Knowing how a mouse's brain works is not necessarily important, as long as you can design a system which performs the same kind of external behaviors (or improved ones if you like).
Deep Blue didn't have to work like Kasparov's brain to be able to beat him at chess. It performed externally better than Kasparov did.
Great point! This is something that really harms Ebay (and other better known companies).
Reminds me of a project idea I had some time ago, similar to price comparison sites in a way:
A 'special' search engine that would take company names and/or product references as the search criteria and display a listing of all the competitors (companies and/or actual products):) Examples: Search: McDonalds; Results: Burger King, KFC, etc... Search: MS Windows NT; Results: Linux, BSD, BeOS, etc...
Such a tool would not rip the actual content (such as prices) from the different company's sites but instead just keep a list of competitors for every company and product in the market. I guess the nature of this tool rules out automatic indexers, which makes it more like a directory. However, existing directories like Yahoo! don't directly map competitors at the product reference level.
This would be great for people who need a product but are not familiar with the extisting market. Many times people go with the big names because they don't even know that there are alternatives...
This reminds me of a lecturer at University who was shocked when I told him that my indexer actually downloaded the HTML from the web servers and stored it locally (even if just in volatile memory):) I guess we must put public domain material into perspective when most people's web browsers keep copies of the pages visited in a cache directory. I guess we're not automatically re-displaying the content to 3rd parties but we're all still keeping copies of it.
I hope you agree that everyone who puts material on a web page and misses out the robots.txt is automatically granting access to conventional search engines to index each and every page. Other webmasters may even manually create links to that person's site with a small description or even a site review, unless there is a clear note on the site saying otherwise obviously. If this is not the case then I think we're taking the public domain case too far.
I agree that Ebay is entitled to sue anyone who's breaching their robots.txt file, but I feel that any site without this protection or an explicit copyright advice will not have a case against content rippers.
This whole think reminds me of a meta-search engine called Electric Monk(?). It ripped the search results of Altavista and other search engines and redisplayed the whole content without even paying credit to these sites. I rembember Altavista 'hitting' them hard... haven't heard anything about them for over 2 years. Anyone one knows what happened?
I agree that spiders should obey the robots.txt file, and that a company should only be able to sue anyone if they breached areas of the site protected using this method.
However I see a clear difference between conventional search engine spiders (such as the ones used by Altavista or Lycos) and other spiders used by price comparison sites. Altavista displays listings with the title, URL and brief description of each page, it doesn't rip the contents of the page apart (such as product prices) and reformats the ripped information from different sites. This is surely the main difference between the two indexing methods, the former probably welcomed by e-commerce sites and the other worth the trouble of going to the courts.
This is why I think the robots.txt standard should be revised and allow for the same clear distinction between spiders:
pure indexers, and
content rippers.
This means an e-commerce site can allow a search engine to index most of the pages (this is not only useful to the company selling the product, but also to the consumers who will be exposed to a larger range of choices) and at the same time prevent those unwelcome content rippers.
As a consumer I obviously welcome the idea of price comparison tools which do most of the dirty work, but we must understand the position of the companies. On the other hand, as an internet user (and hardcore web searcher) the last thing I want is for all the companies to ban access to their sites through the robots.txt file, preventing normal search engines from doing their work.
Here's a company who's doing a great job at doing it all wrong.
I don't agree with the 'don't have to prevent hacking; they have to manage their risks.' bit. If cracking was prevented they wouldn't have to spend so much money 'managing their risks'. And one of the best ways to keep crackers away is to make sure they don't know about you. This is something Lloyds of London is not achieving with this kind of news coming up in Sci-Tech sites... I bet they just got all kinds of crackers lining up to 'test' their new insurance...
It's understandable why projects like Honda's P3 get some much media attention, since they're meeting most people's idea of an artificial intelligent robot, even if in reality its more of an achievement in robotics, and not some much in 'pure' AI. Personally I must admit that research in androids produces very cool prototypes (see Android World) but it isn't necessarily the approach which takes AI research further. I agree that the way forward is to give AI ways of interacting with the world, and therefore robotics play and essential role, but I can't help feeling that the android approach diverts researcher's attention from the main problem, the 'Intelligence'.
I guess there is always a reason to invest in androids as far as interaction with humans is necessary, but I always saw intelligent robotics being most useful where humans are not needed, i.e. space/deep sea exploration, miscellaneous operations in hostile environments, and as controversial as it may be, warfare. All these areas require autonomous machines as substitutes to humans, and I'm not surprised if we see some of the most exciting advances in AI coming from the Department of Defence, not Honda.
Slashdot Mirror
Knowing how a mouse's brain works is not necessarily important, as long as you can design a system which performs the same kind of external behaviors (or improved ones if you like).
Deep Blue didn't have to work like Kasparov's brain to be able to beat him at chess. It performed externally better than Kasparov did.
Reminds me of a project idea I had some time ago, similar to price comparison sites in a way:
A 'special' search engine that would take company names and/or product references as the search criteria and display a listing of all the competitors (companies and/or actual products) :)
Examples:
Search: McDonalds; Results: Burger King, KFC, etc...
Search: MS Windows NT; Results: Linux, BSD, BeOS, etc...
Such a tool would not rip the actual content (such as prices) from the different company's sites but instead just keep a list of competitors for every company and product in the market.
I guess the nature of this tool rules out automatic indexers, which makes it more like a directory. However, existing directories like Yahoo! don't directly map competitors at the product reference level.
This would be great for people who need a product but are not familiar with the extisting market. Many times people go with the big names because they don't even know that there are alternatives...
I guess we must put public domain material into perspective when most people's web browsers keep copies of the pages visited in a cache directory. I guess we're not automatically re-displaying the content to 3rd parties but we're all still keeping copies of it.
I hope you agree that everyone who puts material on a web page and misses out the robots.txt is automatically granting access to conventional search engines to index each and every page. Other webmasters may even manually create links to that person's site with a small description or even a site review, unless there is a clear note on the site saying otherwise obviously. If this is not the case then I think we're taking the public domain case too far.
I agree that Ebay is entitled to sue anyone who's breaching their robots.txt file, but I feel that any site without this protection or an explicit copyright advice will not have a case against content rippers.
This whole think reminds me of a meta-search engine called Electric Monk(?). It ripped the search results of Altavista and other search engines and redisplayed the whole content without even paying credit to these sites. I rembember Altavista 'hitting' them hard... haven't heard anything about them for over 2 years. Anyone one knows what happened?
Well, I guess this makes my post redundant... ;)
Next time I'll refresh the page before submitting my 'TookOneHourToWrite' comments
However I see a clear difference between conventional search engine spiders (such as the ones used by Altavista or Lycos) and other spiders used by price comparison sites. Altavista displays listings with the title, URL and brief description of each page, it doesn't rip the contents of the page apart (such as product prices) and reformats the ripped information from different sites. This is surely the main difference between the two indexing methods, the former probably welcomed by e-commerce sites and the other worth the trouble of going to the courts.
This is why I think the robots.txt standard should be revised and allow for the same clear distinction between spiders:
pure indexers, and
content rippers.
This means an e-commerce site can allow a search engine to index most of the pages (this is not only useful to the company selling the product, but also to the consumers who will be exposed to a larger range of choices) and at the same time prevent those unwelcome content rippers.
As a consumer I obviously welcome the idea of price comparison tools which do most of the dirty work, but we must understand the position of the companies. On the other hand, as an internet user (and hardcore web searcher) the last thing I want is for all the companies to ban access to their sites through the robots.txt file, preventing normal search engines from doing their work.
Here's a company who's doing a great job at doing it all wrong.
I don't agree with the 'don't have to prevent hacking; they have to manage their risks.' bit.
If cracking was prevented they wouldn't have to spend so much money 'managing their risks'.
And one of the best ways to keep crackers away is to make sure they don't know about you. This is something Lloyds of London is not achieving with this kind of news coming up in Sci-Tech sites...
I bet they just got all kinds of crackers lining up to 'test' their new insurance...
Personally I must admit that research in androids produces very cool prototypes (see Android World) but it isn't necessarily the approach which takes AI research further. I agree that the way forward is to give AI ways of interacting with the world, and therefore robotics play and essential role, but I can't help feeling that the android approach diverts researcher's attention from the main problem, the 'Intelligence'.
I guess there is always a reason to invest in androids as far as interaction with humans is necessary, but I always saw intelligent robotics being most useful where humans are not needed, i.e. space/deep sea exploration, miscellaneous operations in hostile environments, and as controversial as it may be, warfare.
All these areas require autonomous machines as substitutes to humans, and I'm not surprised if we see some of the most exciting advances in AI coming from the Department of Defence, not Honda.