyone.
Simply astonishing. Kurzweil's law in practice.
And a limited AlphaFold-1-version (CASP13-version) was made available on GitHub for everyone.