> improving search relevance by x% is tremendously socially valuable, but probably makes Google's bottom line go up by a thousandth of x, and they have a very direct understanding of the connection.
I'm not sure this is correct. In a vacuum of real competition, the most profitable ought to be how it is right now, when search results are kind of ambiguously bad, so you need to click and skim through a few results to maybe find what you want, multiplying the number of ad impressions.
We can talk about the Importance Of This until we die of old age. Google failed us, Google is Borg, Yandex is FSB, Bing is ... Bing, etc. However, the fact that there is a problem to be solved and It Is Important doesn't mean that the EU will solve it. If anything, it's just the set-up for another political play that will have damaging consequences to the internet as a whole, just like GDPR, EU's poster-child "internet project".
They made GDPR just strong enough legislatively to be annoying, but not strong enough to actually change anything. Companies can still store EU citizens' data anywhere they want and do whatever they want. There's no insight into this. It is an unenforceable law and the only artifact of it existing is that I have to have "I don't care about cookies" installed, so that Avast Antivirus can eventually decide to silently collect and sell on my data.
For all intents and purposes OpenWebSearch is most likely not meant to succeed at anything either, and is just going to be a political stepping stone towards legislature that will be awkward and make the internet worse for all users. EU has a long history of creating or hanging onto laws that betray a misunderstanding of how the digital age works, how the internet works, how data can be copied or moved around for free.
Here's an example. Every country in Europe has some sort of legal construct in place that will prevent you from secretly recording a conversation you're having and uploading it to the internet. So for example, take Austria. They accomplish it by preventing you from publishing it on the internet. However, they don't prevent you from recording it. There are laws against secretly recording a conversation you are not a part of, but there's no such law for the situation when you are part of the conversation. So you can record and upload, just not publish. However, you can get on a train and go to a different country which has laws that prevent you from recording, but has no laws against publishing. Then you're in the clear. Or you can just use a VPN so that it looks like you uploaded and published it from the other country. Or you can just upload to YouTube, which will not show where the thing was published from - and claim that you did so during your tourist visit to Vietnam or whatever. And if someone brings a civil lawsuit? Good luck trawling the Vietnamese legal system for clues about that specific issue. Hope you know this rare language with virtually no legal experts who at the same time speak Vietnamese and your particular local European language. The costs would be on the order of tens of thousands at least, which is out of reach for anyone but the wealthiest EU citizens.
Want more? Egregious copyright related laws are known by everyone.
So are dns blocks of shunned sites. Or the recent Austrian project to block cloudflare IPs which literally broke the internet.
The fact such laws are still in place from before the internet - or are even still being put out - and are effectively unenforceable while making the internet worse for everyone - makes it clear that the governments in place simply have a misunderstanding of how the internet works.
None of that will stop politicians from coming up with BS excuses breaking the world with fever-dream laws in order to push their latest political agenda. The real question is: if such clearly unfit legislature is being put in place for political folly in industries we understand - how much of that is happening in industries we don't understand? Health Care, Food, Agriculture, Civil Engineering, Patent Law, and so on. My guess? You can tell what my guess is.
As for Open Web Search, what everyone should really be asking themselves is: ok, so what's the scam that's going to be pulled here?
As I'm one of the people who is working within the Open Web Search project please allow me to feel strongly about your statement. I've been involved in campaigning for this project since around 2014. This project did not originate as a 4d-chess move of some political game. It exists because of the hard work of a group of people, some of which are researchers, some of which are working at smaller search engines and some of which are involved in civil rights organisations. Currently still sitting in the kick-off meeting I can tell you that we are actively discussing how to get this project to produce useful results for a european open web index. Getting the EU to draft problematic legislation is neither in our power nor in our interest.
Google's priority with ads is to make sure ad buyers are happy with the results they are getting. This means making sure the people who click the ads go on to buy the product being advertised. That's what ad buyers measure.
That's a major fallacy you're opening with. If something is important to us (as a society), we will find ways to make it serve us well even when it's under attack. The rest of your post even makes a similar point: the social value of well-working search is greater than the economic value for even the biggest search monopolist on the planet. So why not socialize it?
Being idealistic about it won't change the outcome.
As far as I know, this hasn't really been tried, definitely not with the kinds of resources this project will have available.
PageRank was initially much less prone to SEO shenanigans because it relied on signals from other sites (incoming links) to decide how important a result was. Of course, as Google became more popular, people started sharing links on other pages and so on to cheat the PageRank algorithm. And Google have been caught in a fight with SEO ever since.
AFAIK, nobody has ever tried that because of the dangers of vote buying and coercion. Essentially, gaming the system.
(I don't think this proves anything! Simply wanted to suggest that comparing search to democracy doesn't significantly change the analysis wrt opacity.)
The correct strategy isn’t to hand this over to a giant bureaucracy, but to create an atmosphere where we can have a dozen alternatives to Google.
The only search product that comes close to this is Amazon search or booking.com (or maybe YouTube).
All multi billion dollar businesses. And when smaller ones emerge like the flight scanner one, google buys them.
I don’t see google going anywhere soon. They have a lot of faults but they’re still too good.
What is so great about those search examples?
However, there's no way Geizhals will ever expand into general search. The reason why they are so much better than the competition is that they are focussing on a small, profitable niche and presumably use manual data entry to ensure they have the best data.
[1]: there's a UK version too "skinflint.co.uk"
I don't think I've seen a website that does that before and I think it's worth praising.
But at the end of the day the resulting dataset would be sold to third parties so they can rank the results appropriately. Which to me seems to be the only sane way forward. Only a government could run something the scale of the Google crawler and succeed at doing so. And then everyone can build search engines on top of that.
That bit troubles me. If the index is maintained by a government agency, and every search engine is using the same index, then that's a massive censorship avenue. I wonder how "open" Open Web Index is going to be.
The vast majority of the net is, after all, now indexed, so you can run your own indexer to cover whatever they didn't cover.