What happens when you let algorithms make all your decisions for you?

life by algorithm algorithms tell me what to do for a day — Image used with permission by copyright holder

Like a lot of people my age, I love and depend on the Internet. It’s my secretary, accountant, keepsake, journal, and sometimes the type of friend you fall back on when you’ve exhausted all other options. And now it can play another role: Magic 8 Ball.

The rise of the so-called “recommendation engine” started with apps like Pandora. It practically changed my life in college when I discovered I could just plug in a song I liked to let a computer find more songs I liked, no need to set down my Solo cup. Since then, algorithms have mastered a lot more than just music. You can now find a recommendation engine for everything: what to eat, wear, read, give as a gift, where to shop, travel … almost anything.

So what would life be like if we actually listened? Would we be happy with the paths it sets us on? Could it make life easier, more efficient, and break us of boring routines?

I hoped so, but there was only one way to find out. So I turned as many life decisions as possible over to the Internet for 24 hours. Here’s how it went.

6:45 a.m. – What workout should I do?

Perhaps surprisingly, an app to suggest workouts based on algorithms was by far the most difficult to find, but a search did turn up Sworkit. After entering the type of routine you’re looking for (cardio or strength training) and how much time you have, the app builds a workout for you from a randomized database of circuit-training exercises.

In order to truly allow the Internet to make such suggestions for you, you really have to let go of your hangups

Here’s the thing: It’s pretty hard to do some of these at home, which the app doesn’t account for. For example, I wanted to do cardio … but Sworkit wanted to start me off with a minute-long sprint. With access to an indoor track or treadmill at the gym, fine. But sprinting up the hill next to the busy road I live on seemed unsafe.

I chose a strength-training workout instead, and got in about 45 extra minutes of fitness I otherwise wouldn’t have. But because I’m like an obnoxious border collie if I don’t spend enough energy, I already knew I’d be going for a run after work. Still, I was sweaty and tired after the session, and I’ll definitely reuse Sworkit. But I want an app that asks me more questions (What kind of workout do you want to do? Where? For how long? What are the results you’re looking for? What kind of shape are you in?) and then makes suggestions like how long I should run for, how many crunches I should do, or how many squats.

7:48 a.m. – What should I eat for breakfast?

This is what I’ve most been looking forward to: food. I like food but hate decisions and waiting, which means I eat a lot of the same things. Cereal, rice cakes with peanut butter, bananas, yogurt, eggs, cheese, avocados. Sometimes these are combined into real meals, other times I eat scoops of raw avocado with a spoon.

For breakfast, I hit up AllRecipes Dinner Spinner app, which luckily has a breakfast option. The app lets you pick dish type, ingredients, and prep time, and from there spits out your options. You can get more specific as well (so you paleo weirdos out there can avoid … everything delicious).

After selecting “Breakfast,” “Vegetables,” and “ready in 20 minutes or less,” I scrolled through the results, most of which weren’t breakfast items. The robust database gave me plenty of choices, but that choosing also negates the convenience of having a computer choose for you. I settled on cilantro and garlic eggs with parmesan, but substituted baby spinach for cilantro because I don’t hate myself enough to eat cilantro.

Is it good? Yes. Would I have rather had a mini yogurt from the fridge? Hell yes.

11:43 a.m. – What should I wear?

As someone who works from home three days a week and usually doesn’t change out of sweatpants until noon, this may be the area where I most need an algorithm to help me. In my search to find the best free services, I realized I needed two: One that had style in mind, the other weather.

The What Should I Wear app tells you what to wear based on weather. Surprise: I live in the Pacific Northwest, it’s raining, and telling me to wear a coat isn’t exactly revolutionary. The utilitarian aspects of dressing myself aren’t what elude me. If it’s cold, I’ll wear a sweater; if it’s not, I won’t. It’s choosing what colors or fabrics I should don that I need a hand with. Because hell if I know.

Dailydressme.com has a solution. The app makes its picks based on weather as well as what pop culture says is fashion. It was developed by a 16-year-old who, according to the site, wanted to “answer every girl’s biggest questions in the morning – ‘What’s the weather?’ and ‘What should I wear?'” I’m usually asking “Why is it so bright? What did I do to deserve this hellish nightmare called waking up at 7:30?!” but that’s probably why I need this site. So I typed in my city and this is what the technology suggested I wear:

Monday — Image used with permission by copyright holder

Is that a hat that lost a fight with a scarf? What am I’m looking at? I’m sorry, but I am just nowhere near cool enough to pull this off. And since I wasn’t born and orphaned in a hip downtown alley and then raised by Urban Outfitter employees, I don’t even own the requisite clothing items.

I managed to pull together the normal person version of this: jeans, shirt, coat, boots. I even added the hat.

Given the fact that I usually can’t be pried out of running leggings and sweatshirts, I was feeling overly fancy. Still, when I hit refresh and saw what other abominations I could’ve been suggested …

ddm other — Image used with permission by copyright holder

… I immediately knew I wouldn’t be returning to Dailydressme.com.

12:00 p.m. – What should I listen to?

Since I spend ample time working online, I can almost always be found Spotifying. But I prefer to rely on my own playlists instead of using the radio feature, which I’ve only found mildly successful. But I’m giving myself over to algorithms, so I tuned Spotify radio to the Sol radio station. According to Spotify, it works based on “a combination of our own data based on billions of hours of on-demand listening from our users, and social data and thumbs data (liking it/saving track to playlists).”

Occasionally, I heard a song I liked a lot … otherwise they all sort of blended together. Actually, that was a good thing: It was like a background hum that kept me productive, entertained but not entranced. Here’s the station.

2:30 p.m. – What should I do?

If the Internet has taught us anything, it’s that we are constantly bored. Some of the most common Google searches are “what to do in [insert city here]” or “things to do when you’re bored.”

In order to truly allow the Internet to make such suggestions for you, you really have to let go of your hangups. Hangups being things like a desire to stay inside, inability to go somewhere without a crew, or not wanting to pay for anything. But in the spirit of this experiment, I’m going to see what a few apps come up with.

I like food but hate decisions and waiting, which means I eat a lot of the same things.

First I tried FindGravy, which basically accumulates all of the stuff advertised happening in your area, and you can sort it based on mood and time. Feeling classy? Like you want to be outdoors? Artsy? All options you can choose. Unfortunately the options for the day I tried it were expensive, far away, not actually happening that day, or horrifying. (Open mic night? Can’t do it, won’t do it.) Many were LivingSocial deals, which is a FindGravy partner.

Being the recluse that I am (not really, only sort of), I decided to try Wtfshouldidotoday.com, which suggests random activities as well as time-absorbing websites for you to waste time on.

And waste time I did. On this. And this. And this.

So I headed to Shouldidoit.com, a true Magic 8 ball-type app which gives you yes or no answers. There was a brief break in the rain so I asked, “Should I go for a walk?”

Yes.

Look how easy that was!

“Should I go get coffee?”

Yes. Again with the straight answers! Obviously, this is a last option – and not to be taken seriously, as its disclaimer explains.

4:00 p.m. – What should I read?

One of the things Wtfshouldidotoday.com told me to do was read a new book. So instead of aimlessly wandering Borders or flipping through suggestions on my Kindle, I headed for the apps.

the ask book I tried Whatshouldireadnext.com. The site’s instructions are simple: “Enter a book you like and the site will analyse our huge database of real readers’ favorite books to provide book recommendations and suggestions for what to read next.”

I entered The Corrections, and one of the top recommendations was Nigella Lawson’s Feast Food that Celebrates Life. That wouldn’t do. So I tried Jonathon Franzen, which directed me to a few of his novels to click. I chose Freedom, which I love, and it suggested The Ask by Sam Lipstyne.

Major success: I am interested in a variety of things about this novel. For one, as someone who previously worked in academic fundraising, I appreciate any and all stories about this strange and terrifying world. Secondly, financial struggles for recent college grads? Yes please!

6:15 p.m. – What should I make for dinner?

After my breakfast experience with AllRecipes DinnerSpinner, I delegated the decision making for dinner to Whatthefuckshouldimakefordinner.com. The site is completely random, meaning it’s not actually algorithm-based. Instead, like WtfshouldIdotoday.com, it sends you meandering through the Internet’s trove of recipes, in crude fashion.

According to Whatthefuckshouldimakefordinner.com, I needed to eat Orecchiette with pesto rosso. Which sounded like a lot more work than the string cheese and rice cake with peanut butter I planned on “making.” Now properly dressed by an algorithm, I took myself down to QFC to find the necessary ingredients. Except I don’t want pesto because it’s salty. And some of the items were way too expensive, so I improvised. Orecchiette with tomatoes, spinach, and zucchini in tomato sauce it is!

dinner It was slightly more work than a rice cake, but worth it. It also inspired me to buy and eat zucchini, which I thought I hated, but in fact, love.

7:00 p.m. – What should I watch?

I have a few shows I religiously return to: 30 Rock, Breaking Bad, The Wire, or Parks n’ Rec. Instead of doing this, I used Popcorn.fm and Jinni to find something new.

Jinni was built off the human intelligence of entertainment and language experts to determine what you want to watch based on your feelings. The app gives highly customizable filters; you’re encouraged to get very specific. Based on what I filled out (comedy, witty, TV) I got The Big Bang Theory and M.A.S.H. Neither are available on Netflix. So I tried Popcorn.fm, which like so many of the apps I’ve used today, asks you to name something you like and finds something else with similar qualities. I told the recommendation engine I like the movie Like Crazy and it recommended The Art of Getting By. Also not on Netflix.

This was frustrating.

Finally, I tried basing the suggestions on a few TV shows I like to get another recommendation: Chasing Amy. I had never seen it, it was available on Netflix, so away we went!

My thoughts? It’s not my new favorite, but it’s a solid movie, and I’m glad I didn’t just rewatch season four of 30 Rock.

The verdict

First off, algorithms are exhausting. I thought making my own decisions all day was tiresome, but allowing a recommendation engine to do it is far more energy consuming. Whereas I will factor in convenience, my energy level, and how badly I really want this thing, a computer won’t – and usually has far higher expectations for me than it should.

Which … might be the best part of letting algorithms control your life. I’m all for keeping the bar low and manageable in certain respects, but challenging yourself obviously has its perks. Decisions via algorithms lead you to new discoveries, and while some of those will just flat out frustrate you (hello, dress-me algorithms!), others will just push and prod you behind your comfort zone. And then, you will eat zucchini. And you will like it.

As with all things, using algorithms for daily living taken to the extreme is not something I’d suggest. But handing yourself over to the power of such applications, if but for a day, will yield some happy surprises.

I let algorithms tell me what to do for a day, and this is what happened