Should The American Court System Adopt Artificial General Intelligence?
As wild of an idea as it may sound, one day in the very near future we could be (and likely will be) living in a much different world. A world more out of science fiction than one might think today. A world where artificial intelligence has become much more integrated in our everyday lives. As an avid lover of science fiction (both movies, novels and comic books) I can both see the incredible benefit and incredible danger in this technology. Even Elon Musk and Stephen Hawking have warned about the potential danger in "killer robots" in the foreseeable future.
Artificial intelligence has many practical uses in the real world across numerous industries. The obvious downside would be the loss of jobs on a potentially massive scale as industries adopt A.I. There are ways society could alleviate these problems like providing Americans with a universal basic income.
What is artificial general intelligence you may be wondering? AGI (artificial general intelligence) is the ultimate end goal of A.I. research. It is the ability for a robot, or android, to make intellectual decisions on the same level a human can...if not better. Within the scope of criminal justice, it would be the ability for a machine to weigh a case with a human level A.I. but without the biases that come along with humans. So a jury that didn't assume guilt of a black man or innocence of law enforcement officers. This is another radical solution that can be added to a list of radical solutions I've already suggested in the past.
Now back in October of 2016, the Committee of Technology within the White House released a report called 'Preparing for the Future of Artificial Intelligence.' It's an interesting report that raises questions about ethics and control of artificial general intelligence in the near future. One important thing that I saw in this report was the acknowledgement by the White House that A.I. should be used by the government to better and more effectively serve the public. Not only that, but the criminal justice system was mentioned (along with other sectors such as education and defense) as being an area where the progression of artificial general intelligence research could have positive impacts.
The report threw out a few suggestions in ways this technology could be used in the near future, but like always, none of them were as radical as I felt they needed to be. In the situations mentioned, it was always the A.I. being used as a complementary tool. Not a tool to replace the current structure in its entirety, or at the very least in certain cases that involve crimes committed by actors within the State (so for instance law enforcement). So when the A.G.I. becomes powerful enough, maybe the entire process from filing charges, to the jury (if there is a Jury) should be automated independently of Human beings. If there is not Jury and it's a bench trial, then perhaps the Judge should be an android with human-level A.I.
If one thinks that'll never be possible, think about this - computer scientists from University College London have already created an artificially-intelligent "Judge" that could not only weigh legal evidence and moral questions of right and wrong, but was also used to successfully predict the outcome of hundreds of cases! The A.I. "Judge" that the British scientists created also came to the same verdict in 4 out of 5 human rights cases involving torture, degrading treatment and privacy!
The human brain right now has the edge in terms of general intelligence. If one were to think of neurons in the human brain as being equivalent to a computer's CPU (aka core), then advanced supercomputers have up to 10 million cores. The human brain on the other hand has about 100 billion neurons. So we're still several decades away from human-level artificial general intelligence. The point though is the upward trend and eventually we will reach a point where the rubber meets the road as the saying goes.
In recent years, the application of technology to aid law enforcement has been completely and totally slanted. The government gets all the toys to monitor the public, our internet activity and our purchasing habits. Drones are now being used at major events and during protests by law enforcement as well. I think it's been past time to begin to use technology to actually aid the public and not the national security state when it comes to issues of criminal justice. Recent events have shown flaws in our system so deep and entrenched that one is forced to wonder whether or not unbias in the system is even possible...since bias seems to be the de facto starting point. Artificial general intelligence can begin to nullify some of those problems.
The technology would clearly have to pass very rigorous tests and the coding algorithm and approval process should all be open to the public. There should be no question of whether or not A.I. robots and androids are being built to further exasperate the problem.
To make a long story short, this technology could significantly reduce wrongful convictions. One could even foresee a near-future where A.I. has surpassed human intelligence and can even read a person's mind - a type of polygraph on steroids and not nearly as flawed. One thing is certain, our criminal justice system that we have now is controlled by humans and humans have biases and emotions that cloud their judgement and make it hard, if not impossible in some cases, to get a fair trial. Even with checks and balances in place, the system is thoroughly corroded and full of holes - whether by design or not. An imperfect and flawed justice system can never be expected to maintain freedom in a democracy - and ours will fail too. It already is failing.