Goverp wrote:What do you mean by "manipulated", and what would an "unmanipulated" search result look like?
An unmanipulated search result is exactly what you get when you use a tool like find or locate: the algorithm simply takes your search terms and applies them to the search scope without saying "OK, there are a whole bunch of files matching that expression here...but I'm sure the searcher is a moron who didn't mean to type what s/he typed, so I'm not going to report those...instead, I'm going to output all these other files over here...and it helps that their authors paid me to highlight them too". We wouldn't put up with it if our shell search tools worked this way, but we've become so used to it with Internet searches that some people have forgotten unmanipulated search is actually a thing.
Admittedly it came with its own problems...the manipulation is why Google rose to dominance, because if there were 1,000 sites claiming to offer something and one of them actually provided it, a raw search returned all of them and you had to trawl through them manually...whereas Google (to our delight in the early days) would use its manipulations to serve up the best site first. We also benefit a lot from Google's efforts to thwart deliberate SEO: for every company or individual out there trying to manipulate the manipulations to their advantage, Google does attempt to counter-manipulate their manipulations of Google's manipulations (!). Without that activity we'd be at the mercy of millions of selfish individuals and greedy companies throwing everything they could at raw search to manipulate it (from their end) to get us clicking on their links...however...without Google something would have evolved organically to deal with that (FOSS is regulated by voluntary communities like Gentoo, and search could have been similarly regulated by user-developed algorithms for broadcasting the good links and filtering out the bad ones).
Goverp wrote:Google use AI and lots of "stuff" to try to (a) provide the most useful and relevant answer while (b) making revenue for Google. And as they're providing the result for free, what else do you expect?
If we had that attitude to FOSS ("hey, folks are providing this stuff for free: why should we expect it to work for us, rather than for them?") we'd be sitting here not worrying about it while upstream developers sold our information to the highest bidders, etc. What else do I expect? When it's now arguably the main source of information for human beings worldwide (people Google rather than going to the library and searching for stuff in books), and given that information or misinformation is the difference between a peaceful prosperous community and a violent fanatical brainwashed community, I expect them to say "OK, maybe a trillion dollars is enough...maybe we can get by on that...the hundreds of billions were a bit tight, but now that selling up and splitting the money equally would give every single Google employee over 8 billion dollars each, maybe we can stop tuning our algorithms to maximise our profits, and start tuning them to maximise the benefits to our search users".
People of my generation grew up in a world that wasn't shaped by the Internet so at some deep level many of them still imagine the world is like that: rationally they know it isn't, but emotionally they still feel as though the world out there is too fixed and stable to be tossed around by the twitching of fingers on a keyboard. Google are basically a bunch of guys (mostly) who grew up typing BASIC games into their home computers from magazines: I think it just hasn't sunk in for them yet that their code decides what people see globally when they search for answers about "Covid vaccine safety" or "Trump justice policy" or "Islam peaceful or violent" or whatever. They're steering the course of history but still imagine they're just a tech start-up and it's no big deal to use profit for themselves as the basis of what they're doing. When it actually dawns on them how important their code is now, I think they'll open it up. If they don't, they'll eventually be displaced by a FOSS alternative. Google's huge, impressive-looking data centres are not running this show. They're only able to do so because millions of FOSS servers using FOSS protocols choose to point people at those data centres and feed the results back to them: if the developers of that FOSS software decided to collaborate with data scientists and code a distributed search system into the Net and push Google out of it, Google's trillion dollar value could go straight down the toilet.
