Do you think there will be a 1LTC subset in the next century?

[Cuber142857]

1LTC would stand for 1-Look the cube. It would have 43 252 003 274 489 856 000 (including solved state) algorithms. Imagine if cubers in the future (Tymon or Max) memorise that subset…

1LTC would include every single 3x3 algorithms there are and a bunch of algorithms for random cases.

Who would ever memorise 43.25 quintillion algorithms?

Imagine anyone memorising those algs and getting sub 2 averages…

 

[xyzzy]

Simply not possible.

A typical human life lasts around 2 to 3 billion seconds. You'd have to cram multiple billions of algs per second, from the day you're born to the day you die, in order to finish learning this. Now, you can reduce the number of algs needed with symmetry, but you're still looking at hundreds of millions of algs per second regardless.

This would be possible if we look beyond purely biological human beings, i.e. cyborgs with computerised brains. But we already have computer algorithms (that work now, today!) that can compute 21-move solutions in a millisecond so there's no real advantage to treating them as individual algs to be rote-memorised.

 

[abunickabhi]

1LTC would stand for 1-Look the cube. It would have 43 252 003 274 489 856 000 (including solved state) algorithms. Imagine if cubers in the future (Tymon or Max) memorise that subset…

1LTC would include every single 3x3 algorithms there are and a bunch of algorithms for random cases.

Who would ever memorise 43.25 quintillion algorithms?

Imagine anyone memorising those algs and getting sub 2 averages…

Sub-2 averages not possible.

I am working on a 1 look, 7 algs method to solve the entire cube, method called 5-style.

A large part of the community calls it infeasible already, I don't think anything more grander than that will be even considered.

 

[DGCubes]

As other people have said, the way to do this is definitely not by memorizing 43 quintillion algorithms - I'm generally a very optimistic guy when it comes to cubing improvement but even I'll admit that's impossible. However, I do believe that sub-2 human averages are possible on 3x3, given many many years of method optimization. Currently, domino reduction is used by FMC experts to get move averages in the low-20s - but it takes an hour of trial and error to find these solutions, so it is not feasible for speedsolving in its current state. That said, I believe that with enough time and human ingenuity, we will devise a method to find near-optimal solutions (say 25-30 moves) in the inspection period. Once that can be done, it just comes down to turning speed. Ruihang Xu has already hit 15 TPS in solves - so if we can match that TPS for a sub-30 move solve, boom, sub-2.

None of this progress will be even remotely easy, but the cubing community has proven people wrong time and time again about the limits of what's possible. I believe that if cubing survives long enough and top solvers continue to push each other to their limits, we will see sub-2 averages someday.

 

[Cuber142857]

As other people have said, the way to do this is definitely not by memorizing 43 quintillion algorithms - I'm generally a very optimistic guy when it comes to cubing improvement but even I'll admit that's impossible. However, I do believe that sub-2 human averages are possible on 3x3, given many many years of method optimization. Currently, domino reduction is used by FMC experts to get move averages in the low-20s - but it takes an hour of trial and error to find these solutions, so it is not feasible for speedsolving in its current state. That said, I believe that with enough time and human ingenuity, we will devise a method to find near-optimal solutions (say 25-30 moves) in the inspection period. Once that can be done, it just comes down to turning speed. Ruihang Xu has already hit 15 TPS in solves - so if we can match that TPS for a sub-30 move solve, boom, sub-2.

None of this progress will be even remotely easy, but the cubing community has proven people wrong time and time again about the limits of what's possible. I believe that if cubing survives long enough and top solvers continue to push each other to their limits, we will see sub-2 averages someday.

Yes. It’s impossible but just imagine…

 

[ProStar]

In a centruy theres a good chance human implants will become omnipresent, supercomputers will be commonplace, and radical life extension technology may also be developed. With a theoretically immortal person with brain and finger implants, it is feasable that our brains could download all 43 quintillion states and achieve incredible turning speeds. Even without finger/hand implants some very fast solves could be made.

As other people have said, the way to do this is definitely not by memorizing 43 quintillion algorithms - I'm generally a very optimistic guy when it comes to cubing improvement but even I'll admit that's impossible. However, I do believe that sub-2 human averages are possible on 3x3, given many many years of method optimization. Currently, domino reduction is used by FMC experts to get move averages in the low-20s - but it takes an hour of trial and error to find these solutions, so it is not feasible for speedsolving in its current state. That said, I believe that with enough time and human ingenuity, we will devise a method to find near-optimal solutions (say 25-30 moves) in the inspection period. Once that can be done, it just comes down to turning speed. Ruihang Xu has already hit 15 TPS in solves - so if we can match that TPS for a sub-30 move solve, boom, sub-2.

None of this progress will be even remotely easy, but the cubing community has proven people wrong time and time again about the limits of what's possible. I believe that if cubing survives long enough and top solvers continue to push each other to their limits, we will see sub-2 averages someday.

Sorry, but wouldn't 15TPS be impossible on (near) move-optimal solutions? Since they are almost never ergonomic, they may even be slower than a longer (say 40 moves) solve that has better ergonomics and allows for higher TPS

 

[NmeCuber]

In a centruy theres a good chance human implants will become omnipresent, supercomputers will be commonplace, and radical life extension technology may also be developed. With a theoretically immortal person with brain and finger implants, it is feasable that our brains could download all 43 quintillion states and achieve incredible turning speeds. Even without finger/hand implants some very fast solves could be made.

But would finger implants be allowed though?

Sorry, but wouldn't 15TPS be impossible on (near) move-optimal solutions? Since they are almost never ergonomic, they may even be slower than a longer (say 40 moves) solve that has better ergonomics and allows for higher TPS

Sometimes efficiency is better than TPS

 

[ProStar]

But would finger implants be allowed though?

If we're at the point where every single human on earth has them (not unlikely if the timescale is a hundred years), then it's hard to disallow. Although at that point cubing probably won't be a thing anyway

Sometimes efficiency is better than TPS

True, but 30 moves at 10 TPS (3 second solve) is worse than 40 moves at 15 TPS (2.66 second solve). High TPS will be very difficult with near-optimal solutions becuase of how unergonomic they usually are

 

[Deleted member 36445]

Forget about robotic fingers

imagine a new type of cube like a virtual one that allows to turn almost instantaneously, so the only barrier would be our minds

like virtual reality + brain signals controller

EDIT: Also, no need to coin a new term like 1LTC, there is already F3L

 

[Isaiah Scott]

sub 2 is possible and if you live 43,252,003,274,489,856,000 lives then memorizing all those algs is possible

 

[NmeCuber]

True, but 30 moves at 10 TPS (3 second solve) is worse than 40 moves at 15 TPS (2.66 second solve). High TPS will be very difficult with near-optimal solutions becuase of how unergonomic they usually are

But would a 40 move solution have good enough finger tricks for 15 tps? Ruihang is like the only person with 25 tps rn

 

[ProStar]

But would a 40 move solution have good enough finger tricks for 15 tps? Ruihang is like the only person with 25 tps rn

A 40 move solution definitely would have far superior ergonomics to a 25-30 move solution. And if a human is already getting 15 TPS, then in 100 years it can be far more than that I imagine.

An important thing to note in this conversation is just how long 100 years is. 100 years ago from today we were in the middle of the roaring 20s, and since technology generally improves faster and faster as time goes on (until it reaches some sort of soft cap), the next 100 years will likely contain even more advancements than the last 100