A not-so-subtle reading shows Google is doubling down on ecommerce applications here:
> It could also understand that, in the context of hiking, to “prepare” could include things like fitness training as well as finding the right gear.
> fall is the rainy season on Mt. Fuji so you might need a waterproof jacket.
> MUM could also surface helpful subtopics for deeper exploration — like the top-rated gear or best training exercises
> you might see results like where to enjoy the best views of the mountain, onsen in the area and popular souvenir shops
Or, my favorite line:
> MUM would understand the image and connect it with your question to let you know your boots would work just fine. It could then point you to a blog with a list of recommended gear.
(in other words: "Thanks for showing you're interested in hiking gear. Here's a lot of hiking gear you can buy.")
There's an even bigger picture than possibly monetizing ecommerce revenue (through... ads?). The biggest impact is that they get to use all the content generated on the Internet to create these search "results" that synthesize information from multiple sources without ever having to share traffic or ad revenue with those content sources. Clever.
This really is a section that needs regulation. You basically have to use and allow Google to crawl your site if you want a website findable by 95%+ of Americans, so websites really should be able to tell google how they're allowed to use the scraped data instead of just 'for anything'. Maybe a meta tag would work well.
> websites really should be able to tell google how they're allowed to use the scraped data
Isn't it a bit more complicated than that though? While you certainly aren't entitled to republish things, you (ie anyone and everyone) have traditionally been free to consume public material in whatever way you see fit. The precedent from the recent LinkedIn case regarding scraping supports this.
Also you focus on Google, but anyone with sufficient resources can scrape the public web (anti-bot cat and mouse games notwithstanding I suppose).
(Paywalled sites that allow Google to scrape them for search indexing purposes are an interesting edge case though.)
I'm not proposing a link tax, i'm just saying that websites should be able to opt-out from their content being used for some Google products just because they have to give google a very permissive license to transact online business and be found in search results.
I am sure your left field question has a fancy term in the encyclopedia of fallacies, but are you equating Google's crawlers and AI equivalent of humans.
Even if you got an answer to your question, how would it affect the the parents assertion?
Should the internet provider of the person reading the blog post get royalties for the book the person wrote? Should you be paid for your snarky question, as the generator of content?
We shouldn't extend IP protections to general knowledge gleaned is my point. If we start overreaching into that for AI there would be little stopping overreaching into that for humans.
There is a reasonable risk that future research material will not be publicly available by the traditional copyright system. Effectively an "uber" google could leverage all previous content to answer your query severely limiting the utility of producing any new content or maintaining accessibility to pre-existing content.
If you break the distribution model of content producers there won't be any new content, and the old content will simply rot away.
Record disks broke the distribution of taverns and pubs.
Radio broke the distribution model of record disks.
YouTube broke the distribution model of old content producers.
People still make music and more content than ever before. People are not gonna suddenly stop generating knowledge because of Wikipedia. The prediction that "there won't be any new content" is hyperbolic and just plain wrong as demonstrated by history.
When the content people create is pushed aside and instead used by a giant corp. without any payment, it is safe to assume that people will stop writing content.
Books are typically paid for indeed (even if you get them through a library etc.).
Blog posts are an edge case, but the rise of paywalled sites indicate that more authors think they should be paid for their work.
The interesting point here is: can a non-free site let Google crawl it in order to be searchable, while not allowing them to exploit their content otherwise?
The site can ask for such a setup, but Google is very much in a position to refuse and just don't index that site, unless it's one of, say, top 1000 sites.
I don't see any legal ground for such a limitation, so I can't imagine what would e.g. eBay or Walmart put in their suit if they tried to put such a limitation and then saw Google not honor it. Maybe someone with a legal background could comment.
There isn't currently a legal ground for that (in the US). The context of this thread, though, was to propose that such grounds ought to be introduced by the legislature.
I can see the rationale but at some point Google needs to make money on what it contributes to the picture.
On one hand, Google is in large part responsible for the idea that they’re just a free service and anyone can use them without payment, so they’ve trained users to in essence think of them as a non-profit utility which they’re obviously not.
On the other hand, how do you allocate Google’s value when no-one has ever paid for it or been asked to pay for it?
You can legislate Google into the ground (unlikely, but theoretically possible), but what would replace it? Any other similar service operating at their scale would need to make money too.
I don't know if this is actually true or not but I suspect that a big part of their thinking is that "we are just presenting 'facts' and facts are not subject to copyright laws".
Forgive the analogy but that sounds like a parasitic relationship, and one that might kill off, or at least impoverish, its host. Even if Google isn't doing that, the potential exists. The counter is paywall, I suppose.
Another not-so-subtle reading shows google doubling down on being "responsible" which has a lot of collateral damage when they block or de-emphasize legitimate results that don't fit their own goals.
It rings a little hollow when they fired members of/disbanded their nominally independent internal ML Ethics unit after a member published a paper raising some flags on the kind of models Google is betting its future on.
It’s also just a huge step back from ‘ethical’. Responsible implies one can hide behind ‘technical limitations’ or ‘business concerns’. It means you took the unethical option but you at least weighed the pros and cons first
I don't think the language you've quoted was explicitly intended that way. But I think you're onto something. I think high-context answers open up all kinds of new contextual surfaces where ads can be placed, products + product categories suggested.
I don’t know if it’s “e-commerce” specifically, or just a more general fact that Google own a search engine, and want to surface URLs from their index as answers to questions, when appropriate. And, when you think about it, why would you be linking to a page — rather than giving a straightforward answer — unless you’re linking to a product page / review / other page that offers you a direct means of solving a problem that goes beyond a conversational answer?
> unless you’re linking to a product page / review / other page that offers you a direct means of solving a problem
Embedded in this answer seems to be the mindset that only buying things will solve problems.
Don't get me wrong -- I'm not a consumerist luddite, I use my credit card points like any good and proper citizen -- but when your mindset is "all problems can be solved by buying more shit", well, that's a pretty lonely existence.
Google's gotta make money, and helping people buy useful shit is a fine way of doing it, but just don't fall into the mindset trap that everything solution in life is just a Google Pay away.
No, let me rephrase: often, problems can be solved with words. In those cases, the conversational agent wouldn't link to anything. It would just solve your problem.
But if a problem being solved necessitates linking to something, then what kind of problem is that likely to be? Usually one where you need to stare at something, mull over a bunch of details, and make a decision. What kinds of webpages are those? Usually — for public clients — those are product pages.
(Another potential use-case is that a conversational agent could help people configure to software/services by deep-linking to configuration screens — but that's not really a thing Google Search could integrate with.)
Or maybe it's a personal assistant that lives in your phone and asks you how you're doing everyday, acts like a friend, inquires about your mental health and well-being... And then subtly nudges you in the direction of buying X,Y,Z thing or service to help you fill that existential void in your life.
> It could also understand that, in the context of hiking, to “prepare” could include things like fitness training as well as finding the right gear.
> fall is the rainy season on Mt. Fuji so you might need a waterproof jacket.
> MUM could also surface helpful subtopics for deeper exploration — like the top-rated gear or best training exercises
> you might see results like where to enjoy the best views of the mountain, onsen in the area and popular souvenir shops
Or, my favorite line:
> MUM would understand the image and connect it with your question to let you know your boots would work just fine. It could then point you to a blog with a list of recommended gear.
(in other words: "Thanks for showing you're interested in hiking gear. Here's a lot of hiking gear you can buy.")