Do you need AI to solve crime?

One of the biggest friction points I hear from investigators is whether they even need AI to solve crimes.

Why would they? They’ve spent several years doing it. They’ve built the intuition and have the creativity to spot things that most people can’t.

It’s a feel thing, right?

Yeah, and I mostly agree with you, but it’s never been about replacing investigators with AI.

It’s about equipping them with the tools to solve crime faster and more efficiently.

And that’s exactly what AI does. Below, I’m going to break down the problems I see with how people currently approach investigations. Then, I’m going to point to a few other technological advancements that people didn’t think we needed and outline the 2nd and 3rd order effects most people don’t consider.

The Problem with How We Investigate Crime Today

I come from the startup world. In the startup world, we’re always looking for a problem to solve to make things better, faster, or more efficient. When you break down the word technology, it basically means using a new technique (technology = technique) to solve a problem.

I’m uniquely positioned to tackle this problem because it has directly affected me with the disappearance of my mother. If my mother were still here, I would have never gone on this mission.

Why?

Because it’s not fun, you generally want to choose a problem to solve in technology that’s fun. Thinking about murders or missing people can be traumatizing, and that’s why I also empathize with investigators.

They have to spend all day juggling dozens of cases and thousands of files while uncovering the most disturbing side of human beings. How can someone wake up the next day and maintain a positive attitude or faith in the human race when they see the worst side of people?

I have no clue, and I commend them for doing that, but the fact of the matter is that something isn’t working. I point to the numbers all the time, but unsolved murders continue to increase.

The reason as to why is crystal clear:

Resource constraints.

Police stations don’t have the time or resources to dedicate energy to solving some of these cases. There are bottlenecks in their workflow. There are bottlenecks in their chain of command. There are bottlenecks in their file management.

The list goes on and on. To top it off, they have to do all of this surrounded by red tape and bureaucracy.

One of the primary reasons Wheatridge PD won’t release my mother's case files is that they're afraid of what a defense attorney will argue in court. This raises another issue: an inability to accurately assess risk.

If your entire job is to find a way to avoid and minimize risk, it will be very difficult to make advancements in your industry. The startup world embraces risk. Risk is what results in some of the world's greatest innovations.

People had to take a risk to see an opportunity and then convince the world it was a better opportunity. Without risk, innovation is stifled. Investigators aren’t equipped to take risks because they are dealing with people's lives. If they take the right risk and it works, there is no upside for them.

If they take the right risk and they’re wrong, it’s career suicide. It’s a very difficult position to be placed into.

But the truth is, it’s time to take risks in the criminal justice system because the current system isn’t working correctly. Again, I point to the numbers. If the numbers are going up, you need to reassess how you solve problems, and the criminal justice system and investigations need to be reassessed.

When I sat down with the Wheatridge PD, one of the detectives kept saying things like “I don’t see how AI can help us. We looked at everything repeatedly. We’ve been thorough, and now we have another detective looking at things. How can AI help?”

This mindset is prevalent throughout the history of technological advancement. There is always a group of people who say they can’t see how something new could help society.

When cars started to emerge, people likely said, "What's the point when we already have a horse and buggy?" It does a good enough job to get from point A to point B. Besides, cars are dangerous, and you can get hurt or kill someone!

Well, the 2nd and 3rd order of effects for a car is that you get ambulances, which means if someone gets hurt, you can get them help faster. You can also transport materials faster, so you can build infrastructure to save more people.

Could you have done that with a horse and buggy, right? Sure, but a horse and buggy has too many constraints and variables you have to manage. To put it simply, it took too long to get from point A to point B, and you had to worry about keeping your horse alive rather than filling up your car with gas.

Think back to what it was like in medieval times. Only a few people had a horse, and if you got sick or injured in a village, it was likely over! Now, no matter where you are in a city, or even most of the country, there is infrastructure to send an ambulance to come get you, or you can get in your car and go to the doctor's office.

When cars first emerged, people didn’t consider all the additional use cases or the ways they would help society. They dismissed it and said they didn’t need it.

The same thing is happening with AI and investigations. The truth about AI is that most people don’t know what the 2nd and 3rd order affects are on investigations. They are likely to be very positive and drastic, though they could take a long time to play out.

What I do know is that AI is extremely good at parsing thousands of files and extracting information quickly and efficiently, something that humans can’t do. What AI does in minutes takes investigators weeks, months, or years.

The truth is that there is likely a lot of human error in the criminal justice system. It’s immeasurable, and likely will never be measured. That’s besides the point. How do I know this?

Because the investigators working these cases are bogged down with dozens of cases and thousands of files. They are surrounded by legal red tape and incentivized to minimize risk. This is a perfect recipe for human error, and it’s not really their fault.

What I do know is that AI gives them their time back, allowing them to focus on solving the case rather than digging through files manually.

Treating AI as one of the world's best junior investigators is a mindset unlock I’d like to see more investigators use, and I think it will eventually lead to breakthroughs, such as solving hundreds or thousands of cold cases.