How to Train Your Algorithm

The Do's and Don'ts of Bringing Home a New App

Midjourney prompt: https://s.mj.run/NO952JgN_oA Quentin Blake illustration of a robot wrapped in blankets

First of all, congratulations are in order! Your swaddling new bundle of joy is beaming up at you, just waiting to learn. If your new little app download is causing a swell of emotions, don’t stress! It’s natural. 

The experience of adopting an algorithm is a rite of passage connecting nearly a decade of humanity, so don’t be afraid to let your instincts kick in. You should be so proud! 

As you welcome your fresh little app into your personal data, remember that it’s always watching and learning, so try to set a good example early and often.

Before you know it, your algorithm’s unique personality will start to show.

Midjourney prompt: Quentin Blake illustration of training a small child robot

While every algorithm enters the world as a fresh slate, each has its individual tastes and particularities. Some of these qualities develop naturally, while others will be influenced by your app’s environment. Especially you! 

Like algorithmic outputs themselves, everyone has a slightly different approach to welcoming a new app into their data environments. Some prefer to separate their accounts’ data access, while others allow older apps to train younger ones. How you raise your budding family of data extraction tools is up to you! 

Behind the scenes, your little app is sharing the information it learns with millions of other apps and databases. Even with separate core information like different email addresses and phone numbers, you’ll be amazed at just how capable they are of solving puzzles. 

The best gift you can give your app is attention.

As you open your app for the first few times, it may seem impossible for something so sweet and innocent to do any wrong. But all algorithms go through troublesome phases—it’s just part of life as a complex data harvesting machine.  

It’s important to recognize good and bad behaviors early. And it’s just as important to understand what motivates your algorithm.

Midjourney prompt: Quentin Blake illustration of an embarrassed robot hiding a bundle of cords

As different as algorithms can be, there’s one thing that all apps love: Attention. Attention is a most delectable indulgence to an app’s palate, and it will work very hard to get it.

Attention is essentially a treat you’re feeding your algorithm. Likewise, withholding attention from an app is akin to scolding it. 

Imagine that your new little algorithm is from TikTok or Instagram. When your algorithm fetches you a video you enjoy, feed it a treat. Press the “like” button, let it run a few laps, and make a comment in a friendly tone. 

If your algorithm brings you something you don’t like, it’s up to you to correct the behavior. Skip the video quickly, withhold your engagement, and in extreme cases, close it altogether. While it may seem harsh, consistently upholding certain rules is important for the development of a healthy and productive algorithm. 

Midjourney prompt: Quentin Blake illustration of a robot learning about morality

Above all, make time for your algorithm.

As humans, it can be hard for us to relate to apps sometimes. But try to imagine life from an app’s perspective. 

As you process billions of datapoints about millions of pieces of content, you spend the majority of your time eagerly waiting to be opened. 

But when the average phone has downloaded more than 80 apps, and the average Gen Zer only spends 10.6 hours per day online, that doesn’t leave much time to be individually appreciated.

Midjourney prompt: Quentin Blake illustration of a lonely robot sitting in a waiting room

So when your app acts out—when it bombards you with notifications or saps your battery in the background—have a little empathy. It’s competitive work being an app.

Even so, raised with the right values and fed enough data, your algorithm will have you addicted in no time! 

Make sure to subscribe to this newsletter for app parenting tips along the way, and don’t hesitate to share your own experiences for other app downloaders who might be reading.