4 ms·
> Google is a search engine not a truth engine IMO "truth engine"s of various forms are the future of search. Think e.g. Wolfram Alpha or any of the newfangled
by throwaway729 10y ago
> Google is a search engine not a truth engine
IMO "truth engine"s of various forms are the future of search. Think e.g. Wolfram Alpha or any of the newfangled chat bots. Or even, increasingly, Google itself.
I don't want a link to some guy's page; I want a definitive answer. Maybe the answer I want is the truth (probably more often than not). Maybe the answer I want is my truth (probably more often than not for political topics). But in any case, this sort of curation is going to happen, so at some point search engines are going to have to wrestle with this reality -- will you be the truth engine people want, or will you be the truth engine the people need? ;-)
- jayajay 10y agoYou're right -- it seems to me that search engines right now are really great at self-validation and confirming almost any opinion, or any point of view. It's a double-edged sword. Like you can look up "why it's ok to kill a citizen" and you'll almost certainly find something that agrees with you and seems reputable. But hey, I guess we can't blame the gun because of the people that decide use it.
- wutbrodo 10y agoThis is the direction Google has been moving in for uncontroversial facts for years: the barriers to them doing this completely are mostly technical, not product-based. For questions where there's any sort of controversy, this broad a solution is by definition unsolvable. It requires taking an explicit side on _literally every controversial issue the human race is talking about_, which means they'd spend roughly 100% of their time fending off mobs demanding to know why they said X. Since we're on the topic, let's use history as an example. There are TONS of things in history that "everyone knew" where the "cranks" turned out to be right and overturned our entire understanding of a given time and part of the world. The scientific and academic methods are messy and iterative and (for the soft sciences) often subjective, attempting to provide a "best guess" for what the human race knows. To ask one entity to explicitly answer questions, stamp out unsupported theories but somehow magically spare contrarian theories that ultimately end up being correct is simply insane.