3 ms·

In prior life, I worked at place where we had access to data that contained user searches. We could plainly see that google searches, due to autocomplete, "qua

by jethkl 4y ago

In prior life, I worked at place where we had access to data that contained user searches. We could plainly see that google searches, due to autocomplete, "quantize" results in a way that bing and others don't. The result is that google searches are almost inorganic crafted things, and they contain unusual artifacts. The impact on society is definitely real.

guerrilla 4y ago
> The impact on society is definitely real. What impact do you mean and how do you know?
- jethkl 4y ago
  For "what" -- I don't specify. Regarding "how I know", I start with several premises: 1) search engines are the dominant way to find information now. 2) conservatively, ≥80% of searches in US are through google. 3) google is not a passive observer of user search queries, but instead actively suggests queries to users. It follows that there must be some impact. The "engagement" effect is this: I start a search for Depeche Mode, but get distracted by suggestions for J Depp. In contrast, DDG suggests Depakote, Depression, Dept of Revenue and a few other things. Other search engines seem to be less aggressive.
  - guerrilla 4y ago
    > For "what" -- I don't specify. Yeah, that's why I'm asking...
    - jethkl 4y ago
      There is hint of "what" at end of my previous comment, but no larger assertion is made.
      - guerrilla 4y ago
        I wasn't asking for a hint, I was asking what you meant.
      - darkerside 4y ago
        I think the problem here is that your statements have all the hallmarks of a conspiracy theory
        jethkl 4y ago
        interesting take, but I don't see any problems here. Conspiratorial hallmarks? I'll try to defend. My comments about ≥80% market etc. are common knowledge and the conclusion is consistent. And my post prior to that is about my observations from a few years ago. Whatever you think of that claim, it is consistent with what is widely known (or easily inferred) about Google: they have top-tier engineering talent that has incentive to get tidy searches sent to them by dynamically generating "is similar to" canonical and trending matches, and aggressively pushing those matches to users before hitting submit. My claim is simply that I noticed Google succeeded in building a better version of this than their competition. I don't see anything suspect there. The interesting artifact of this is that the dominant set of queries becomes a very small proper subset of natural/organic queries.

by jethkl 4y ago

guerrilla 4y ago

> The impact on society is definitely real. What impact do you mean and how do you know?

jethkl 4y ago
For "what" -- I don't specify. Regarding "how I know", I start with several premises: 1) search engines are the dominant way to find information now. 2) conservatively, ≥80% of searches in US are through google. 3) google is not a passive observer of user search queries, but instead actively suggests queries to users. It follows that there must be some impact. The "engagement" effect is this: I start a search for Depeche Mode, but get distracted by suggestions for J Depp. In contrast, DDG suggests Depakote, Depression, Dept of Revenue and a few other things. Other search engines seem to be less aggressive.
- guerrilla 4y ago
  > For "what" -- I don't specify. Yeah, that's why I'm asking...
  - jethkl 4y ago
    There is hint of "what" at end of my previous comment, but no larger assertion is made.
    - guerrilla 4y ago
      I wasn't asking for a hint, I was asking what you meant.
    - darkerside 4y ago
      I think the problem here is that your statements have all the hallmarks of a conspiracy theory
      - jethkl 4y ago
        interesting take, but I don't see any problems here. Conspiratorial hallmarks? I'll try to defend. My comments about ≥80% market etc. are common knowledge and the conclusion is consistent. And my post prior to that is about my observations from a few years ago. Whatever you think of that claim, it is consistent with what is widely known (or easily inferred) about Google: they have top-tier engineering talent that has incentive to get tidy searches sent to them by dynamically generating "is similar to" canonical and trending matches, and aggressively pushing those matches to users before hitting submit. My claim is simply that I noticed Google succeeded in building a better version of this than their competition. I don't see anything suspect there. The interesting artifact of this is that the dominant set of queries becomes a very small proper subset of natural/organic queries.