The brief history of Facebook apologizing
boingboing.net
boingboing.net
1. A/B testing is done all the time and has undefined emotional effects on users other than to maximize addiction & ad revenue. Google does this constantly and is worshipped for it's ingenuity.
2. The whole point is that they didn't know whether there would be emotional contagion, which is why they conducted the experiment. And yet similar experiments are happening all the time, so the experiment didn't add to the emotional burden of users - it just uncovered it.
[edit: I'd go further and suggest that maybe it would be unethical for Facebook not to do experiments like this. They need to understand this kind of effect in order to make responsible design decisions.]
People don't trust Google. They've ruined their brand by abusing their users' trust, and Facebook is on the same path.
As for what's "reasonable". Apple is operating a large messaging network as well - iMessage. And they also published a paper recently.
However, their paper was about how their architecture prevents anyone, not just third parties, but in most cases Apple's own employees from reading and manipulating user data. No one at Apple can read a message you send between two iOS devices. Apple aren't saints, but I like their idea of reasonable way more than what you're trying to sell here.
I think you're right about Google's brand, but I think they can simultaneously be distrusted and worshipped at the same time.
I also think that Apple is showing that compromising privacy is a consequence of Google and Facebook's business models, and not some fundamental rule of the web.
But, I don't think any of this changes my view that this particular study is not unethical in the way people are claiming given that Facebook exists and behaves as it does. If you think that Facebook operates an unethical business, I'd still argue that this doesn't make it worse, and that becoming less ignorant about how they affect people is a good thing.
This particular apology is even worse because they're not apologising for doing the research; they're apologising "for the way the paper described our research and any anxiety it caused."
They're not saying "We will never do this again."
They're not saying "We will never do this again without first obtaining ethical approval from an appropriate review board."
They're not saying "We will never change your Facebook newsfeed to affect your emotional state in order to make you more susceptible to adverts for particular products."
They're not saying "We will never accept payment from an advertiser to change your Facebook feed in order to affect your emotional state."
So, what conclusions should we draw here?
It's a free service that you agree to use. You should read the fine-print if you care about anything like this.
If it can be proven that experiments or anything else worth apologizing for is detrimental in some way, then something needs to be said. But otherwise, I think they're free to do with they want with their data.