Our legal system, at least in the US but I imagine in many other countries as well, has gradually become so broken that it barely serves its original purpose anymore, and while by all rights it should be improving, it is instead collapsing into a wildly manipulative affair that falls a long way from, “justice.” There are multiple factors behind this, but I’ll stick to the larger ones because I’m not going to do a treatise, and there’s a point that I’m about to examine.

Let’s start with:

Over-reliance on human testimony. By far, our courts rely on witness statements, which is perhaps the weakest factor that they could. Humans are notoriously prone to subjectivity, emotional bias, suggestibility, false confidence, and even “gut feelings,” something that a recent experience with jury duty reinforced, suggesting that we the jury ask ourselves whether we thought the witness’ statements felt correct. This is patently ridiculous, since courts are there to present evidence, not solicit feelings, and objectivity is intended to abolish whether we like some aspect or not – this is likely never completely possible, but at the very least the attempts should be actively encouraged, rather than subverted in this manner.

More specifically, however, humans make terrible witnesses, and this has been established and clinically proven time and time again. What someone says that they saw or heard should be taken with a grain of salt, especially if there is nothing solid to corroborate their statements. We tend to forget how often people lie, to say nothing of how often they’re simply dead wrong, and/or wildly mistaken in what they’ve seen or heard. Attorneys actually know this, because they’ll coach their clients, again and again, on exactly what to say, trying to establish confidence through repetition. The further people are from an event, the more likely their recollection of it is damaged or altered, and this has been well-established too, and preyed on just as much by attorneys.

Humans are notoriously biased. You can instruct jurors on penalty of expulsion or mistrial not to let prejudice enter into their decisions, and it won’t stop the lifelong, ingrained biases that they possess. The idea is that this balances out with multiple jurors, but in reality? Not so much, especially in some cultures, or portions of a city, and especially not if the attorneys are tasked with helping select jurors. And this says nothing whatsoever about judges, public defenders, and so on. It’s even hard to determine to what extent this takes place, though long-term examination of data and decisions often reveals some nasty little trends.

Expert testimony is too often anything but. It’s fairly well known that police officers are usually considered impeccable witnesses – certainly their testimony will be considered much more reliable than any other participant unless strong countermanding evidence is presented, as if becoming an officer instills perfect objectivity while also eliminating bias – just saying that highlights how ludicrous this is, though most people never even raise the question. But at the same time, outside testimony, for instance by metallurgists or handwriting experts, is often selected carefully by the attorneys because such testimony hews closest to their desired outcome. No one is a certified “expert” that cannot be wrong, and such testimony thus comes down to how convincing they make their statements and how well, or poorly, the opposing attorney has sought their own expert witness. Moreover, precision in many of the sciences simply doesn’t exist, and despite the impression that DNA testing and such offers incontrovertible evidence, almost no one steps in and points out how often such things can and do go wrong.

Nearly everyone involved is more motivated by their own career, advancement, and ego. Judges, attorneys, police officers, expert witnesses – the idea of serving proper “justice” is not usually paramount to most of them. Instead, it’s their record, and their politics, how many ‘wins’ they garner, how high-profile the case is, what kind of settlement it’s likely to produce, and so on. Everyone is well aware of the “hanging judge,” or the public defender that doesn’t put much effort into drug cases, and so on; attorneys routinely juggle the court dates of their clients to aim for the desired bias.

It’s ridiculously, irrevocably, undeniably money-driven. With few exceptions, the client with the most money will win their case, and attorneys will even refuse cases based on who they’re up against rather than the merits of the case itself. While it may not be as distinct as implied here, a legal team receiving more money will devote more time to reviewing case histories for applicable precedents, seek out better testimony, and most especially, bend over backwards to ensure a larger settlement and thus a greater payoff from their percentage (which is often ruinous, and not in the slightest commensurate with effort involved or an ‘hourly rate.’)

It’s also disturbingly true that representing oneself is almost guaranteed to fail; the legal system is so byzantine and convoluted that trying to negotiate it without years of law school is impossible. Clients that cannot afford a decent attorney will therefore never get representation, in the vast majority of cases, meaning that many, many crimes simply never go to trial, something that large corporations prey upon ruthlessly.

There’s a lot more, but suffice to say, it’s fucked up.

Now here’s the curious, potential solution that’s been nagging at the back of my brain for a little while now: this is what artificial intelligence can actually provide for us.

Not in its current state, no – what we have now is not even remotely artificial intelligence, it’s just algorithms to find patterns. And it’s possible that we may never develop a system that mimics natural intelligence to any serious degree. But bear with me for a moment.

The biggest issue with artificial intelligence pursuits right now, that many don’t even realize is an issue, is that the biological imperative defines what we are and what’s important to us. Actually getting humans to think rationally requires effort – otherwise, we default to indulgences and reactions and feelings. It’s what worked best for us among the choices that arose, all during our evolutionary development, but it’s so intertwined with our thought processes that it defines us and guides us, and depending on what purpose we might want to develop AI for, it could be exceedingly difficult to ‘program in.’ It’s also what introduces a tremendous amount of the difficulty in producing a functional, objective, and fair trial of justice. Hell, we can even argue about what justice is and what our morals should encompass.

Very little of that would actually have to be present for a justice AI, and the lack thereof is what would produce the fairest and most focused system in the first place. No ulterior motives, no biases, no underlying self-aggrandizement, no past experiences that color perception, and especially, no corruption. As the saying goes, just the facts, ma’am.

It would take a lot of work, without question – but quite possibly a lot less than self-driving cars. Attorneys would still be necessary (that’s a hard sentence to type) to prepare the cases for presentation, at least for a while, though a system that can search a broad database for precedents and interpretations is well within our abilities now. it even has the advantage of being able to determine, through the same kind of data searching, what forms of ‘expert’ testimony are anything but, and where the flaws lie.

Right now the idea is only half-baked in my own head, and needs more consideration; witness statements, for instance, which should probably take place in the forms of immediate recordings and specific questions to eliminate the subjective and exaggerating aspects – though this should be the case regardless. Cross-examination? Is this possible for an AI to accomplish functionally?

But overall, it suggests that it could potentially produce the environment that we’ve striven for all along, the elimination of human pettiness and imperfect concepts of justice, as well as the aforementioned biases and unrelated motivations. I’m at least finding the idea rather intriguing…