qengho writes "Vstone's programmable and automonous humanoid robot Robovie-M can play soccer, throw a ball both over- and underhand, and even demonstrate shame (over a missed shot, presumably). Lots of QuickTime movies of the bot in action."
There's a big difference between demonstrating shame and feeling shame. It's a pretty paltry feature compared to the ability to automate its motions so well. I wonder how well it'll do in the RoboCup?
Carnegie Mellon AIBOs pwn this
by
ElGnomo
·
· Score: 5, Informative
Im actually a student at Carnegie mellon University where we have our own teams of robotic soccer players, ranging from cute little Sony Aibos dogs that pwn in international Aibo dog soccer competitions ( not kidding! )to an up and comming segway team!
check it out here
Re:Carnegie Mellon AIBOs pwn this
by
metlin
·
· Score: 4, Informative
I'm impressed by this
by
SimianOverlord
·
· Score: 4, Insightful
This little guy shows that as far as the goal of a fully working robotic (human-avatar) intelligence, mechanistically we're there. We can create an avatar which is robust enough to do the full human movements. Now we just need to improve the AI side of things. Let's face it, AI is severly lacking at the moment. I wouldn't say I'm an expert by any means, but I do try to keep up with the relevant research in the field.
Most experts agree the real test of artificial intelligence is, of course, found in game theory, with chess programs in particular gaining a lot of attention and publicity because of their year on year increase in performance thanks to better and better AI. But these certainly aren't quite up to scratch - a good example is a game I played just 10 minutes ago against the computer at so-called "Intermediate II" level.
I find most computer chess programs good at opening moves, because most have built in opening move databases (which I regard as cheating). I played the two knights defense as always, leading with 1....d4. The computer answered with its own two knights defense which is particularly annoying and insipid. That's one mark down for AI, mimicking human behaviour.
The trouble is chess AI is reactive rather than proactive and so whereas I was working to take down the right flank of the computers defence using a Queen bishop rook combo, it was just skirmishing around the centre to take a few pawns off me. Very poor AI.
Frankly, it's stategic thinking that the computer can't match us at, which was clear by move 27, where my move 27...KnC3-E4 forked a queen and rook. That sort of manouvering is quite obvious to the human intellect, and the advantage in points meant I went on to win the game quite comfortably again, with a rook v bishop endgame and quick checkmate.
Really all this just goes to illustrate that AI has a long way to go...I wouldn't say I was much more than a very good chess player.
-- Meine Schwester ist sehr, sehr reizvoll - Nietzsche
Re:What about this?
by
levram2
·
· Score: 5, Informative
That "robot" is a cgi stealth marketing ad for the mini cooper. See this previous slashdot discussion.
1869 Rutgers and Princeton played a college soccer football game, the first ever, November 6. The game used modified London Football Association rules. During the next seven years, rugby gained favor with the major eastern schools over soccer, and modern football began to develop from rugby.
1876 At the Massasoit convention, the first rules for American football were written. Walter Camp, who would become known as the father of American football, first became involved with the game.
So, American Football (Gridiron) is based on Rugby, which is a sibling of Football (the game where you kick the ball all the time).
Slashdot Mirror
Hot robot soccer moms?
Comment removed based on user account deletion
"even demonstrate shame"
There's a big difference between demonstrating shame and feeling shame. It's a pretty paltry feature compared to the ability to automate its motions so well. I wonder how well it'll do in the RoboCup?
Im actually a student at Carnegie mellon University where we have our own teams of robotic soccer players, ranging from cute little Sony Aibos dogs that pwn in international Aibo dog soccer competitions ( not kidding! )to an up and comming segway team!
check it out here
Yes, but does it rip off its shirt after a cup-winning goal shot? Everyone knows that the crowd likes it when the players hamm it up.
(ducks, runs for cover)
Please help metamoderate.
This little guy shows that as far as the goal of a fully working robotic (human-avatar) intelligence, mechanistically we're there. We can create an avatar which is robust enough to do the full human movements. Now we just need to improve the AI side of things. Let's face it, AI is severly lacking at the moment. I wouldn't say I'm an expert by any means, but I do try to keep up with the relevant research in the field.
Most experts agree the real test of artificial intelligence is, of course, found in game theory, with chess programs in particular gaining a lot of attention and publicity because of their year on year increase in performance thanks to better and better AI. But these certainly aren't quite up to scratch - a good example is a game I played just 10 minutes ago against the computer at so-called "Intermediate II" level.
I find most computer chess programs good at opening moves, because most have built in opening move databases (which I regard as cheating). I played the two knights defense as always, leading with 1....d4. The computer answered with its own two knights defense which is particularly annoying and insipid. That's one mark down for AI, mimicking human behaviour.
The trouble is chess AI is reactive rather than proactive and so whereas I was working to take down the right flank of the computers defence using a Queen bishop rook combo, it was just skirmishing around the centre to take a few pawns off me. Very poor AI.
Frankly, it's stategic thinking that the computer can't match us at, which was clear by move 27, where my move 27...KnC3-E4 forked a queen and rook. That sort of manouvering is quite obvious to the human intellect, and the advantage in points meant I went on to win the game quite comfortably again, with a rook v bishop endgame and quick checkmate.
Really all this just goes to illustrate that AI has a long way to go...I wouldn't say I was much more than a very good chess player.
Meine Schwester ist sehr, sehr reizvoll - Nietzsche
That "robot" is a cgi stealth marketing ad for the mini cooper. See this previous slashdot discussion.
Is shame *really* what we want in a robot?
509 - Server Is Expressing Shame for not handling
slashdot load.
Table-ized A.I.
Initialise:
-Take money
-Find ball
-Take money
-Move towards goal
-Take money
Game end:
-Return to changing rooms
-Overclock servos
-Refuse overclocking test
-Disqualified
-Take vacation
-Take money
-Have baby: name=[city|band|corporate product]
General:
-Take money
-Take knight-hood
Error Handling:
-Avoid tabloid photographers
-Avoid mid-season injuries
Additional: if (popularity > 10)
-Reality TV spin-off
-Reality TV spin-off
-Cameo
-Release single
This comment does not represent the views or opinions of the user.
which says:
1869
Rutgers and Princeton played a college soccer football game, the first ever, November 6. The game used modified London Football Association rules. During the next seven years, rugby gained favor with the major eastern schools over soccer, and modern football began to develop from rugby.
1876
At the Massasoit convention, the first rules for American football were written. Walter Camp, who would become known as the father of American football, first became involved with the game.
So, American Football (Gridiron) is based on Rugby, which is a sibling of Football (the game where you kick the ball all the time).