5 ms·

"Broken" is a stretch. It gives the only result that could be reasonably expected by anyone making that query in good faith. If you're genuinely expecting a res

"Broken" is a stretch. It gives the only result that could be reasonably expected by anyone making that query in good faith. If you're genuinely expecting a result like "never," then I wonder, would you also expect a result of "never" for the query "When did Neal Armstrong set foot on the Moon?" (notice the misspelling of his name)

lordnacho 5y ago
For instance I saw that many Russian generals had been killed during the war. I wondered how many Ukrainian generals had been killed over the same period. Still can't figure it out. Lots and lots of articles about dead Russian generals. There's got to be some way where the search engine doesn't second guess you. You can never find the answer to something adjacent to a popular question with the current state of things.
- ______-_-______ 5y ago
  From google's point of view, it's not second guessing you. It just associates words it decides are similar. You'd probably accept results for "dead generals", "killed generals", or "killed commanders". To google, dead = killed, generals = commanders, Russian = Ukranian, and moon = mars. It will expand the meanings of words until it has a page full of results and ads to show you.
- PeterisP 5y ago
  Ukraine has repeatedly asserted that the HQs of all their major units are intact and as far as I know there have been no announcements of any Ukrainian generals killed and as far as I see even Russian media have not contested that. With some searching on the Russian internet, I find some assertions in Russian social media from early March that Lt Gen Юрій Содоль commanding the Ukraine Marines was killed, but not anything conclusive. So the number seems to be 0, or perhaps 1. And IMHO that isn't surprising, it's quite normal to fight major wars without many generals KIA, it is very surprising that Russia lost so many of them. Even on colonel level the losses seem to be small - I think that this https://www.republicworld.com/world-news/russia-ukraine-crisis/ukrainian-colonel-leading-anti-terrorist-operation-killed-in-mykolayiv-by-russian-forces-articleshow.html https://www.republicworld.com/world-news/russia-ukraine-cris... this was the first Ukrainian ground force colonel KIA that I recall, there was at least one air force colonel before that.
  - Spooky23 5y ago
    Ukraine has better signals security so we don’t know. The other thing is that Russia is operating WW2 style. It’s a top down system where the lower level people have no autonomy. American colonels and generals die in helicopter shoot downs and accidents. Russian generals get assassinated on the front screaming at soldiers to move trucks, etc.
tremon 5y ago
But "Mars" is not simply a misspelling of the correct object, it confuses two distinct objects. A better comparison would be "When did Louis Armstrong set foot on the Moon?"
- tshaddox 5y ago
  Indeed, it's a confusion between two distinct objects, and the intended object is overwhelmingly clear, which is why it's appropriate to correct the confusion and give the intended response.
- JoshuaDavid 5y ago
  Amusingly, "what date did Louis Armstrong land on the Moon" displays the same behavior of showing the date "July 20, 1969" and no indication that it answered a different question than the one you asked.
dave5104 5y ago
I'd consider it broken. In many other cases, if I make a typo in my search, Google will (rather prominently, IMO) at the top of the search display something like: > Showing results for "When did Neil Armstrong set foot on the Moon?" > Show instead for "When did Neil Armwrong set foot on the Moon?" However, it seems to only work for actual spelling errors (e.g. if I typed Armstrong incorrectly) vs. context errors. So to answer your question, I'd expect Google calling out its interpretation, but not necessarily a "never".
makeitdouble 5y ago
What is “good faith” in that case though ? Some people will really not know (think kids for instance, but not only) The interesting part to me is that Google will show redirected queries from typos (“Did you mean “When did *Neil* Armstrong set foot on Mars” ?”) but not disclose the term it completely ignored. I makes it look like it parsed and validated all the search terms before coming up with the prominently displayed date, which makes it worse. On your point, it’s broken because it shouldn’t display “Never”, and instead skip the date widget and show that the page results are for the moon, like it does for typos and other kind of corrected search terms. Or alternatively _always_ show which part of the query the results are based on. Failure to be transparent creates the problem.
- tshaddox 5y ago
  > Some people will really not know (think kids for instance, but not only) It seems very improbable that a kid would ask that question without having any preconceived notion whatsoever of a person with that name famously stepping foot on an extraterrestrial body. It's overwhelmingly likely that they have the famous 1969 event in mind, and that's why Google's response is appropriate. Of course I'm not disputing that it wouldn't be even better for the Google response to explicitly correct the mistake in the query. I'm only disputing the fact that it's "broken." If you were asking this question of a human as some sort of test (but the human didn't know they were being tested), it would clearly be a "gotcha," and you could likely "fool" highly educated people who know full well that it was the Moon and not Mars.
  - cryptoz 5y ago
    The answer Google gives to the question is wrong. End of story. Google aims to be a question-answering machine with the Info box, so giving the wrong answer is a broken state. Not sure you can really reason you way into how it's kinda-understandable-and-gotcha! to be 'not broken'. Aside: Going to get real confusing when another Neil Armstrong born in the 2010s does set foot on Mars in the 2030s. I wonder if we will still get the first response from Google, 1969, and if you will still consider it to be 'not broken' then too.
  - 5y ago
wdevanny 5y ago
At the very least I would expect a "Did you mean..." message at the top. In fact, it correctly does this for the "Neal" query. The lack of one indicates a failure on google's part to understand the answer it gives. Yes, people probably don't make this exact query in good faith very often, but it raises significant concerns for other queries where google tries to supply an answer. See for example: https://gizmodo.com/googles-algorithm-is-lying-to-you-about-onions-and-blam-1793057789 https://gizmodo.com/googles-algorithm-is-lying-to-you-about-...

lordnacho 5y ago

For instance I saw that many Russian generals had been killed during the war. I wondered how many Ukrainian generals had been killed over the same period. Still can't figure it out. Lots and lots of articles about dead Russian generals. There's got to be some way where the search engine doesn't second guess you. You can never find the answer to something adjacent to a popular question with the current state of things.

______-_-______ 5y ago
From google's point of view, it's not second guessing you. It just associates words it decides are similar. You'd probably accept results for "dead generals", "killed generals", or "killed commanders". To google, dead = killed, generals = commanders, Russian = Ukranian, and moon = mars. It will expand the meanings of words until it has a page full of results and ads to show you.
PeterisP 5y ago
Ukraine has repeatedly asserted that the HQs of all their major units are intact and as far as I know there have been no announcements of any Ukrainian generals killed and as far as I see even Russian media have not contested that. With some searching on the Russian internet, I find some assertions in Russian social media from early March that Lt Gen Юрій Содоль commanding the Ukraine Marines was killed, but not anything conclusive. So the number seems to be 0, or perhaps 1. And IMHO that isn't surprising, it's quite normal to fight major wars without many generals KIA, it is very surprising that Russia lost so many of them. Even on colonel level the losses seem to be small - I think that this https://www.republicworld.com/world-news/russia-ukraine-crisis/ukrainian-colonel-leading-anti-terrorist-operation-killed-in-mykolayiv-by-russian-forces-articleshow.html https://www.republicworld.com/world-news/russia-ukraine-cris... this was the first Ukrainian ground force colonel KIA that I recall, there was at least one air force colonel before that.
- Spooky23 5y ago
  Ukraine has better signals security so we don’t know. The other thing is that Russia is operating WW2 style. It’s a top down system where the lower level people have no autonomy. American colonels and generals die in helicopter shoot downs and accidents. Russian generals get assassinated on the front screaming at soldiers to move trucks, etc.

tremon 5y ago

But "Mars" is not simply a misspelling of the correct object, it confuses two distinct objects. A better comparison would be "When did Louis Armstrong set foot on the Moon?"

tshaddox 5y ago
Indeed, it's a confusion between two distinct objects, and the intended object is overwhelmingly clear, which is why it's appropriate to correct the confusion and give the intended response.
JoshuaDavid 5y ago
Amusingly, "what date did Louis Armstrong land on the Moon" displays the same behavior of showing the date "July 20, 1969" and no indication that it answered a different question than the one you asked.

dave5104 5y ago

I'd consider it broken. In many other cases, if I make a typo in my search, Google will (rather prominently, IMO) at the top of the search display something like: > Showing results for "When did Neil Armstrong set foot on the Moon?" > Show instead for "When did Neil Armwrong set foot on the Moon?" However, it seems to only work for actual spelling errors (e.g. if I typed Armstrong incorrectly) vs. context errors. So to answer your question, I'd expect Google calling out its interpretation, but not necessarily a "never".

makeitdouble 5y ago

What is “good faith” in that case though ? Some people will really not know (think kids for instance, but not only) The interesting part to me is that Google will show redirected queries from typos (“Did you mean “When did *Neil* Armstrong set foot on Mars” ?”) but not disclose the term it completely ignored. I makes it look like it parsed and validated all the search terms before coming up with the prominently displayed date, which makes it worse. On your point, it’s broken because it shouldn’t display “Never”, and instead skip the date widget and show that the page results are for the moon, like it does for typos and other kind of corrected search terms. Or alternatively _always_ show which part of the query the results are based on. Failure to be transparent creates the problem.

tshaddox 5y ago
> Some people will really not know (think kids for instance, but not only) It seems very improbable that a kid would ask that question without having any preconceived notion whatsoever of a person with that name famously stepping foot on an extraterrestrial body. It's overwhelmingly likely that they have the famous 1969 event in mind, and that's why Google's response is appropriate. Of course I'm not disputing that it wouldn't be even better for the Google response to explicitly correct the mistake in the query. I'm only disputing the fact that it's "broken." If you were asking this question of a human as some sort of test (but the human didn't know they were being tested), it would clearly be a "gotcha," and you could likely "fool" highly educated people who know full well that it was the Moon and not Mars.
- cryptoz 5y ago
  The answer Google gives to the question is wrong. End of story. Google aims to be a question-answering machine with the Info box, so giving the wrong answer is a broken state. Not sure you can really reason you way into how it's kinda-understandable-and-gotcha! to be 'not broken'. Aside: Going to get real confusing when another Neil Armstrong born in the 2010s does set foot on Mars in the 2030s. I wonder if we will still get the first response from Google, 1969, and if you will still consider it to be 'not broken' then too.
- 5y ago

wdevanny 5y ago

At the very least I would expect a "Did you mean..." message at the top. In fact, it correctly does this for the "Neal" query. The lack of one indicates a failure on google's part to understand the answer it gives. Yes, people probably don't make this exact query in good faith very often, but it raises significant concerns for other queries where google tries to supply an answer. See for example: https://gizmodo.com/googles-algorithm-is-lying-to-you-about-onions-and-blam-1793057789 https://gizmodo.com/googles-algorithm-is-lying-to-you-about-...