Except that people will stop using google if it can’t help them find useful information. Google’s leadership seems to not understand or not care about how the internet works

I think it's highly naive to suggest that Google's leadership is operating in complete ignorance of their own product. This is the company of analytics - whatever approach is being taken is probably backed by a gargantuan amount of data.

That doesn't mean it's good for the web, but "what (the most) users want" is Google's strong point. And this is par for the "the internet is now 4 websites" course, so I don't think it's possible to claim that this isn't "how the internet works" anymore.

If they are operating on top of a mountain of data, then they have a problem (what honestly would be good, even too good to believe).

People are not the memoryless automata that statistics handles best. People react positively to a bad change up to a point where they don't anymore, you don't even need to keep changing to trigger this.

I really hope you're right, but who's to say Google might not be one of the few that manages to correctly figure out how far they can go based on their historically unprecedented amount of data? I wouldn't rely on it, either way.

>This is the company of analytics - whatever approach is being taken is probably backed by a gargantuan amount of data.

Take a look outside your window. Have you understood quantum physics yet? No? I guess all of the available data doesn't help you, because it still requires effort and careful observation to glean insights from it.

The same goes for Google. Data by itself does not guarantee anything. You can look at a piece of data and not reach any conclusion, or reach the wrong conclusion. More data does not mean more insight.

In my field (biology) we are suddenly drowning in data - we are producing new datasets much faster than we can analyze them. Our understanding is not increasing at a commensurate pace because we lack the ability to turn these data into useful knowledge.

I don't think it's "what users want" as much as it's "what makes the most money out of the things that we can convince users to want". And "what can we manipulate our engineers into believing about what they are building so that they won't protest".

The thing about data is it only tells you about the past.

The unfortunate truth is that even as I do my level best to use another search engine, DuckDuckGo most recently, I still find myself disappointed and trying the Google results at least half the time.

I still think Google today isn’t as good as Google 10+ years ago, but I don’t think anything is as good as Google 10+ years ago. They don’t have to be flawless, they just have to be better than the rest, and they are. I settle for DuckDuckGo when I can, but when the results matter, I’ll still end up using Google.

Yeah, I'm on DDG but not too infrequently (particularly with news search) I find myself wondering why I can't find what I'm looking for before remembering to try !g and lo and behold Google is able to pull up what I'm looking for.

That might've been true in the past when the internet was still young and there wasn't as much data around to monopolize. Look at the history of the auto industry to see how the development of a new tech industry plays out long term. To be overly reductive, thousands of companies got distilled to GM + Ford (like Google + Facebook).

Data is the most valuable commodity today. The mountain of data and network effects Google has cultivated ensure they will remain on top for the foreseeable future. It's similar to the crazy expensive industrial supply chains that manufacturers in the past leaned on to maintain their market position. No other company (outside maybe Facebook/Amazon) has that kind of a data moat.

food for thought: there used to be 30-40 american car makers, now there are 3.

3 big ones. Tesla and like a dozen smaller ones.

It depends. Most people, only want the answer to the “question” they typed. And if that is answered in a widget, cloaking the original content, they don’t care. That’s what google realized.

Earlier today I wanted to search for info concerning whether a person had to turn in everything they found in the street. Which one returned any relevant results?

Bing: https://www.bing.com/search?q=do+i+have+to+turn+in+everythin...

Google: https://www.google.com/search?q=do+i+have+to+turn+in+everyth...

Wow... that's the single, most impressive thing I've seen supporting the use of Bing over Google...

So, Google better get the right answer every time. Because when they were only pointing to a larger knowledge base, people understood that they should search a bit for their answer, but if they show it on a Google frame, it better be correct.

Funny thing is that I moved to DDG because Google couldn't reliably guess my answers, and I am satisfied with an engine that tries to guess to a much larger extent. This is because DDG guesses are better, and they don't hide the web results so I can fall back to them.