6 ms·
And I'm sure they're right. It's important to remember that Facebook can precisely measure the impact of news articles like this. It's tempting to think of Fac
by sillysaurusx 6y ago
And I'm sure they're right. It's important to remember that Facebook can precisely measure the impact of news articles like this.
It's tempting to think of Facebook as omniscient, but I'm not sure how precisely it can be measured. Economically, yes, maybe. But morally?
Remember, the Timnit controversy ended up with letters from congress to Google's CEO. In bigcos, not everyone forecasts everything.
- infogulch 6y agoWhat makes you think that Facebook even considers the moral impact of its actions?
- sillysaurusx 6y agoIt's not so much that they care about the moral impact of their actions, as that society might backlash them if they're not too careful. Large != untouchable. I wish I'd thought of a better word to use than "morally," though, since it didn't quite capture what I meant to say. I meant, people being upset about their actions is hard to quantify.
- infogulch 6y agoAh, I can agree with that. Facebook probably does consider its actions from a moral standpoint if only to calculate the outrage potential.
- edoceo 6y agoOh, rough! A strongly worded letter from Congress. I'll read it after I finish counting all my money.
- andrewljohnson 6y agoWhat effect do you think those letters will have?
- sillysaurusx 6y agoIt's hard to say. In the worst case, it could be a prelude to regulating AI in general. I'm legitimately concerned it might be, if you read between the lines: https://twitter.com/RepYvetteClarke/status/1339297358815834113 https://twitter.com/RepYvetteClarke/status/13392973588158341...
- sosborn 6y agoIt can't even really be measured economically beyond how much it costs to run the ad. They will see the result of the ad, but FB will never know how it would turn out if they didn't run it.
- marcosdumay 6y agoHum... Facebook has some large experiments on mood analysis and manipulation on their knowledge base. If anybody can measure this, it's them. What can still happen is that the campaign backfires so quickly that hey don't have time to react. I don't know their reaction time, but they've been doing mood manipulation for long enough to think it's short.
- wpietri 6y agoIt's true that not everybody forecasts everything, but I think it can be pretty precisely measured via things that they already track, like "how much does this person visit Facebook" and "based on engagement, what are their interests". They were very big on measuring factors related to growth; I'd be very surprised if they weren't measuring what causes people to leave.