Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

A chess robot in Moscow has broken the finger of its human opponent

newsbeezer.com

Post reply on HN
Print view View original

21–30 of 509 posts

Re: A chess robot in Moscow has broken the finger of its human opponent

#21

Post by ip26 »

twayt wrote: post #18
est wrote: post #11

Earlier quoted context omitted.

From the looks of it, that's an industrial robot arm which was designed to operate no where near humans. The move paths were hard programed and has no obstacle avoidance procedure nor sensors. The robot arm had quite a strong pressure as the the victim were unable to free his/her finger.

Yea it is a very poorly designed chess robot but if they were going to use it in a competition like this, they should have at least briefed the users appropriately. Or maybe they did brief them appropriately but the child make a mistake. There isn't enough context in the video or article. But what's beyond stupid is claiming that a "chess robot broke the finger" of a kid as though it went rogue or operated beyond its…

Briefed the users? You mean the second graders it played against?

Re: A chess robot in Moscow has broken the finger of its human opponent

#23

Post by twayt »

carbadtraingood wrote: post #14

The answer from the federation is unconscionable. They blamed the boy and then said they could not be held responsible. Fuck off, a kid made a reasonable kid movement. If the robot wasn't ready to be around children, it shouldn't have been deployed around children. And there should have been a big red button that immediately everything, within reach of every player.

No post body was provided.

Re: A chess robot in Moscow has broken the finger of its human opponent

#24

Post by twayt »

ip26 wrote: post #21
twayt wrote: post #18

Earlier quoted context omitted.

Yea it is a very poorly designed chess robot but if they were going to use it in a competition like this, they should have at least briefed the users appropriately. Or maybe they did brief them appropriately but the child make a mistake. There isn't enough context in the video or article. But what's beyond stupid is claiming that a "chess robot broke the finger" of a kid as though it went rogue or operated beyond its…

Briefed the users? You mean the second graders it played against?

Yes, or the parents who let them play against it.

Re: A chess robot in Moscow has broken the finger of its human opponent

#25

Post by twayt »

klyrs wrote: post #20
twayt wrote: post #5

Quoted post unavailable.

A killswitch wouldn't help here, the kid's finger was probably broken before he even felt the pain. Given the bot's behavior, it looks like it's programmed to move its arm down until it touches the table. It didn't notice that it encountered resistance, and kept pushing down. That's incredibly unsafe, and easy to fix: track actual vs expected position, and give up if any surprises are encountered.

Even if it wouldn't help in this exact situation, it should be a pre-requisite for any such tournaments just in case.

It is an incredibly unsafe robot, but if the chess federation outlined the risks, who is at fault?

Re: A chess robot in Moscow has broken the finger of its human opponent

#28

Post by yjftsjthsd-h »

twayt wrote: post #18
est wrote: post #11

Earlier quoted context omitted.

From the looks of it, that's an industrial robot arm which was designed to operate no where near humans. The move paths were hard programed and has no obstacle avoidance procedure nor sensors. The robot arm had quite a strong pressure as the the victim were unable to free his/her finger.

Yea it is a very poorly designed chess robot but if they were going to use it in a competition like this, they should have at least briefed the users appropriately. Or maybe they did brief them appropriately but the child make a mistake. There isn't enough context in the video or article. But what's beyond stupid is claiming that a "chess robot broke the finger" of a kid as though it went rogue or operated beyond its…

> But what's beyond stupid is claiming that a "chess robot broke the finger" of a kid as though it went rogue or operated beyond its domain.

No, I'm comfortable arguing that being able to exert enough force to break human bones is outside the scope of what a chess playing robot should be able to do. Or is "beyond its domain" as you say.

Re: A chess robot in Moscow has broken the finger of its human opponent

#29

Post by kstenerud »

twayt wrote: post #23
carbadtraingood wrote: post #14

The answer from the federation is unconscionable. They blamed the boy and then said they could not be held responsible. Fuck off, a kid made a reasonable kid movement. If the robot wasn't ready to be around children, it shouldn't have been deployed around children. And there should have been a big red button that immediately everything, within reach of every player.

Quoted post unavailable.

This is why in civilized societies the government mandates minimum safety standards so that a company is not even allowed to place such dangerous equipment around people.

You wouldn't be allowed to build a mangler anymore in a civilized society, waiver or no.

Re: A chess robot in Moscow has broken the finger of its human opponent

#30

Post by Nextgrid »

kstenerud wrote: post #22

One has to wonder why the hell they'd make a chess robot with a grip so strong that it takes the concerted effort of two adults to pry it loose. And blaming it on the kid was real classy.

The robot is a standard industrial robot arm that happens to be used for chess - it it was not made to be a chess robot. This is absolutely overkill for chess even just from a cost perspective so you wouldn't build something like this if you actually were building a chess robot.

The amount of pressure this robot applies is totally adequate for its normal operating environment which is far away from humans.

There might be a way to vary the level of force applied but it may have been abstracted away (maybe you need to go one level lower to get that level of control) by a simpler API for applications where a "default" level of pressure is acceptable.

Post reply on HN