[soxpopuli]

What is a Search Engine? Is this some kind of legally regulated category that says you a) must only return Web results and b) in the form of 10 blue links?

Should a search which a search engine knows the direct answer to (word definitions, the current weather or time, etc) send you to another site for the answer. Why are "search engines" prohibited from direct answers, but voice agents like Siri allowed? Is there something fundamentally different between a text box and a voice input?

Likewise, if a search engine has indexed data and can return deep links to other sites formatted differently, why is that different? For example, if you search for "Playstation 4" and Google simply returned the first 10 hits (Amazon.com, eBay, Walmart, etc) as a page of 10 blue links like it did in 2006, instead of formatting them in a nice box at the top of the screen with summary price extracted, would this still be illegal? Why is it legal to display organic search results as blue links, but if you display them in a box and call it "Product Search", it's suddenly illegal? This makes no sense to me. The only difference between the Google Product Search box at the top, and displaying the links is simply better visual presentation.

The world has moved on from ten blue links. Mobile devices have even more constrained real estate and network latency pushing the need for summarization and smart presentation even further. A new class of consumer expects these devices to almost act like intelligent agents when answering queries.

Is the European Commission saying it will be illegal to build JARVIS or the Star Trek computer, because a smarter search that doesn't delegate to other niche search engines, and instead returns direct answers, is unfair competition?

At the heart of this seems to be the idea that Google search should return links to other shopping comparison engines instead of direct links to Amazon, et al. That frankly seems like a good way to hurt customer experience. If you have a good product comparison engine these days, you're probably going to end up as a native app anyway.

By the time this EU case winds down (Microsoft's took a decade), the traditional web search engine might not even exist anymore.

[pjc50]

There is the rather delicate issue of copyright here. If you google for something and the answer is on another page, lifting it from that page is (a) making a copy and (b) potentially depriving the linked page of revenue.

There is also the question of using the threat to ban people from your search results (which is calamitous for most businesses) to resolve your disputes with them.

[soxpopuli]

But the linked page is there. Google is indexing Amazon, extracting the photo, description, and price of the product, displaying it in a box, and making it a link back to Amazon where it got the data from, what's the problem?

Product Search is just another form of summarization/snippeting that just presents the data in more digestable format.

Remember Google Fusion Tables? That was an attempt to extract facts from pages and put them into tables, so if you ask "What's the masses of the planets of the solar system", you could get a table of 8 planets and masses, with the results coming from 8 different sites. But the links could still be there to the original site, it's just formatted as a table instead of as 8 blue links with summary paragraphs, which is harder for humans to process.

Where do we draw the line? You've seen how Google has a new system that can automatically caption images with deep neural networks. (http://techcrunch.com/2014/11/18/new-google-research-project...)

Now what if this same system eventually allows the search engine to summarize your web page by 'reading it', and then auto-generating a paragraph that explains what it thought it was about?

There'd be no actual direct copying of text (like there is with search snippets), instead it would be more like a human going to a library, reading a book, and writing a review

Would this also violate copyright?

[fixermark]

That question is moot in the shopping context; the products that show up in that system are fed directly to Google by the owners of the product listings.

https://support.google.com/merchants/answer/188478?hl=en

[tajen]

> deep links, formatted differently

Strictly speaking, you are only allowed to connect to a remote computer if you agree with their Terms of Service. I guess there are legitimate round corners, like you need to fetch the ToS first. It's not applicable as-is for the WWW, but you need to accept the websites have a say in how their results should be displayed.

Btw, is Robots.txt legally enforceable? http://www.robotstxt.org/faq/legal.html

[crazy1van]

> By the time this EU case winds down (Microsoft's took a decade), the traditional web search engine might not even exist anymore.

Exactly. The court system moves so much slower than the technology industry. It wasn't the US justice department that disrupted IE. Instead it was Firefox, Safari, and Chrome. It took me 5 minutes to install a different, better web browser. It will take me even less time to start using a different, better search engine whenever one comes along.

[ocdtrekkie]

"Building the Star Trek computer" was always a funny marketing angle for Google, given that the Federation would frown on a greedy corporation controlling the flow of information.