[Member Intro] Potential Software to Create Faster Method

[Smv1264836]

Hello Everyone,

I was thinking about the potential in using Artificial intelligence to create a new method. Not just a any method. One that has potential to be faster than CFOP and Roux with enough practice. I don’t know if I would pursue this but I defiantly am thinking about it as it can really change speed cubing. I am also making this post so that it can maybe spark someone else’s interest in this. AI has become very powerful.Certain Parameters can be set allowing the AI to solve the Rubik’s cube hundreds of times using the same method every time. This is not just solving the Rubik’s cube in 20 moves every time. This is potentially solving the cube in under 30 moves every time whilst using a single method that the AI can figure out. I am more experienced with web applications but I have worked with Artificial Intelligence. Though I have little Knowledge, I don’t think implementing this would be to hard. Thank you - Smv

 

[Smv1264836]

Hello Everyone,

I was thinking about the potential in using Artificial intelligence to create a new method. Not just a any method. One that has potential to be faster than CFOP and Roux with enough practice. I don’t know if I would pursue this but I defiantly am thinking about it as it can really change speed cubing. I am also making this post so that it can maybe spark someone else’s interest in this. AI has become very powerful.Certain Parameters can be set allowing the AI to solve the Rubik’s cube hundreds of times using the same method every time. This is not just solving the Rubik’s cube in 20 moves every time. This is potentially solving the cube in under 30 moves every time whilst using a single method that the AI can figure out. I am more experienced with web applications but I have worked with Artificial Intelligence. Though I have little Knowledge, I don’t think implementing this would be to hard. Thank you - Smv

P.S. Sorry for some typos:
Not just any*
Too hard**

 

[Nmile7300]

P.S. Sorry for some typos:
Not just any*
Too hard**

FYI you can edit your post

 

[Smv1264836]

FYI you can edit your post

Haha thanks. What do you think about my idea? Thanks.

 

[TheSlykrCubr]

yea you could make a database

 

[Smv1264836]

Yeah. I don’t know much but I definetly could try. I do see a lot of Potential in this.

 

[Nmile7300]

Haha thanks. What do you think about my idea? Thanks.

The thing is, cube methods depend on more than move count to be good (see how CFOP became insanely popular with a way worse move count than other methods). Methods also depend on good ergonomics and fast recognition. I feel like this is a cool idea, but I doubt that the method used by the computer will be the best method for humans. Still really cool though.

 

[Smv1264836]

The thing is, cube methods depend on more than move count to be good (see how CFOP became insanely popular with a way worse move count than other methods). Methods also depend on good ergonomics and fast recognition. I feel like this is a cool idea, but I doubt that the method used by the computer will be the best method for humans. Still really cool though.

Yeah. That is why some parameters need to be set making it hard. Parameters that make it so that the AI formulated a method good for human turning and those type of things. Setting those parameters is where it gets really hard. This also can be used to create new alg sets. I originally thought this would be good for roux because there aren’t many subsets for it and I use roux. I also thought that it can be used for a new method. I still think it is possible but there are many factors that would make it difficult to code. Difficult not Impossible though.