undefined | Better HN

0 pointscwp12y ago0 comments

So it's our duty as IT professionals to prevent people from using IT?

The services that large-scale IT provides could not exist without gathering the volume of data they do. Google couldn't provide relevant search results with just an index of the web, it also relies on vast amounts of data about user behaviour. The same is true of Facebook, Twitter, LinkedIn etc.

It's non-sensical to talk about risk without talking about the actual scenarios you're worried about. How bad are they? How probable are they? What can we do to reduce the harm and/or likelihood of harm? How much, if at all, will those measures reduce the benefits of the services based on big data? That's the discussion we should be having. Most of the "concern" about big data boils down to an appeal to emotion: "Just think what they could do with all that information!"

0 comments

3 comments · 1 top-level

pfraze12y ago· 2 in thread

> So it's our duty as IT professionals to prevent people from using IT?

Yes, within reason. I advise people not to use Facebook, and I spend a fair amount of working hours on alternatives to Google's app infrastructure. For the grayer areas like improving search, you deal with complicated questions and push the industry to innovate on effective distributed designs. You're making a big assumption saying that collecting a centralized database on users is the only way make good search results. It's certainly one of the easiest ways...

> It's non-sensical to talk about risk without talking about the actual scenarios you're worried about. How bad are they? How probable are they?

If you don't have personal control over your information, detailed knowledge of how the information is used, or the ability to revoke access, you basically have to assume the probability is 100%. There's barely any legal framework in place on the issue, and so we're left with social pressure and business ethics.

Can you really not guess at a negative outcome for personal information collection and passive surveillance?

cwpOP12y ago

Can you really not specify the negative outcomes you're trying to prevent? You're accusing me (and most of the members of this forum) of a moral failing, an abrogation of our responsibility to the world. That's a serious accusation, and I'm taking it seriously.

If you're right, then I should go into the next room and tell my wife not to post any more pictures of our daughter to Facebook. I should tell my mother that she won't be seeing any more pictures of her grand-daughter, and we're going to have to stop having video chats on Skype. Of course, if I do that, they're going to want to know why. What should I tell them? It better be good, because if not, they'll think I'm just being a jerk, and they'll be right.

I agree with you that Facebook has data that could be used in undesirable ways. But it's already being used, right now, on a massive scale, in ways that are clearly and demonstrably desirable. I think the best way to live up to our responsibility is to do two things:

  • clearly enumerate the scenarios we want to avoid, the reasons why, and advocate for public and private policies that prevent them
  • be on the lookout for unintended consequences and negative effects that we didn't anticipate, so that we can prevent them from occurring on a large scale

pfraze12y ago

Ok, that's fair. Here's a list of things that worry me, specifically related to Facebook, but they can apply in other cases.

- Private messages. These can be really dangerous for personal relationships and community standing if leaked.

- Profiling data: friendship graphs, location pings, browsing history, political alignment, sexuality, religion. Can be used to make targeted attacks on individuals by predicting their behaviors; to track an individual's influence in communities; to create credible misinformation about the target or their community; to make strategic decisions that will minimize a person's/community's influence in the outcome (for instance, politically).

- Images. Can be used with automated systems to deanonymize crowds and thus passively surveil.

The issue isn't so much zero-sum. I'm not worried about a sudden crackdown on citizens that's orchestrated through this data, because there are other more relevant safeguards (like cultural health and the functioning political process). I'm more worried about the value of the information as intelligence in political strategy.

Anthony Weiner's embarrassing blow-up is a good extreme example - not that he was targeted (it appeared to have been self-inflicted) but it shows what embarrassing information can do to a politician. Facebook has that in spades, and I'm certain there will be politicians looking to buy.

The more subtle applications are more troubling. Applied well, the dataset could be queried for potential or active dissidents and used to discredit them. It wouldn't be hard, either. Make an anonymous tip to someone's SO that they cheated once, perhaps - if their marriage blows up publicly, that's a success. Think of a pastor or a CEO, in that case.

This sort of thing doesn't happen yet (I hope) but information "wants to be free," which is why you can't let it out once. As I suggested before, there's system breaches, unethical sale of data (perhaps even by employees) government mandates. It's not like any of it is unheard of even before the big data era. And even if most citizens are not a target, by complying with the arrangement, we endanger the people who could protect us by dissenting. The health of our cultures and our republics relies on distributed pockets of inalienable authority - in this case, over private information.

Should you stop using Facebook? I don't know, it's all about trade-offs. I do my best, but I've still got a gmail account and a linode setup, the former because I'm more worried about system breaches with my email, the latter because I can't afford dedicated hosting. I don't like either situation. You can call me a bit hypocritical for that too, but the situation is partially social, so I advocate for change as hard as I can.

Lately I've been thinking the best strategy possible would be to call for Facebook to release self-hosting software that's still compatible with their network. But there's also the possibility of somebody creating a really superior alternative, and that would be nice too. That's what I work on, but it's really hard to beat the featureset.

j / k navigate · click thread line to collapse