Do computers need to forget?

[bascule]

http://arstechnica.com/news.ars/post...to-forget.html

Mayer-Schönberger suggests that we are creating a Benthamist panopticon by archiving so many bits of knowledge for so long. The accumulated weight of stored Google searches, thousands of family photographs, millions of books, credit bureau information, air travel reservations, massive government databases, archived e-mail, etc., can actually be a detriment to speech and action, he argues.

I disagree. I think panopticon is inevitable and we just have to learn to like it. Google probably knows more about me than I know about myself at this point.

What do you think? Are you scared of the continuous archiving of everything you do online?

[Deviant Ollam]

Re: Do computers need to forget?

Are you scared of the continuous archiving of everything you do online?

i'm slightly worried about the potential for such an archive, but i still feel that most civilian applications where logs are concerned can be overcome or at least counter-acted with Tor and other similar technologies. Maybe i'm just naive, but i don't think covert channel data is being hoarded or packet latency timing is being calculated or any other similar things are being done at this present time in order to tie my to each and every one of my many google searches which take place from loads of different machines, IPs, and so forth.

[bascule]

Re: Do computers need to forget?

What about just your search terms?

[Floydr47]

Re: Do computers need to forget?

Archived knowledge is the gravel of the information highway. Let's face it, the governments of the world would know just as much about us without all the archives. It would just take government agencies longer to organize the data that they have on us. If someone is using search terms that could possibly prove harmful to others then by all means I think that search terms should be linked to IPs. If a person is researching the best way to grow roses, then what is the harm in linking that search to his/her IP? On the other hand if someone is researching how to build a nuke in their basement, well, that search may bear scrutiny. Just my thoughts....

[hackajar]

Re: Do computers need to forget?

Perhaps if more people that have to live up to previous statements, the more 1.) You have to be careful of what they say in general 2.) Some odd ball statements will seem less important.

People love to lache onto obscure irrelivent statements to try and defame a person. If EVERYONE's obscure irrelivent statemenst were public, then maybe we could talk about current issues and not what someone said in passing 20 years ago.

[bascule]

Re: Do computers need to forget?

i'm slightly worried about the potential for such an archive, but i still feel that most civilian applications where logs are concerned can be overcome or at least counter-acted with Tor and other similar technologies. Maybe i'm just naive, but i don't think covert channel data is being hoarded or packet latency timing is being calculated or any other similar things are being done at this present time in order to tie my to each and every one of my many google searches which take place from loads of different machines, IPs, and so forth.

Microsoft is developing software to do just that:

http://www.newscientisttech.com/arti...ng-human_rss20

You can probably bet Google is doing it as well

[theprez98]

Re: Do computers need to forget?

I have multiple computers at home for multiple users. But to Google or anyone else "outside", I have one IP address. Outside of actually creating individual accounts that save searches/preferences (iGoogle), how much can Google (or anyone else for that matter) really find out about me? It seems to me that my preferences (and those of my family) are largely obscured by each other into a rather undecipherable amalgam.

For all the interest it received, LikeBetter faired no better than 50% in guessing very generalized information about me based on my photo preferences.

[bascule]

Re: Do computers need to forget?

Google has unique tracking cookies per-computer. Beyond that, your computers probably have different User-Agent strings.

And beyond that, the software I linked above doesn't need an IP address to identify you uniquely. It can classify you by your behavior patterns.

[theprez98]

Re: Do computers need to forget?

Google has unique tracking cookies per-computer. Beyond that, your computers probably have different User-Agent strings.

And beyond that, the software I linked above doesn't need an IP address to identify you uniquely. It can classify you by your behavior patterns.

Do you honestly believe any software developed by Microsoft will work as well as they advertise it?

On a serious note, how about several people (a family perhaps) using the same computer and the same logon. Google would have a difficult time differentiating the users, would they not?

I'm not suggesting that Google and/or Microsoft and/or whomever can't gather a ton of information on any particular person, but I believe it is a lot more difficult that these articles suggest. There is a lot more noise they'd have to filter out to get things more accurate.

Furthermore, I think we are a longs ways off before software can accurately and uniquely classify me by my patterns assuming that the software does not know my IP address.

[Shinobi]

Re: Do computers need to forget?

Hopefully google won't take into account my spelling mistakes. I'd end up with some strange recommendations.

[bascule]

Re: Do computers need to forget?

Do you honestly believe any software developed by Microsoft will work as well as they advertise it?

Sure, just give it a couple of patches after they release SP1, heh

On a serious note, how about several people (a family perhaps) using the same computer and the same logon. Google would have a difficult time differentiating the users, would they not?

Given Google has the best minds in the field, I'm sure if they wanted to they could make it happen. As to whether or not they're doing it, I can't really say.

I'm not suggesting that Google and/or Microsoft and/or whomever can't gather a ton of information on any particular person, but I believe it is a lot more difficult that these articles suggest. There is a lot more noise they'd have to filter out to get things more accurate.

I'm sure there'd be a lot of indeterminate cases, or false positives. I don't think the process would be perfect. But if you know that only one person is using a given IP/cookie combination at a time, it becomes a whole lot easier.

Furthermore, I think we are a longs ways off before software can accurately and uniquely classify me by my patterns assuming that the software does not know my IP address.

Yeah, I agree that if you were just to hand a program a ton of search queries which, say, only had timestamps but no other uniquely identifying information, it would make things a lot harder.

That's why the case of the family computer is different: you're not really getting a mixed batch of data. You're trying to pick from a small number of people who can only use a computer one at a time and probably make several searches in a row. That makes the task considerably easier.

[philomath]

Re: Do computers need to forget?

Just a thought. what if someone were to write a web crawler that hit random sites at all times of the day? I would wonder if this would eat up mass amount of space and be hard to sort? just wondering.

[theprez98]

Re: Do computers need to forget?

Just a thought. what if someone were to write a web crawler that hit random sites at all times of the day? I would wonder if this would eat up mass amount of space and be hard to sort? just wondering.

There's a firefox plugin that does "random" searches at regular intervals which is designed to muddy the waters. I don't remember what it's called though.

[TaGmAn]

Re: Do computers need to forget?

In response to the the first post, an archive of everything we do (and eat) isn't a bad thing. It makes me think of Star Trek and how they would constantly ask the computer for Historical Records, and would then use these records to get out of a jam. When used appropriately such an archive could be a powerful tool.

[Deviant Ollam]

Re: Do computers need to forget?

an archive of everything we do (and eat) isn't a bad thing.

B to the U to the Double-L to the SHIT.

sorry to be so blunt, but i'm of the opinion that long, comprehensive records of ANYTHING that a person does are generally a Bad Thing™ to have maintained. "The premise that privacy is about hiding a wrong [is fundamentally incorrect]. Privacy is an inherent human right, and a requirement for maintaining the human condition with dignity and respect." -- Bruce Schneier

in this modern age, records themselves are an anathema to privacy since they are easier to compromise, very easy to disseminate if compromised, and impossible to reclaim/delete once they leak out.