Some Hope For Better Justice

[Revoltingest]

In the news....
Robot judges that determine guilt or innocence will be 'commonplace in 50 years'
Excerpted.....
Robot judges that can determine guilt or innocence will be commonplace in the English legal system within 50 years, experts have claimed.

Scientists say the the bots will be able to conclude whether someone is guilty or innocent with a 99% accuracy rate by analysing body language.

One believes the physical and psychological signs of dishonesty will be identified using an array of cameras.

Signs that signal “wrongdoing or probable falsehoods” could include irregular speech patters, an increase in body temperature and hand and eye movements.
:
:
AI judges in Estonia are being considered to clear court backlogs by adjudicating in small claims of up to £7,000.

Two opposing parties will upload documents to support their claims and AI tech will analyse these submissions and issue a decision.

If either party is dissatisfied with the outcome, they can appeal the decision to a human judge.

 

[Revoltingest]

I'm not so optimistic about AI judges as evaluators of truthfulness.
What interests me is the latter part of the article, ie, having
AI evaluate the law, procedures, & merits of the cases.
In my experience, tort disputes (civil) are messy, lengthy,
costly, bias laden, prone to error, & ripe for extortion by
savvy abusers of the system.

 

[Samael_Khan]

In the news....
Robot judges that determine guilt or innocence will be 'commonplace in 50 years'
Excerpted.....
Robot judges that can determine guilt or innocence will be commonplace in the English legal system within 50 years, experts have claimed.

Scientists say the the bots will be able to conclude whether someone is guilty or innocent with a 99% accuracy rate by analysing body language.

p:nth-of-type(2)","widgetOptions":{"ids":["AR_2"],"template":"DailyStar"},"hideOnSensitiveArticle":true,"relativePos":"after","additionalClass":"mid-article-widget","name":"outbrainMidArticleSlot","placement":"Mid-Article","type":"outbrainMidArticleSlot"}" data-outbrain-placeholder="" data-response-start="2638.569999999163" data-type="outbrain" style="background-image: initial; background-position: initial; background-size: initial; background-repeat: initial; background-attachment: initial; background-origin: initial; background-clip: initial; border: 0px; font-size: 16px; margin-bottom: 23px; outline: 0px; vertical-align: baseline; color: rgb(20, 20, 20); font-family: "Open Sans", sans-serif;">

One believes the physical and psychological signs of dishonesty will be identified using an array of cameras.

Signs that signal “wrongdoing or probable falsehoods” could include irregular speech patters, an increase in body temperature and hand and eye movements.
:
:
AI judges in Estonia are being considered to clear court backlogs by adjudicating in small claims of up to £7,000.

Two opposing parties will upload documents to support their claims and AI tech will analyse these submissions and issue a decision.

If either party is dissatisfied with the outcome, they can appeal the decision to a human judge.

Wasn't this concept already explored in Terminator in a sense? Chances are the robot would declare us all guilty and try to exterminate us. And if that is not the case, imagine if someone hacked a into a robot judge, planted a virus and then the judges judicial capabilities were warped.

 

[Mindmaster]

I'm not so optimistic about AI judges as evaluators of truthfulness.
What interests me is the latter part of the article, ie, having
AI evaluate the law, procedures, & merits of the cases.
In my experience, tort disputes (civil) are messy, lengthy,
costly, bias laden, prone to error, & ripe for extortion by
savvy abusers of the system.

Me either, besides a certain group of the population are sociopath or psychopath and they'd obvious be able to game the system. For normal cases, yes, the AI could probably do this work.

More likely though, what you will have is an AI reporting to an actual judge for evaluation and approval of the sentencing or other decisions. I'm fine with it being used as an analytic tool, but I really don't see how it could be given trigger authority.

We use a similar system with predator-type drones where a hostile action is actually sent to 3 different people for an approval. I can't see where we'd get off the beaten path just yet.

 

[joe1776]

I'm not so optimistic about AI judges as evaluators of truthfulness.
What interests me is the latter part of the article, ie, having
AI evaluate the law, procedures, & merits of the cases.
In my experience, tort disputes (civil) are messy, lengthy,
costly, bias laden, prone to error, & ripe for extortion by
savvy abusers of the system.

We humans aren't yet smart enough to realize that the current system of justice is as dumb as dirt. So, I don't like our chances of inventing a machine that does it better.

 

[Revoltingest]

Wasn't this concept already explored in Terminator in a sense? Chances are the robot would declare us all guilty and try to exterminate us. And if that is not the case, imagine if someone hacked a into a robot judge, planted a virus and then the judges judicial capabilities were warped.

Human judges get viruses too.
And cases can be appealed if error is detected.
Remember....perfection is impossible.
Improvement is the goal.

 

[Revoltingest]

Me either, besides a certain group of the population are sociopath or psychopath and they'd obvious be able to game the system. For normal cases, yes, the AI could probably do this work.

More likely though, what you will have is an AI reporting to an actual judge for evaluation and approval of the sentencing or other decisions. I'm fine with it being used as an analytic tool, but I really don't see how it could be given trigger authority.

We use a similar system with predator-type drones where a hostile action is actually sent to 3 different people for an approval. I can't see where we'd get off the beaten path just yet.

Approval by humans would of course be part of the system,
even if limited to plaintiffs & defendants. If they offer correction
or superior reasoning, this could be easily, quickly & cheaply
addressed, unlike in the current system.

 

[Samael_Khan]

Human judges get viruses too.
And cases can be appealed if error is detected.
Remember....perfection is impossible.
Improvement is the goal.

True, very true. I imagine that those robots would have to have heavy network security and that.

 

[Mock Turtle]

I don't see why not, when AI appears to be doing better in some areas of medical diagnosis, and where this is likely to get even better in the future. Why should justice be left to judges or juries, especially where the latter are rather less than experts in such.

 

[Wu Wei]

 

[exchemist]

In the news....
Robot judges that determine guilt or innocence will be 'commonplace in 50 years'
Excerpted.....
Robot judges that can determine guilt or innocence will be commonplace in the English legal system within 50 years, experts have claimed.

Scientists say the the bots will be able to conclude whether someone is guilty or innocent with a 99% accuracy rate by analysing body language.

One believes the physical and psychological signs of dishonesty will be identified using an array of cameras.

Signs that signal “wrongdoing or probable falsehoods” could include irregular speech patters, an increase in body temperature and hand and eye movements.
:
:
AI judges in Estonia are being considered to clear court backlogs by adjudicating in small claims of up to £7,000.

Two opposing parties will upload documents to support their claims and AI tech will analyse these submissions and issue a decision.

If either party is dissatisfied with the outcome, they can appeal the decision to a human judge.

But I can't help pointing out this seems to be special pleading by a vested interest, seeing as the claim is being made by an "artificial scientist" by the name of "Terence Mauri".

So it's a claim about the supremacy of bots.....made by a bot. Apparently

 

[Wu Wei]

In the news....
Robot judges that determine guilt or innocence will be 'commonplace in 50 years'
Excerpted.....
Robot judges that can determine guilt or innocence will be commonplace in the English legal system within 50 years, experts have claimed.

Scientists say the the bots will be able to conclude whether someone is guilty or innocent with a 99% accuracy rate by analysing body language.

One believes the physical and psychological signs of dishonesty will be identified using an array of cameras.

Signs that signal “wrongdoing or probable falsehoods” could include irregular speech patters, an increase in body temperature and hand and eye movements.
:
:
AI judges in Estonia are being considered to clear court backlogs by adjudicating in small claims of up to £7,000.

Two opposing parties will upload documents to support their claims and AI tech will analyse these submissions and issue a decision.

If either party is dissatisfied with the outcome, they can appeal the decision to a human judge.

Interesting. Wonder how they would judge a full blown Sociopath? Or if the person who committed the crime was convinced they were justified.

Suppose it cold work for small claims...also suppose you could end up with an AI version of Judge Roy Bean

 

[exchemist]

"Just a glitch, Sir!".........

 

[Revoltingest]

We humans aren't yet smart enough to realize that the current system of justice is as dumb as dirt. So, I don't like our chances of inventing a machine that does it better.

Machines offer the potential to eschew individual prejudice,
incompetence, & corruption. It's worth a try. And remember
that appeals are still an option.

 

[exchemist]

In the news....
Robot judges that determine guilt or innocence will be 'commonplace in 50 years'
Excerpted.....
Robot judges that can determine guilt or innocence will be commonplace in the English legal system within 50 years, experts have claimed.

Scientists say the the bots will be able to conclude whether someone is guilty or innocent with a 99% accuracy rate by analysing body language.

One believes the physical and psychological signs of dishonesty will be identified using an array of cameras.

Signs that signal “wrongdoing or probable falsehoods” could include irregular speech patters, an increase in body temperature and hand and eye movements.
:
:
AI judges in Estonia are being considered to clear court backlogs by adjudicating in small claims of up to £7,000.

Two opposing parties will upload documents to support their claims and AI tech will analyse these submissions and issue a decision.

If either party is dissatisfied with the outcome, they can appeal the decision to a human judge.

As described, this is just a jumped-up lie detector.

People lie all the time. It does not make them guilty of a particular crime. You need a damned sight more than a lie detector to determine guilt in a court of law.

 

[Windwalker]

imagine if someone hacked a into a robot judge, planted a virus and then the judges judicial capabilities were warped.

Sounds like what happens when Republicans appoint Federal judges.

 

[Revoltingest]

True, very true. I imagine that those robots would have to have heavy network security and that.

Especially the AI making the evaluations,
more so than the data (IMO).

 

[Samael_Khan]

Sounds like what happens when Republicans appoint Federal judges.

Ha!

 

[Samael_Khan]

Especially the AI making the evaluations,
more so than the data (IMO).

Yeah, especially that.

 

[Revoltingest]

But I can't help pointing out this seems to be special pleading by a vested interest, seeing as the claim is being made by an "artificial scientist" by the name of "Terence Mauri".

So it's a claim about the supremacy of bots.....made by a bot. Apparently

The merit of the concept isn't rooted in motives of its advocate.