Live data from Hacker News

Why one search engine for the whole internet is not a good idea anymore

dynamicguy.com

11–20 of 28 posts

Re: Why one search engine for the whole internet is not a good idea anymore

#11
post #5

Personally, when I know that a certain topic is of high value to advertisers, I don't even bother googling anymore, and try to find someone qualified instead. For example, I recently was trying to look up theory behind working out, how differenr muscles work, how nutrition works and all that - but there're so many people who're trying to sell their courses, apps, books and stuff, that I have no idea how to go around…

That's a problem now only AI could solve. All other attempts at non-AI algorithms are futile. We need to be able to search through a curated internet that looks somewhat like a mix of Wikipedia and the old Yahoo directory and has to be tailored to each individual user search goals (I unlike you may be interested in commercial workout products). AI is the only "curator" capable of ranking and selecting results with such abstract filters.

Re: Why one search engine for the whole internet is not a good idea anymore

#12
One search engine per country seems like the wrong way to slice the problem up to me.

Locality is missing from many queries because the data is generally not relevant or is relevant but not available.

I would propose that instead of breaking it per country, breaking the search based on what sort of content you are looking for.

Programming data or food recipes have nothing to do with locality but finding the nearest place to get chocolate covered coffee beans is.

I know this is going to seem like total self-promotion but this article caught my eye and thought that it would be valid to say I have been thinking about this problem a while, which is why I built https://www.closient.com/ - mind you I am still building it but it does work.

Maybe I am not seeing the same problem as the author is but I do see locality as being a critical part of merging the logical/virtual world with the physical.

Re: Why one search engine for the whole internet is not a good idea anymore

#13

I work on Google Search. Having one search engine per country doesn't seem like the correct approach to the problem. > Every country has a mammoth collection of valid results for your query. Having seen the corpus of content available from each language, this is categorically false. Consider Wikipedia, which is a fairly ubiquitous information source on the web that provides answers to tons of searches. English docume…

Agree and, as I work for Seznam, especially agree on the ability to train models properly to satisfy local users. Not to mention that the user can choose what should be the language of the results.

I guess I wasn't too far off geographically with Romanian Wikipedia :)

Keep up the great work at https://www.seznam.cz/ ! Lots of interesting search features, maps, etc.

Re: Why one search engine for the whole internet is not a good idea anymore

#14
post #5

Personally, when I know that a certain topic is of high value to advertisers, I don't even bother googling anymore, and try to find someone qualified instead. For example, I recently was trying to look up theory behind working out, how differenr muscles work, how nutrition works and all that - but there're so many people who're trying to sell their courses, apps, books and stuff, that I have no idea how to go around…

Thankfully we still have journals and conference papers. Look for the ones with high impact factor and go from there. That's what I did when I was looking into creatine supplementation.

You can't even ask experts for their opinion since they are the ones selling the stuff!

Re: Why one search engine for the whole internet is not a good idea anymore

#15

One search engine per country seems like the wrong way to slice the problem up to me. Locality is missing from many queries because the data is generally not relevant or is relevant but not available. I would propose that instead of breaking it per country, breaking the search based on what sort of content you are looking for. Programming data or food recipes have nothing to do with locality but finding the nearest p…

How does it work? Searching for book store (https://www.closient.com/?q=book_store the underscore was autocompleted) gave me results that were 1200+ km away from my current location. Other queries did the same.

Re: Why one search engine for the whole internet is not a good idea anymore

#16
post #5

Personally, when I know that a certain topic is of high value to advertisers, I don't even bother googling anymore, and try to find someone qualified instead. For example, I recently was trying to look up theory behind working out, how differenr muscles work, how nutrition works and all that - but there're so many people who're trying to sell their courses, apps, books and stuff, that I have no idea how to go around…

That's a problem now only AI could solve. All other attempts at non-AI algorithms are futile. We need to be able to search through a curated internet that looks somewhat like a mix of Wikipedia and the old Yahoo directory and has to be tailored to each individual user search goals (I unlike you may be interested in commercial workout products). AI is the only "curator" capable of ranking and selecting results with su…

Just like Facebook's AI is solving the feed problem? Do you trust the proprietary machine learning black box, that even the creators don't completely understand, ran by a for-profit corporation to decide what information you consume?

It's not like I have a better suggestion, but AI is definitely not a silver bullet for this.

Re: Why one search engine for the whole internet is not a good idea anymore

#17
post #5

Personally, when I know that a certain topic is of high value to advertisers, I don't even bother googling anymore, and try to find someone qualified instead. For example, I recently was trying to look up theory behind working out, how differenr muscles work, how nutrition works and all that - but there're so many people who're trying to sell their courses, apps, books and stuff, that I have no idea how to go around…

Thankfully we still have journals and conference papers. Look for the ones with high impact factor and go from there. That's what I did when I was looking into creatine supplementation. You can't even ask experts for their opinion since they are the ones selling the stuff!

Good advice - but in my experience, reading the latest research in any field requires a lot of more basic knowledge first.

Re: Why one search engine for the whole internet is not a good idea anymore

#18
I couldn't agree more - one search engine is a terrible thing. But for other reasons. It's amazing to see how in other areas people complain about monopolies but when it comes to our window to the digital world, most of us seem to be fine with it. One US monopoly that dictates what should be important for us, filters what we see based on secret algorithms, forces us to "optimize" our websites in this and that way along their "standards". Google has good technology but we should wake up and see they are not the altruists they like to tell everyone.

Re: Why one search engine for the whole internet is not a good idea anymore

#19
Of course Google has its short comings. But I don't see "a search engine per country" is going to change anything. First of all, why limit the boundary to countries? There could be vastly different areas in the same country or two countries with similar culture and languages. Second not every country can afford / has enough resource for setting up their own Google. Search engine requires expertise that builds up through many years of experience. Instead of reinventing the wheel for every country, why not work on better localisation of existing search engines?

Re: Why one search engine for the whole internet is not a good idea anymore

#20
The author says a lot of correct things. When information belongs to one organization, then there is no future. Do you really think that we will receive information only from 1 organization? Monopoly on information, like any other monopoly will not be eternal.

Today the information growing rapidly and they try to place everything on first page - it is impossible to display all the relevant information on 1 page, it is very restrictive.

And I'm not talking about the fact that on this first page there are a lot of advertising links and artificially advanced sites.

Model Yahoo has a future. And we are developing it: we are developing the Bubblehunt project, where each user can create his own search engine. This is decentralization, where each user can act as an independent information provider, like miniGoogle.

A bit of advertising, sorry: https://bubblehunt.com - tell me, what do we need to improve?

We strive to enable each user to influence the quality of Internet search and make it more social, transparent and dynamic.

People are increasingly talking about the need to change something in search.

Google is an excellent search engine, but it hides from us a huge amount of information. Every day born a lot of interesting resources and we get information only from sites that are on the 1 page of search. They become only more popular. And you need to pay money so people see your website!

Is this correct in your opinion? In my opinion, Google will make a huge profit here and this is not very correct. People just do not have alternatives.

Post reply on HN