Comment by dframe | Hacker News Reader

Comment by dframe | Hacker News Reader

I have a serious question : If google uses a person's search results internally to determine that they are a person of interest for hiring, whose to say they aren't using a similar process to strip Intellectual property from searches as it relates to projects they are working on?

i.e : They are internally working on self driving cars and, in particular, are looking for ways to integrate algorithm 'X' in their solution. So, google decides to setup a special filter for searches related to certain profiles that search for 'X and self driving ....Results get fed to the group conducting the work. In effect, allowing them to internally see what 'profiles of interest' are up to as it relates to Google's flagged area of interest.

It would appear that this could easily be made 'legal' under the premise that the individuals are anonymized. However, it does give google a unique advantage in being able to see 'what everyone is up to'.

The fact that they are already using people's specialized searches for flagging them for interviewing makes me strongly believe they are already internally snagging people's searches/clicked links to give them advantage on projects.

I think everyone should take a pause and consider the implications of this as they are quite damning.

You vastly overestimate the amount of human intelligence that can be gathered from looking at search queries.

Working for search quality, two things I quickly learned from watching a lot of random sample queries are:

(1) People like porn.

(2) People can't spell.

In other words, if you are working for algorithm "X", and if you gather one hundred random queries mentioning X, a lot of them will be along the lines of: "python library X", "CLR ch 12 X solution", "knuth X 1985", "given figure 12 what is the amortized time complexity of X" (or some other exact copy-paste of a homework problem).

People don't type "Amortized time complexity of X can be improved when f and g are monotonically increasing" into a search box.

Besides, exactly who's going to use the result? Imagine you meet a Googler who has PhD in Algorithm X, working for a super secret project, and tell them "You know, I found random guys on the internet searching for X and I thought their disjointed query terms might teach you a thing or two about how to use Algorithm X."

(Not to mention that almost everyone will view it as a gross breach of ethics.)

> I have a serious question : If google uses a person's search results internally to determine that they are a person of interest for hiring, whose to say they aren't using a similar process to strip Intellectual property from searches as it relates to projects they are working on?

Question for you to consider: if google uses a person's search results internally to determine if they should show an ad or a knowledge card, whose to say they aren't using a similar process to strip Intellectual property from searches as it relates to projects they are working on?

It looks like it's a keyword match. You should be just as worried as when you saw the "do a barrel roll" thing.

I agree. These kind of posts make me a little nauseated to be honest. The only reason this person is bragging about being tracked and monitored online is because it resulted in a job at a prestigious marketing company. The vast majority are just being tracked and monitored, no job offered.

It is nauseating because it's a 1000 word humblebrag.

Look at this guy's bio:

> Max has a degree in mathematics from Yale University where he received multiple awards from the math department including the Barge Prize and the Runk Prize. He previously worked at The Boston Consulting Group where he worked on projects in computer hardware and biopharmaceuticals. His interests include data analysis, hiking, pad thai, and skiing.

This essay reads like he is just discovering the magic of the internet for the first time, and is swept away by the user experience that Google has provided just for him!

No, somehow I suspect that he would have gotten a job wherever he decided to apply.

But apropos of the first comment, I wonder how much Google already knew about him before the challenge was displayed. In other words, is it just triggered by keywords? Or does the algorithm also take into account what else is known about the user doing the searching?

>>>> Google already knew about him before the challenge was displayed.

This seems a little unsettling when you think about it. As if they had been tracking him and just waiting for the right moment to spring this on him. I wonder how receptive he would have been had they approached him in a more traditional manner. Perhaps because of the uniqueness of their approach he was more receptive to going along with it?

Had this happened to me, I probably would have done an Elliot Alderson and tossed my PC in the microwave.

It would be a missed opportunity not to include such features. Google are not ones to miss opportunities when it comes to mining personal information.

Pretend multiplication is unknown to Google then write out the series of search queries which teaches Google multiplication. If you cannot complete this exercise sensibly, then you're worried about nothing. Spoiler alert: you're worried about nothing.

If one really wanted to worry about Google using search queries against the interest of searchers, one would probably ask "Do they loop search queries into either search quality against the interest of particular webmasters, or do they loop them into advertising pricing against the interest of particular advertisers?" The answer to both of which is "Yes."

It would be way more efficient to mark their IP not at the search processing level (where they only have so many milliseconds to go around) but at the indexing level... Where they can bring to bear a lot of CPU power.

They can use your searches any way they want to. If you don't like it, don't use google - or if it's bad enough, you can try an antitrust lawsuit, but I wouldn't hold your breath.

Reply on news.ycombinator.com