Did you happen use google to search for misdeeds by google?
Anyways, searching "artificial 2 second wait to non-AMP ads" is inconclusive. However, if you quote "non-AMP", you get the expected results. Google seems to have a peculiar way of handling dashes and the "non" in "non-AMP" seems to get lost.
You might call that "sentiment analysis". Google has published papers on it. They've refined it so much that they even turned it into a well-documented feature of their Natural Language API [0].
> Google could probably pull it off but for what? Google hate drives views, and ironically, these are commonly monetized by Google itself.
If you could assign a monetary value to showing someone an advertisement (or several), would you? Google has.
If you could assign a monetary value to someone's knowledge about government/justice actions against yourself, would you?
If you could categorize sentiment into something that can generate revenue... or perhaps something that could hurt revenue... would you?
There's money in censorship.
> Anyways, searching "artificial 2 second wait to non-AMP ads" is inconclusive
Let me make it much less inconclusive . The two searches [1] [2] both reveal the lawsuit by the Texas AG which has explicit allegations [3].
[0]: https://cloud.google.com/natural-language/docs/sentiment-tut...
[1]: https://duckduckgo.com/?q=Google+AMP+second+delay+lawsuit
[2]: https://www.google.com/search?q=Google+AMP+second+delay+laws...
[3]: https://www.texasattorneygeneral.gov/sites/default/files/ima...
A quick search I imagined doing (I even imagined that I used a VPN) turned up 10,000,000 results, 49 of which were news articles from the last 15 years detailing what, but not how, Google Search search results were censored. When I imagined clicking "Next" to see the rest of these results, I imagined that I was presented with a No CAPTCHA, asking me to identify which colored block of visual noise most closely resembled hell itself.
Edit: Oh, and then I did the search for real, and got 4.95m results of similar composition.