This is how Google will collapse (2019)
hackernoon.com
hackernoon.com
It is way too "Googly" for its own good. I did a shootout of six visual recognition APIs a few years ago. I got five of them to work in under fifteen minutes. Google Cloud has an authentication system modeled after
https://www.youtube.com/watch?v=sWEvp217Tzw&feature=emb_logo
and took much longer, trashing my Anaconda installation along the way. (IBM Cloud sticks out because it is still sending me spam.)
I'd /never/ use GCP in any sort of business setting. Too high odds an automated Google algorithm would take out my business.
Googley automation worked well enough for B2C, and when Google hired the best people. When Google has a mid-tier workforce, buggy algorithms, and bugs can take out entire businesses, you run into problems. Combined with the old culture of arrogance, I don't see GCP ever competing with Amazon, or even Azure.
https://web.archive.org/web/20200731131748/http://www.sfu.ca...
(Some images seem to be missing)
(Side note, can anyone suggest a better site than imgur? They are pushing mobile users to their app instead of website.)
With Android, we're already seeing forks in China. A bit more stickiness, but not great stickiness.
Ads is still a big two-sided market.
And I've switched Runnaroo.com which has been presented here just some weeks ago. It already feels much better than Google. In rare cases when it fails me I fallback to DuckDuckGo. If DuckDuckGo doesn't manage either - only then I go to Google. In the majority of cases Runnaroo results are the best.
Minor UX quirk: I need to hit <enter> twice to search. That's probably enough friction to lose a lot of people.