With the right rules changes could humans beat computers?

For some reason it looks more appealing than current chess, to me...

Houdini 3 could never beat Nakamura at bullet chess PROVIDED the computer was limited to the size of the human brain and it physically had to pick up and move its pieces and punch the clock. If computers had to move their own pieces in bullet chess i.e. a robot physically moved the pieces, it would take at least 50 years before computers were World Champion.

The reason I predict this is that computers are tied to sequential logic i.e. first they have to calculate the move; only then can they decide to move a piece and punch the clock. A human might well make a split second decision to change a move. If a computer is caught in mid-calculation, it might well start playing second and third best moves. Even Houdini 3 has a limited depth of search when it has to calculate in 1/10th -1/20th of a second---and then move.

In science we perform experiments. I am offering a very simple cheap experimental design. Change the rules to make them more human friendly and a set of "reasonable" rules changes will permit humans to become World Champion again. It is just a question of tweaking the rules, convinving the public that they are fair, like requiring programmers to limit the computer to the size of the human brain and requiring the computer to physically move its pieces and punch the clock.

What you are saying is no different than saying why should we be upset that Lance Armstrong use PAD's? Computers are given an unfair advantage. They have unlimited access to opening and endgame theory and play. Shouldn't humans be given the same benefit?

Computers can never make a touch move violation. If we eliminate time delay, then computer programmers would be forced to decide when to shift to "blitz" mode i.e. if they know they are faced with the one second rule, they might set aside five minutes at the end of the game for blitz being able to make about 250 moves in five minutes.

For there to be a "true" contest between humans and computers then we should require computers to have their CPU size limited to the size of the CPU of a human being studying a chess position. If programmers want to contest the one second rule, then let them build a computer no bigger than a human brain that must control the movement of its own pieces and punch the clock. You would be surprised how much space will be required to get a computer to play bullet chess well if it has to move its own pieces and punch the clock.

What you are proposing is that we limit human effectiveness by forcing humans to obey rules that are patently unfair and benefit computers.

I don't really see the point in this.

It is obviously very easy to give the human an edge by requiring the computer to perform physical tasks in addition to just calculating a move.

I would just accept that a computer is better at the chess game than a human and don't try to find ridiculous backdoors to turn that around with some made up rules.

Chess could become the most popular game in chess, if, with simple rules changes, humans could win again. You dismiss the inherent inefficiency of the human mind that must share different functions other than calculating chess moves. I am simply saying that if we are to truly have a fair contest that #1 Both sides should have the same access to all opening and endgame theory, #2 If computer programmers object to having humans play consultation games, then let them use other computers in series to study the position, #3 If we are going to remove any touch move violation of computers then they have to give up something comparable like eliminating time delay and implementing the one second rule.

You are missing the main point of this thread---if, for any reason, humans could defeat computers at chess by implementing rules that "level the playing field", this has phenomenal monetary implications for chess professionals from National Masters on up. It would be a classic case of John Henry defeating the pile driver, humanity conquering the machine.

For fifty years we have focused on making computers stronger and stronger and then when they beat humans feel sympathy for our puny efforts. We don't know what humans in consultation teams could do against computers. One player in a consultation game could crunch the move while a second player could scour the literature to find relevant theory and play. Together they might find a winning strategy.

We will never know until hundreds of strong chess players against a variety of strong chess programs engage in consultation matches at differing time limits with the rules set forth here, or modifications of them. If they all fail then this would support the prevailing belief that it was a futile effort. If they are successful it will bring vast wealth and press coverage to chess players.

Yes, there is a way. Change chess into Shogi, where the best computers are quite weak

When Deep Blue defeated Kasparov it made world-wide news for several cycles; when a consultation team beats Houdini 3 in a match featuring these new rules, it will be even more newsworthy than when Deep Blue defeating Kasparov.

We are constantly being told that the rise of computers is inevitable; to assure this both at Jeopardy and chess we have implemented rules that are computer friendly. If humans beat computers again at chess using rules that the average person thinks are "reasonable", this will have a tremendously uplifting impact on every human who has been replaced by a machine.

I haven't even addressed a very real advantage of computers that one post member has suggested---we should allow humans on an analysis board shuffle pieces instead of being forced to calculate moves in their minds. Chess programs have a "chessboard in their mind" and shuffle pieces on this virtual chess board. Shouldn't humans be given the same ability?

The amount of press coverage and money that will flow to chess once humans defeat computers at chess cannot be overstated.

There will be tremendous interest in ANY intellectual endeavor where a computer has demonstrated intellectual supremacy only to be beaten by a human at a later date whether it is chess, checkers, or any other intellectual activity. This would extend way beyond the boundaries of chess. You underestimate how much many people look at the advances in artificial intelligence with trepidation. If we "level the playing field" and humans win, there will be phenomenal press coverage, sponsors will line up to fund human/computer contests and a thousands of mid-level professionals will make a good living either as players or teachers.

Leveling the playing field is not crippling the computer. Allowing computers unlimited access to opening and endgame theory, allowing it to "shuffle wood" on its virtual chess board, forcing humans to play games to completion knowing that machines don't tire, taking away any touch move violations by the computer is crippling humans.

Hwy cant just the humans stop making so many stupid moves and make better moves instead? Then they could beat Houdini 3: If they learn to recognize the GOOD moves and avoid all the BAD ones. What do you think about that?

Remember also that humans have created the mackines and that they are here for our wellbeing and in our ownership and control. The fact that the best programs have become so strong is really a human success story if there ever was one.

Good question, but if the rules change too much then chess wouldn't be chess. 

Let's make computers better at chess than humans. Now lets make them worse, because clearly they are too good. We, God of the computers, rather limit their abilities, so what was once ours long ago, can be ours once again. 

The computer isn't crippling humans, but giving the image of it. What was still is. The objective is not to give us (humans) the appearence of greatness, but to be great itself. What this all shows is that there is more to life than logic; that there is more to life than solving problems. Let each do what they were made to do.

"If they learn to recognize the GOOD moves and avoid all the BAD ones."

People spend decades trying to do just that.  Even world champions make bad moves sometimes. 

I was under the impression that the strongest GMs in the world can still beat the strongest engines in the world.

At the present time computers like Houdini 3 Pro and Rybka 4 are rated vastly higher than top humans.

My "exotic" rules are totally unnecessary for humans to be World Champion at chess. All that is needed is to return to the "old rules" i.e 120'40 followed by adjournment. This automatically requires the computer to move its piece and punch the clock i.e. no electronic moves because the computer must be subject to the touch move violation.

The only new rule that is necessary is for humans to be World Champoin is to be able to play consultation games, but only human + human versus computer + computer, but not human/computer versus human/computer. This will eliminate most cheap tactical shots by the computer.

It would be fun to see a consutation match between some top gm`s and one of the best programs. Filming the whole session and seeing the GM`s discuss and analyse together, adding their minds together and discussing to find ever better moves. Add an interviewer to ask some fun and stupid questions and it will be even better. Of course the most fun part will be when they realize that, inspite of all their consulting and discussions and democratic decisions, they will still not stand a chance against the program.

They will also need time odds, in order to have time to talk between them and vote.

Kasparov and Short played a simul; I'd like to see Kasparov, the first World Champion to lose to a computer and GM Judit Polgar, the highest rated woman in the world play the simul. This would have fantastic media appeal.

Together, with existing rules, they would be about 3000 ELO, still not 3300. To get them to 3300, they need to be able to have access to opening and endgame theory as well as adjournment at 40 moves. That gets them up  to 3300 i.e. fully the equal of Houdini 3.

Good ideas all! The key for them to stand a chance and avoid tactical shots is  they need to be able to discuss the position. With 120'40, they would have enough time to chat with the announcer and answer questions. One player can keep track of the position when not much is going on while the other chats with the announcer; since they are used to Game in 60 taking time to talk shouldn't be a problem.