Phrenology brain diagram

Your users want AI. Is your code ready?

Artificial intelligence (AI) is no longer just for science fiction films. Today, intelligence exhibited by machines is a reality in peoples' everyday lives.

How much time do you spend away from your smartphone in a day? Probably very little, if any at all. Smartphones remind us about appointments, tell us not just when to be there, but when to leave given our current location and traffic conditions, all in real-time, and without any human input.

This is just one example of the presence of artificial intelligence in our lives. So, what else will AI take over? Today, software engineers constantly look for ways to automate hard, repeatable data tasks using AI. And artificial intelligence is already taking on some traditional IT jobs, such as network security.  

The Mobile Analytics Playbook: A guide to better testing

Here's a look at what AI will mean for software engineers.

The recipe for AI

Robots and AI are taking on some IT jobs. Algorithms identify potential threats by scanning machine-generated data. As has happened in the manufacturing and agriculture industries, you can get to a much higher scale by automating repeatable tasks.

But AI works especially well when users are relying on data to make a judgment. Humans can only consume so much data before it overwhelms them. At the same time, under certain circumstances, our brains simply don't how to compute all of that information. When decision making is under severe constraints, and you have massive amounts of data, AI can provide the algorithms you need to make informed decisions. 

How does AI already factor into peoples' daily lives? Waze, the community-based traffic and navigation app, is a great example. The service aggregates large volumes of user-provided data on everything from traffic jams to disabled vehicles and police presence, and determines the optimal route (the decision). It’s not forcing the user to crunch all of that data; Waze does it for them.

Software engineers: Watch this space

For software developers, it’s about finding the problems that your end-users need to solve that rely on large amounts of data, that need to be made in relatively short time frames, and that rely ultimately on the user making a judgment that can be supported by artificial intelligence.

People still want control over the hardest, most valuable problems they need to solve. But they place a high value on applications than help prime that judgment in a way that is reliably more effective. What are the hard jobs your users do in which their decisions could strongly benefit from using data plus an algorithm to provide insights? That’s where you want to place your AI efforts.

In so doing you are continuing to find problems for AI to solve, and then implementing those AI features within the applications people are already using. However, if AI solutions are going to be widely adopted, the analytics that drive them must be embedded seamlessly into the interfaces that people already use.

As James Mayes, co-founder of Mind The Product, says:

“The UI is usually visual analytics of some sort. If those analytics are embedded in an application to help solve specific problems, then AI will help us automate data analytics tasks as much as possible. [AI] is very unlikely to replace the human in the process, though—it is likely [we] will make the final decisions.”

Why your users win when you include intelligence amplification

Today, most people believe our brains are different enough that computers will not gain sentience—the ability to prioritize on which problems to work first. But is anybody really sure? Some would argue that we’ve already made the transformation to cyborgs (life forms augmented by technology). Just look at how attached people have become to their smartphones.

Eventually, we may see more direct interfaces between computers (machines) and human brains. But for now, AI is all about using advanced algorithms to move information into context. AI is a learning technology, but a lot of what machines learn still comes—and will likely always come—from human feedback.

The future won't be about robots. It will be about us—the cyborgs. We are already augmenting our lives with technology, but the technology is something we control. It is intelligence amplification.

The Mobile Analytics Playbook: A guide to better testing
Topics: App DevMobile