Reviews | Of course, LinkedIn is experimenting with “weak ties” on you.


There is an aphorism about social media sites that warns consumers that they are not consumers after all – they are the product. That’s true, but there’s more: they’re also test subjects.

A study in the journal Science by researchers from LinkedIn, MIT, Stanford University and Harvard Business School reveals that the professional networking platform has been experimenting with more than 20 million users for five years. The idea was simple: harness the power of big data to confirm the sociological hypothesis that less close acquaintances are more useful for finding a job than closest friends.

The attempt to prove the theory worked, apparently too well. Now, some critics say LinkedIn gave some users a leg up while leaving others languishing — carefully improving their product but carelessly playing with people’s livelihoods. Maybe so, even though LinkedIn was ostensibly trying to help everyone get a job more effectively in the long run. But the existence of the study is hardly surprising.

LinkedIn assessed what’s called “weak link strength” by adjusting its “People You May Know” algorithm so that it recommends weaker contacts (e.g., your roommate’s golf buddy’s boss) to certain stronger contacts (e.g. your college roommate) to others. It’s called A/B testing, and it’s much less controversial when someone else is doing it.

The Post A/B tests headlines, for example, showing some readers a version of one with “Donald Trump” and others a version without to see which gets the most clicks. Fast food A/B testing burgers, such as showing some eaters a version of a digital menu that puts Big Mac combo meals right in the middle of the screen and others a version that promotes Quarter Pounders no soda or fries. Does Ronald McDonald experiment on us ?

The difference is that, as important as lunch (and even breakfast) may be, calorie intake doesn’t feel as personal as job prospects, or, in Facebook’s case, as intimate as photographs. and reflections from family and acquaintances.

This explains why a 2012 Facebook A/B testing unveiled in 2014 inspired anger similar to that of today – but greater: the platform had served some users more positive messages and more negative messages to determine whether the change had changed their mood – which , spoiler alert, did. Mark Zuckerberg manipulated our emotions to make money!

Well yeah. Exactly.

Surely there are ethical questions about the implications of big data research. And we surely feel violated when a study we participated in – but had no earthly idea about – is published. We are not lab mice. But on the other hand, we somehow are – even when there is no official scientific study in progress, and even when everything of us rather than a pre-selected subset or two are affected by the decisions of social media sites. Every adjustment, every step of product development does something for each of us.

YouTube adjusts its artificial intelligence to recommend more extreme versions of what we’ve already watched so we keep watching, and sends those drawn to conspiracy theories down a rabbit hole they can’t get out. Youtube readjusts the algorithm to prevent the spread of harmful false information. It’s not exactly an experiment, but it’s a way to influence human behavior with implications for society as a whole. Twitter is increasing its character limit from 140 to 280, and our political discourse is changing.

The aphorism declaring us a product of these sites comes from the basics of online advertising: Facebook is “free” only because the platform draws users’ attention to companies trying to sell us things. As long as the sites aim to gain from us something as ingrained in our minds as Warningwe will also continue to feel like guinea pigs.

Either this is all a moral disaster, or this is exactly what we signed up for. We’re on Twitter, and YouTube, and Facebook, and even LinkedIn because we want them to do something more meaningful in our lives than a burger and fries: to connect us, or inform us, or make or break. our professional future. We’re onto them because they’re personal, not despite that – and it’s only because we’re attached to them that many of us find it so hard to let them go even when we know full well we’re their mouse lab. .

The options are to log out or continue running through the maze.


Comments are closed.