Hacker News Reader

Comment by _m8fo | Hacker News Reader

I'm curious to the opinions of people here on companies collecting data to build data sets vs. privacy.

Regardless of our individual opinions, probably far more than enough of the general public does not care about or know about securing their online privacy than is needed for an invasive AI startup to get what they need to make a valuable product.

Do you, or anyone else, have any thoughts as to why Americans have, at least stereotypically, a more laissez faire attitude to personal privacy compared to those in Europe? Just the somewhat recent history of secret police in Europe? The American founders were very pro-privacy weren't they?

I couldn't say. I don't know enough about European data sharing habits to really draw a comparison. I would remind you though as Apple proved during the San Bernadino iPhone showdown, the U.S. government is heavily restrained by law from just storming in and taking company data that they claim is necessary for the interests of national security.

Personal privacy is a paper tiger, as Facebook and others have proven again and again. You only need to look at the consequence-free landscape of privacy breaches to understand that people have no meaningful right to avoid direct marketing and spillage of their contact info.

On the other hand, confidential information of corporations is pretty well guarded, and confidential contract terms, costs, pricing, and other information that doesn't get shared among peer companies or competitors is what will propel AI into the economic stratosphere. Finding ways to get confidential data into learning systems and provide actionable feedback is the killer app.

I am not sure I get it. If I am Ford and get access to GMs Salary database or factory electrical meter, do I win much? It's almost certain I know the ballpark frommrunning my own factory and probably have hired three of GMs managers with detailed knowledge in their heads last week.

Or is it more hedge funds - like satellite images of Walmart car parks to estimate the revenue figures?

Either way these don't seem like things we need AI to pick out patterns ?

Or am I missing something?

It's about normalizing pricing and terms in the supply chain. (You chose bad examples, as the larger buyers implicate antitrust/competition rules and illegal collusion.)

If an AI has access to a broad spectrum of confidential information, it can reliably answer questions about the state of the market on an anonymized basis. This has the effect of normalizing sourcing behavior, which reduces purchasing and sales friction and improves overall efficiencies.

Well, why does it have to involve privacy? A camera on a marsh with thousands of hours of footage could be useful data (has bird, doesn't have bird? Water level? Cloud cover?)

A tollgate recording vehicle pass-through with a timestamp, a train station gate recording individual pass-through with a timestamp, etc, with huge volume could be useful data.

Indeed, it is very frightening what the companies can do with user's data. While in U.S, the data of a user may reside in the cloud of multiple companies and needs to be aggregated to be useful, in China, companies like Tencent knows about almost everything about every user: your IM messages, your purchases, the movies/music you like, the news you read, your geolocation. I cannot imagine what they are capable of with all those data. Think black mirror.

Privacy activists have long ignored any benefits of data collection and as we continue to extract more and more value from data this should become more evident and we will be forced to start discussing concrete harms rather than people's general discomfort.

Aside from running large scale analyses over large health data sets, what are some examples where the value derived from large aggregations of personal data is dispersed widely through a society rather than being captured mostly by a single corporation or organization?

Building large data sets doesn't necessarily mean from personal data. Look at open-data initiatives such as http://open.canada.ca/en/open-data . Lots of potential for useful tools to be created if the data is there, which won't happen if even benign data like that are kept under wraps/not collected.

Of course, this is not true. Plenty of harms have been shown, particularly based on past history. But these are dismissed as things that couldn't possibly happen again. Which means that the objections will only be accepted when it's too late.

Marge: Do I have to be dead before you’ll help me?

Wiggum: Well, not dead – dying. [Marge gets up to leave] No, no, no, no. Don’t walk away. How about this: just show me the knife in your back. Not too deep, but it should be able to stand by itself.

Reply on news.ycombinator.com