Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

twitter/the-algorithm

github.com

Post reply on HN
Print view View original

201–210 of 403 posts

Re: twitter/the-algorithm

#201

Post by LegitShady »

unethical_ban wrote: post #36

There are elements of their algo that I think should be openly defined, and perhaps there should be some regulatory branch that reports to Congress that has full access. However, obfuscation is often necessary to countering bad actors.

>perhaps there should be some regulatory branch that reports to Congress

I think only if you offer twitter users the level of first amendment protection they'd expect with a government body. Otherwise reporting to congress would be an a bold faced circumvention the first amendment. Twitter is a privately held company with no need to report to congress.

Re: twitter/the-algorithm

#202

Post by mistermann »

andrewmcwatters wrote: post #133
axg11 wrote: post #95

I've worked on very large scale recommendation systems at a FAANG. If Twitter's system resembles anything like ours, the concept of publishing or open sourcing "the algorithm" doesn't make sense. Even if we were to open source all associated code and publish all related documents it would be very difficult to make sense of the entire system. That is precisely why companies such as Twitter A/B test the hell out of eve…

For someone who worked on recommendation systems, you really don't seem to understand the concept of an "algorithm" across the abstraction of multiple systems and at different layers of the stack you worked on. In fact, a lot of people here really think what people are talking about is the equivalent of what is handled in a subroutine. No, what people are talking about when they talk about "the algorithm" is anything…

Absolutely agree with what you say, but I wanted to pick a nit here:

> Being pedantic about whether or not this happens in an SQL query, or across multiple codebases, or by region, doesn't escape the question.

Actually, epistemic ~"muddying of the waters" is a well proven technique to control perceptions and public discourse. If it works on HN folks, I expect it would work much more easily on amateurs.

Re: twitter/the-algorithm

#203

Post by butterlesstoast »

tonguez wrote: post #177
mhh__ wrote: post #116

Earlier quoted context omitted.

Of course it doesn't make sense. I think it's just a dog whistle to the people who believe google have a guy in a room somewhere turning the "conservative search results" lever down a notch during elections.

Quoted post unavailable.

Ok cool

Re: twitter/the-algorithm

#204

Post by jjulius »

asabjorn wrote: post #179
mhh__ wrote: post #116

Earlier quoted context omitted.

Of course it doesn't make sense. I think it's just a dog whistle to the people who believe google have a guy in a room somewhere turning the "conservative search results" lever down a notch during elections.

Quoted post unavailable.

Got a link to some actual peer reviews of his research? I've seen him trotted out before, but at most the only data people seem to be able to produce related to his work are interviews like the one you linked, and his website (which is just request for donation after request for donation).

Re: twitter/the-algorithm

#205

Post by nullc »

mhh__ wrote: post #150
tyingq wrote: post #140

Earlier quoted context omitted.

Not saying Google is turning down conservative search results, but they absolutely can. Using the same cycle of human raters and tweaking weights they used to push down comparison shopping sites. See http://graphics.wsj.com/google-ftc-report/

It's the man in a room part I was emphasising. It's simultaneously massively incorrect about how software actually works, misunderstands social dynamics inside companies, and massively stokes the fire when it comes to conspiracy theories along these lines (in this case at least)

So you're saying google doesn't have infrastructure to allow a human specified list of keywords or domain names to "Twiddle" the results returned in varrious services such as news, search, and youtube for the purpose of artificial promotion or de-boosting?

Re: twitter/the-algorithm

#206

Post by tonguez »

axg11 wrote: post #95

I've worked on very large scale recommendation systems at a FAANG. If Twitter's system resembles anything like ours, the concept of publishing or open sourcing "the algorithm" doesn't make sense. Even if we were to open source all associated code and publish all related documents it would be very difficult to make sense of the entire system. That is precisely why companies such as Twitter A/B test the hell out of eve…

it’s almost like this thread is nothing but people being pedantic and saying, there isn’t just one algorithm, it’s multiple algorithms. yes do you really think elon doesn’t know that? it’s almost like he’s just trying to get the point across in the most simple and basic way possible. pointing out that the recommendation algorithm isn’t just one algorithm isn’t profound at all. this entire conversation is mostly just people who want to argue and point out that they know something about how large tech companies work. congrats, why don’t you tell elon since you’re obviously so much smarter

Re: twitter/the-algorithm

#207

Post by Traster »

axg11 wrote: post #95

I've worked on very large scale recommendation systems at a FAANG. If Twitter's system resembles anything like ours, the concept of publishing or open sourcing "the algorithm" doesn't make sense. Even if we were to open source all associated code and publish all related documents it would be very difficult to make sense of the entire system. That is precisely why companies such as Twitter A/B test the hell out of eve…

It's so difficult as someone who works in technology to tolerate this idea that there is "an algorithm". Maybe this will get flagged on HN but the overwhelming feeling is to just scream "don't be such a fucking idiot, this is a $43Bn company it doesn't boil down to 50 lines of code" and hell, there are plenty of examples of 50 lines of code at my company I could spend a month really understanding and even then not understand the full ramifications. It's a stupid persons idea of how tech works.

Re: twitter/the-algorithm

#208

Post by JumpCrisscross »

ironmagma wrote: post #162
jonwinstanley wrote: post #27

Earlier quoted context omitted.

If it’s a list of tweets ordered based on a kajillion ML data points that varies per user is it still an algorithm? And does every user have their own algorithm? And could it be made readable to a human?

In that case, the definition of “the algorithm” is the training set.

> the definition of “the algorithm” is the training set

Or the trained model itself. There are people looking for intentional bias. But the insidiousness of the problem likely arises from unintentional bias. Letting researchers brute force the ranking models with hypotheticals could be a win win.

Re: twitter/the-algorithm

#210

Post by a-dub »

it's probably just a ripoff of pagerank with a separate spam filtering and banning system along with an army of contractors manually fixing it up.

if twitter is a game, sinking $43bn into it is kinda like winning or losing the grand final boss level. (unclear which)

wish elon would get back to facilitating the building of useful things. we still don't have a great clean energy generation story.

Post reply on HN