I'm very much against algorithms in criminal justice, particularly where it concerns a person's freedom (as in the case of bail and sentencing). I disagree very much with the entire premise of statutory treatment of alleged criminals. Indeed, many of the things people dislike about "bad judges"-- a lack of empathy/kindness/tolerance, an over- or under-estimation of the seriousness of the crimes or the seriousness of the threat of recidivism, a lack of interest in or consideration of the context in which the crime occurred, and an unreasonable faith in statutory sentencing-- can't be fixed with an algorithm. What the criminal justice system needs is not more rules. It's more flexibility.

>Algorithms would be another, better way to know exactly what makes the determination if a person should be released awaiting trial or not.

Isn't it a source of constant concern that the algorithms that end up being used in law enforcement inevitably end up employing the same stereotypes that their flesh and blood counterparts have?

But wouldn't algorithms be easy to examine? At the least you can test it with similar circumstances and modify the "race" variable and see if the results are consistent or different. Humans have many self-image protective layers when questioned (e.g. "I'm not racist, I have black friends!"), Bbut hopefully algorithms are more transparent...

Race variable is often correlated with other data.

Look at red lining in real estate/banking. They marked black neighbourhoods with red boxes, white neighbourhoods with green. The bank would approve business and home loans depending on if you were in a red or green area. Algorithmically the banks did not engage in a racist practice but systemically they are.

But did they mark red areas because they were black neighborhoods or because historically people in those neighborhoods defaulted on loans?

An algorithm as simple as "mark neighborhoods with high default rates as red, low default rates as green" would probably turn out to be "racist" even if it is also flagging poor white trailer parks with red.

There's no easy way to solve that

At least in the case of historic redlining (i.e. where the term came from), we don't have to wonder - there has been a ton of research done on this, and it was pretty clearly a racially motivated discriminatory practice. A fact which resulting in, among other measures, the fair housing act in the late 60s....

And it is very well known how to test and correct for that. In a human, however, it cannot be corrected. As you point out, it can be very hard to see what rules have to do with races.

I would argue that in some cases it's bloody hard to see that you're doing it yourself.

An algorithm can be racially biased, or have biases that significantly correlate with race, without actually having a "race" variable.

I think what he's suggesting is feeding the input data that differs only on that variable, and if the algorithm produces differing output, then it is by definition racist (and then you send the authors back to the drawing board to figure out why).

That's fair.

Machine learning models in particular are famously, fiendishly difficult to examine.

While agree with your general sentiment, you ought to point out that this very much depends on the algorithm in question. Things like decision trees/forests are quite observable.

> But wouldn't algorithms be easy to examine?

Examine by who? And would they be trustworthy?

They would be, but it turns out that humans make their own similar algorithms without realizing it.

>Algorithms would be another, better way...

Humans are flawed, and any system made by humans will be flawed. That's just reality. Whether the system is composed of human judges or human built mechanical judges, it will display the same biases of-, and will be vulnerable to the same manipulation by-, humans that exists today. People like cops and DA's will figure out, "Hey! Just feed it the right input, I'll get the output that I want!"

So if you want to switch to digital judges because you want to save some bucks by not having to pay the humans, well that works. But if you are expecting a better system at the aggregated mean, you're likely to be very disagreeably surprised.

we have ways to compensate for human flaws - like having a diverse group checking and balancing each other. it’s (seen as) inefficient but that doesn’t account for hidden costs like people’s lives being destroyed by injustice.

"There people always made the point that human judges making just decisions based on fairness."

I once had the 'pleasure' of working with a Magistrate. She was the most bigoted, opinionated person I've ever worked with. Judges are at best, just like everyone else.

I personally think the biggest problem is that people even track their "conviction rate", ESPECIALLY public prosecutors.

But it's worse for prosecutors: they get evaluated on it. Convicting people is all that matters. And then people complain the prisons are too full ...

If we replaced that metric by "how many years of freedom without serious crimes" did you create in defendants, using punishments ... and using just letting them go ... that would make a world of difference.

Those who will live by the algorithm will die by the algorithm.

People can be reasoned with, appealed to, threatened, or compelled in any number of circumstances, and that is human. Algorithms will simply be gamed and abused by those who control them.