Has Google actually invented anything?

[leahcim]

@BOOM

Well, exactly. You write it. Maybe like you write a book or play - is Shakespeare an inventor?

Not quite a perfect analogy for software inventions, but yes.

I’ll certainly grant that most code is mundane, doing what was done before with little to know refinement. If I implement a quicksort to sort a list as part of my program, I am not inventing. But when Tony Hoare first wrote a quicksort algorithm, he was inventing – He saw a problem, sorting of a list, and devised a way to do it better.

[Chronos]

For that matter, though, if you’re using quicksort in a program you’re writing, you’re more likely to just grab a function from a standard library where someone else has already implemented it.

[BOOM]

@ultrafilter

If someone comes up with a way to identify pairs that are not similar by looking at each document once, how is that not an invention?

@leahcim

If I implement a quicksort to sort a list as part of my program, I am not inventing. But when Tony Hoare first wrote a quicksort algorithm, he was inventing – He saw a problem, sorting of a list, and devised a way to do it better.

It’s really tempting to accept these suggestion but, at the same time, isn’t the invention here was the concept of ‘algorithm’, everything subsequent being an method application of that invention …

I guess we’re now talking in circles … maybe there’s a fine line.

[Chronos]

If you’re going to argue that, then wasn’t the original invention that of “machine”, and everything else is just a method application of that?

[leahcim]

@BOOM

…isn’t the invention here was the concept of ‘algorithm’, everything subsequent being an method application of that invention …

That certainly would grossly underestimate the breadth and depth of the field of algorithm design by any conventional standard.

I guess we’re now talking in circles … maybe there’s a fine line.

There certainly is a fine line, and many patent examiners and lawyers make their living based on arguments about whether something constitutes a new invention or just an obvious or trivial extension of an existing invention.

But the location of that fine line is quite far away from “all algorithms are the same invention”.

[Yookeroo]

@clairobscur

I think that’s very exagerated.

I don’t think so.

@KarlGauss

But in larger part, I think it was the cutesy, easy-to-remember name (a name that became a verb and a meme very quickly).

No. Not at all. Or am I being whooshed?

@puzzlegal

ThisAnd then Google came along. And it was almost always better than any of the others. Within a few months, I rarely used any other search engine. I can promise you that had nothing to do with how cute the names were, and everything to do with how happy I was with the results.

Yes.

Google revolutionized searching on the internet. Whatever their issues are now, they did earn their early success.

[brujaja]

@Chronos

Yes, but what they were doing wasn’t what Google started doing, and people liked what Google did better. Plus, it very quickly got to the point where Google was doing what those other search engines were doing, too.

Let’s suppose, for instance, that you’re looking for information on some incredibly obscure, very specific topic, such that there are only a handful of pages on the Web about that topic. AltaVista was great for searching up things like that: It’d give you all four or five hits for your precisely fine-tuned search query, and you could then check through them yourself. Google, for the first year or so of its existence, wasn’t so good for those sorts of searches, since its coverage of the entire Web wasn’t as exhaustive… but then it caught up, and you could do that with Google too.

On the other hand, let’s suppose that you’re searching for some relatively common topic, that has a vast number of webpages about it, but you only want the best pages for that topic. In a case like that, AltaVista was useless: It’d get you all of the hits, all right, all 17,392 of them, and then you’d have to go through all of them to find the best one. Or more likely, all of the first page before giving up. That’s where Google shines: The hits are sorted by relevance, and the relevance is actually meaningful, so the first hit (or at worst, the second or third) is probably going to be the one you want.

Unsurprisingly, people search for common things more often than they search for uncommon things, so Google was usually better, even before it caught up on coverage. And it’s become all the more true as the Web has grown: Nowadays, even the really obscure topics have 17,392 hits, so the case where Altavista was usable has completely disappeared.

Thank you for this excellent explanation. Well done.

[ultrafilter]

I found a recent article about Google Kubernetes, an open source version of one of their internal tools. This is a significant innovation, and like most of what they do, it’s completely invisible to the consumer.

[DSeid]

@brujaja

Thank you for this excellent explanation. Well done.

But will it come up on a Google search?

[BOOM]

Fwiw, in the end ‘innovators’ kind of works for me - ymmv.