Here’s a free marketing pitch that’s sure to be popular: Don’t be Facebook.
While its shares thrive despite the Cambridge Analytica scandal, Facebook has emerged as a convenient punching bag for other technology companies wanting to show off their virtuous bona fides.
It’s part of a trend that’s seen the word “ethical” tossed about in more and more press releases. Like “privacy”, it’s a term that means a lot, but very little without accountability.
At Apple’s developer conference in early June, for example, executives singled out Facebook and its method of collecting user data across the internet via the “Like” button.
“We’ve all seen these ... like buttons and share buttons,” Apple’s senior vice president of software Craig Federighi said, announcing a new anti-tracking option in Safari.
“Well it turns out, these can be used to track you, whether you click on them or not. And so this year, we’re shutting that down.”
Which site did Apple used to illustrate their new feature? Facebook.com, naturally.
If it wanted to get deadly serious, Apple could go a few steps further and boot Facebook out of its app ecosystem.
It won’t, of course. So how seriously should we take this vigorous virtue signalling?
Assistant professor of computer science at Princeton University, Arvind Narayanan, has a strategy for dealing with these sticky questions.
Facebook’s business model, based on targeted advertising, is dependent on collecting personal information from users.
Apple makes cash from selling hardware, with less incentive to collect every speck of data. In Dr Narayanan’s view, that makes their stance on privacy a little more credible.
For companies that want to seem good — unlike the other guy (read: Facebook) — the debate over whether our smartphones and social media are designed to be “addictive” is another opportunity to promote those differences.
Both Google and Apple, for example, recently announced new tools to track and control screen time. Some critics say that’s just another way to measure the problem rather than fix it, but it’s a compelling message.
Speaking at the Semi Permanent design conference in May, Airbnb’s vice president of design Alex Schleifer said it is refreshing to work at the travel accommodation company, whose business doesn’t “value” people spending endless amounts of time on their app.
“I think once you start being incentivised by people just staring at a screen you start putting in these addiction loops, which become really powerful,” he explained.
“We don’t always get it right, but at the very least, it’s great to be in a business that values and tracks just the quality of the human experience rather than the amount of time that they spend on the product.”
Who is being ethical?
It gets even more complicated when companies simply promise to be “ethical”.
A few years ago it was worker’s rights and the environment, these days it’s “killer robots”.
Google, for example, recently launched a set of principles guiding its research into artificial intelligence following controversy over its work for the US military.
It claims it won’t build tools that cause or facilitate injury, as well as aiming to “avoid creating or reinforcing unfair bias”.
But from privacy and addiction management to AI, how will we know these companies have lived up to their promises?
Dr Narayanan said he hoped companies would use “ethics” as a differentiator, but added that when they do, we may not fully trust their motivations or their oversight.
“I think the answers are not going to come entirely from tech companies,” he said.
Ellen Broad, who writes about data issues, said while many major platforms were investing in ethics programs, they can’t meaningfully act without accountability.
“It’s no great comfort to know that these platforms are potentially acting ethically if there’s no mechanism to actually challenge and interrogate those services,” she said.
Take facial recognition: One way to understand the quality of facial recognition software — which in some studies, has been shown to have a racial bias — is to examine the data the system was trained on, she suggested.
For example, what is the representation of different ethnic groups in that data set? When was it collected? Why was it collected originally?
Not to mention, who are your customers?
Employees at Microsoft protested this week over the company’s contract with US Immigration and Customs Enforcement (ICE), as controversy raged about the removal of immigrant children from their parents.
Like privacy, “ethical” can be stretched and moulded to mean anything you like. Norms must be debated and chosen. There must be real consequences for failing to follow them.
In the medical industry after all, the Hippocratic Oath to “do no harm” is still broken — but ideally, investigations are then undertaken and licences revoked.
For doctors, failure to live up to a socially agreed upon benchmark comes with penalties, Ms Broad pointed out. By and large, we’re waiting for technology companies to fix themselves.
“It does not prevent harm from occurring, and gives you no mechanism to respond to harm when it does occur.”
Simply being “not Facebook” is the easy option.