4 ms·
I'm glad they're moving forward, but I still don't buy it. Someone is tossing a big load of money on them (you know that when you see their ad in twitter every
by scoutt 8y ago
I'm glad they're moving forward, but I still don't buy it. Someone is tossing a big load of money on them (you know that when you see their ad in twitter every day, and articles like this), specially in the last year. When that happens, investors will want their return, no matter what.
People saying they respect your privacy, but... how do you really know? Also because "improved privacy" != "no targeted ads". They can still do massive amounts of money from the data we provide.
At some point:
1) they would have to comply with some law somewhere, handling user data (logs).
2) they may become the number 2 search engine. Spammers will target them and they would have to start doing some behavioral analysis on "what links worked well to those kind of users" to fix that.
Are they so transparent and well intentioned? Is their system so simple, so innocuous and filled only with anonymous data? Fine, they should Open-source the engine and servers, prove that they are using the same code we see, give us RO access to the DBs, and the equivalent of a MacDonalds kitchen tour.
Otherwise, they are just relying on nice words.
- Vinnl 8y ago> how do you really know? You don't really know. However, when your entire USP is being privacy-friendly (and really, that's the primary feature that Google can't copy), it would be incredibly damaging if word got out that that wasn't true. And we can be reasonably certain that, if it wasn't true, word would get out. Also note that DDG explicitly says they don't do user tracking. So while ads are targeted in the sense that you'll see ads relevant to the search term you're using (and your location and other data attached to the query), they're not keeping a profile of you and targeting ads based on that. In other words: there should be no significant user data for them to hand over. I'm not sure why they would have to become the number 2 search engine at some point, although regular search engine manipulation does already seem to affect them sometimes. I don't see why they would have to track users to combat that. Sure, open source would be nice. But right now, we do have reason to assume that they're at least significantly better than Google, without having to sacrifice a lot in search result quality.
- scoutt 8y ago> we do have reason to assume > word would get out > no significant user data These, as you say, are assumptions. That's why I don't (still) buy it. They are in "sell" mode right now. They would say anything to gain traction, and it's working. That's why they are pushing hard and investing a lot. Remember it's about making BIG money through an advertisement-based business, what could possibly go wrong? Imagine that you are magically handled the DB for the searches done today. What would you can possible learn from that information? Perhaps you can't track what searches did HN user Vinnl, but the info has surely a value. They may not be honest about what are they doing with that information.
- Vinnl 8y agoAbsolutely those are assumptions - hence my first sentence. That's still more reliable than trusting Google, about whom we don't need to make assumptions. Another assumption I make is that I don't expect then to seriously displace Google any time soon,but rather remain a player that fills the niche of private search. If that ever appears to change, we can always adapt - let's not make perfect the enemy of the good :-)