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

Shirt Without Stripes

github.com

Post reply on HN
Print view View original

391–400 of 639 posts

Re: Shirt Without Stripes

#391

Post by Stranger43 »

bradleybuda wrote: post #365
cscurmudgeon wrote: post #330

Earlier quoted context omitted.

Isn't this just a version of "God works in mysterious ways" applied to AI? "The AI works in mysterious ways. Trust it."

It's the efficient markets hypothesis applied to AI. "If the AI could make more money by showing something else, it would be showing something else".

The original version of which talks of an "Invisible hand" and if that's not an metaphor for divine intervention hidden by an chaotic system i don't know what is.

It's rather surprising how often almost all complex systems theories be it AI, cosmology or economics have an aspects where even the theorists are resorting "to because it is".

Sometimes those statements are based on measured data but it's not always easy or possible to do so accurately for highly interconnected system or worse system where you have actors reacting to theoretical model in a way that changes how the system behaves.

Re: Shirt Without Stripes

#392

Post by leonardopucci »

I think that query analysis in terms of volume of actual people using this query will show that very little people if any actually type "shirt without stripes". Once enough people do it, feedback is accumulated that results are bad (by CTR analysis), and results will auto-correct.

Re: Shirt Without Stripes

#394

Post by jedberg »

holler wrote: post #313
nikanj wrote: post #305

Earlier quoted context omitted.

Your search for "skiing Norway" mostly returns results for skiing in the French Alps, because those pages have much higher visit rates. Google is a dumbass nowadays, and regularly ignores half your search terms to present you with absolutely irrelevant results, that have gotten lots of visits in the past.

I've noticed this too, and frequently wonder why there aren't new and better search startups...

There are, like DuckDuckGo. But the first complaint is usually "their results aren't as good as Google" and that's because Google in reality still gives better results because of their (lack of) privacy.

People want better results but don't want to be tracked, and those things are in opposition to each other.

Re: Shirt Without Stripes

#395

Post by ScoutOrgo »

albertzeyer wrote: post #243

Why should it not be possible to solve this with statistical methods? The model just needs to be able to understand the important meaning of "no" in here, in the context of the whole sentence. I would guess that most modern NNs from the NLP area (Transformer or LSTM) would be able to correctly differentiate the meaning. The problem is, I think there is no fancy NN (yet) behind Google search, and the other web searche…

The OP most likely isn't up to date on latest NN architectures which can learn from the ordering of words. Statistical methods ML, not anymore in NLP.

Re: Shirt Without Stripes

#396

Post by WillPostForFood »

animalCrax0rz wrote: post #314
atq2119 wrote: post #309

Earlier quoted context omitted.

This is true, but isn't really that relevant to the parent's point about statistical methods. Statistical methods (and "deep learning" is such a method) could certainly take the order of words in a sentence into account, for example.

Which part of "The paper I linked to below describes a technique they call "deep parsing." Check it out for more context." could you not parse? I mean if us humans have a difficulty parsing each other's statements, then why should machines do any better?

The number of humans that would have difficulty parsing "shirt without stripes" is very very low.

Re: Shirt Without Stripes

#397

Post by glup »

DenisM wrote: post #197

This problem is known as "attribution" - you have a "no" or "without" in the sentence, but you don't know where it belongs. One could (and one does) argue that the problem cannot be solved with statistical methods (ML), especially not in any domain where accuracy is required, such as medical recored analysis: "no evidence of cancer" and "evidence of no cancer" are very different things. Zooming out, the language fiel…

As a postdoc in computational linguistics, my go-to example for talks is asking Siri not to show me the weather.

Re: Shirt Without Stripes

#398

Post by shanty »

Alex3917 wrote: post #212
rgovostes wrote: post #23

The point that the author is making, in a very understated way, is that all three companies have PR websites that breathlessly describe their advanced AI capabilities, yet they cannot understand a very simple query that young children can.

At least this is relatively innocuous. Until recently if you did a Google Image Search for "person" or "people", it only showed white men.

I think search algorithms still have a long way to go to really understand the intention. Try your image search results for "white person" "black person" "asian person" "white inventors" "black inventors" "asian inventors" Doesn't quite deliver what would be expected.

Re: Shirt Without Stripes

#399

Post by yann63 »

mherdeg wrote: post #310
onurcel wrote: post #293

Earlier quoted context omitted.

The point of the OP is that they claim they understand everything. Example: https://www.blog.google/products/search/search-language-unde...

OK, but what if an Amazon algorithm has actually learned that people who search for "shirt without stripes" are more likely to buy more things if the first image they see is a picture of a striped shirt?

If that's true, should we have the same kind of results when searching with "plain shirts", then?

Which is not the case: searching for "plain shirts" do not give similar results than searching for "shirts without stripes".

Re: Shirt Without Stripes

#400

Post by kevin_thibedeau »

TomJansen wrote: post #351
kevin_thibedeau wrote: post #347

Earlier quoted context omitted.

> If Google was a paid service this problem would be solved the next day. Nope. Cable television was introduced with the promise of no ads. That didn't last long.

Spotify & YouTube Premium run without ads just fine, so why wouldn't it work for search engines? Text is way more compressible then music and video. So at least on serving data it is way more cheaper than the aforementioned services

Because an MBA will realize that they're "leaving money on the table" and not "maximizing shareholder value" by selling ad free subscription services.

Subscription based services also require you to be authenticated and that enables fine grained invasive tracking. Something traditional media couldn't do.

If delivery costs were a factor then I shouldn't be charged $15 for an ebook with near zero distribution costs when a paperback was $5 before ebooks came onto the scene and introduced a new incentive for price gouging.

Post reply on HN