Randall Munroe is learning new things from Google | Hacker News Reader

Randall Munroe is learning new things from Google

twitter.com

130 pointsfortran7744 comments

>I agree but simple things it's really nice ¯\_(ツ)_/¯

Not to pick on this person is particular, but this is why I am so against surfacing untrustworthy information as trustworthy.

Google's AI "knows" that Austin is the capital of Texas with the same system that thinks the Wright Brothers flew through the core of the earth. So why the fuck would anyone trust google that the capital of Texas is Austin? Because you think you can intuit the boundaries of what their AI will and won't fuck up? [0] And that's a reasonable burden to put on dummies who already think search engine results are somehow more authoritative than the simple text matching they're based on at their core?

How is anyone at google with a brain comfortable with this feature being live?

But these low hanging, blatantly obvious, wrong things are actually a good thing. Because once the obvious things are fixed the un-obvious ones will linger as even more unwarranted trust is be placed in the feature. And it will become even harder to get it though people's thick skulls how goddamn stupid it is to outsource reasoning to machines that can't fucking reason.

[0] Sure, xkcd asked fundamentally unanswerable questions, but google notoriously spent years getting "how long does it take to caramelize onions" wrong.

This reminds me of something I once wrote in a conversation about the phrase "when all you have is a hammer, everything looks like a nail" wrt web technologies:

When a tool's convenience grows faster than its reliability and quality, this is what often happens (assuming the tool needs to be more than just convenient). This is just one example.

How is anyone at google with a brain comfortable with this feature being live?

Because people at Google test this feature against their query traffic, measure how many answers are correct and useful, compare it to the number of answers that are wrong and misleading, and use the overall data to decide whether it is on balance more helpful than harmful. With billions of queries flowing through the system, of course you will often make a mistake. The correct strategy is not to set the bar at "zero incorrect answers".

It sure seems like the nature of the question and the likely consequences of answering incorrectly is:

- of profound importance

- obviously unknowable with their current solution, given the absurdity of RM’s screenshots

That’s a pretty dangerous gamble at any moment, but I’d say it’s inherently unethical to take the risk in the current climate.

It’s called Google Brain for a reason

I think this is a UI problem. Google ought to re-state the question they're answering, which may be slightly or entirely different from what was asked.

In other words, if I Google "What year did Tom Hanks land on the moon", the answer box should say "US astronauts landed on the moon in 1970" (not just "1970"). Give users the ability to assess whether the AI was correct.

Why would the factoid "US astronauts landed on the moon in 1970" be anything more then pattern matching on the input?

I don't think I understand your question? Yes, I'd say the system presumably works via some form of pattern matching.

There is no need for a question to answer, just a match.

Oh, I see where you're coming from! Well, if that's the case, that strikes me as something Google needs to fix. It should be pattern matching both the answer and the likely question!

Why would it say that? US astronauts landed on the moon in 1969 (Apollo 11). Tom Hanks starred in the movie Apollo 13, and the real life mission of Apollo 13 happened in 1970.

Whoops! If you can't tell, I didn't actually look up the years! Thanks for the correction.

What is a bit funny is that US astronauts landed on the moon in 1969, 1971, and 1972 but not 1970.

I don’t think it’s a UI problem. You are presuming the system knows the question it is answering.

Yeah, it's not answering questions, it's giving the best answers which they could find themselves. It doesn't means it's the actual answer, nor that it's to the actual question, that's up to you to confirm.

Google isn't an encyclopedia, it's a search engine. It search inside content and try to find the best content that fit your criteria, it could be from trustworthy source, or not, it could be right or wrong on what you meant. For sure their goal is to give you the closest thing to what you asked, with the best sources, but it's not a guarantee, so it shouldn't be trusted as such.

It's crazy that we all came from the "don't trust Wikipedia" and now some people are arguing that we should be able to trust Google results directly.

This is so wrong in so many ways. Google shouldn't restate anything I type in (this is my opinion, but it's based on solid ethics and a decent understanding of good software engineering practices).

Also, everything about Google's "answer" is wrong, from the date, to the people, to the singleton response to a question that should have multiple answers.

Let's all move to DuckDuckGo and Yandex.

If you ask "which president invented coffee", it brings up an inventor named George C. L. Washington who founded a coffee company (and invented a process for instant coffee). So it kind of merges three separate pieces of information (has same name as a president, is an inventor, worked on coffee) in an entirely incorrect way.

https://www.google.com/search?q=which+president+invented+cof...

> So it kind of merges three separate pieces of information (has same name as a president, is an inventor, worked on coffee) in an entirely incorrect way.

Is it an entirely incorrect way? Or does it present a very reasonable pass at what a person was likely to be looking for if they had a recollection that there was a President who invented coffee, but wanted to know who that was?

Entirely incorrect? Or a spectacular feat of induction on a cryptic crossword clue? Where does one draw the line?

I think most humans reading the entirely correct sentence "George Washington invented an early instant coffee process" would interpret that to mean "The first US president invented instant coffee." George Washington (the president) had a wide range of careers like surveying and operating a distillery, and further was an avid coffee drinker (the article google cites for me is one claiming he imported 200 pounds of coffee in 1770). Even if you know when instant coffee was invented, the modifier "early" leaves open the possibility that this was some precursor to what we would today call instant coffee.

These are definitely a bad. Had to tell my girlfriend, who is a high school math teacher, not to trust the info in those boxes.

I think the answer is "google isn't research". Google might be fine for verifying basic/objective fact, such as capitals, but nothing with any sort of nuance, e.g. "Did OJ do it?"

Google is a research tool, not a source to cite.

> ". Google might be fine for verifying basic/objective fact

Unfortunately, the first moon landing happened in 1969, which is an objective fact, so no, Google can't even verify basic facts.

It seems to me like it's just vaguely guessing the question, in a manner of a Google search:

"When did the Simpson's first..." and it leaps ahead, ignoring the nonsense.

"Who flew the first airplane...", again.

I tested the theory with "How big is the Death Star's toilet", and sure enough it promotes a result on the diameter of the Death Star. So, part of this phenomenon is finding queries that are very close to "common" searches, with nonsense attached that's so completely zany that it doesn't push the statistical needle in any particular direction - and it latches on to the "common" search.

edit: I succeeded again with "When did doctor who invent fish fingers", which proudly returns "2010" - which is probably merely the most likely answer to any question involving "when", "doctor who" and "fish fingers".

I actually like this because I often get lots of things wrong about something. It's like I'm giving the engine a whole bunch of detail some of which is misremembered.

"Which movie did tom hanks drill into an asteroid to save the earth"

Tom Hanks wasn't in Armageddon but that's what I was thinking about.

This is a tool to me. I want it to try.

Interesting point.... sometimes it’s a benefit, sometimes a deficit.

I’ve had the experience searching google maps for bagels, and it returns donut shops. Yes, they are both bakeries, but I wanted bagels, not donuts. In this case I don’t want google to ‘guess’ what I’m searching for; I gave it a pretty specific search term (bagels) in context (google maps), yet the results were imho too smart for their own good.

Try using double quotes for literal search (like in web search)

Ironically the top tweet in the More Tweets section for me was "Being good at things is not the point of doing them." https://twitter.com/jodotcom/status/1333728719643570176

I for one welcome all the terribly sourced and miss-informed term papers this will bring.

This should make plagarism detection easier.

I would expect the opposite - presumably people plagiarize (or outright buy) essays which are full of facts and proper citations? I would think kids using nonsense from search engines will produce completely novel (and wrong) papers.

A Googler responded on Twitter, indicating that this may not be at least partially intentional:

> If people ask for things they clearly know aren't true, sometimes featured snippets try to direct them to the most likely related content with context.

They linked to this blog post: https://blog.google/products/search/reintroduction-googles-f...

Why would this only happen when people know they arnt true? Kind of a disingenuous answer, it will lie whether you know what you're asking is silly or not

That sounds like quite the dodge there. I wonder if that Googler was a weatherman in a past life.

At least put up a "We aren't sure what you mean; here is the closest factoid that we think is relevant".

TL;dr - stop presenting everything, including "guesses", as authoritative answers.

Context matters and Google isn't presenting any. There is an implied precision here that is completely unwarranted.

Googles search functionality seems to be getting worse and worse. A lot of the time now, when I try DuckDuckGo it surfaces what I’m looking for in instances where Google just gives me a lot of irrelevant links.

Is it just my imagination (or sample bias) or are other people experiencing the same thing?

> Is it just my imagination (or sample bias) or are other people experiencing the same thing?

I mean, since you asked, I'm still experiencing the opposite when I try DDG every few months or so. Maybe that's my bias, but I do wish it was better.

It's not just your imagination. I tried to look up "how valve index controllers work" and was steered into shopping links and reviews. Not one technical detail about how the Index reads finger gestures. I can't help but suspect Google's algorithm was quite deliberately optimised to steer one into consumerist gestures where it once would steer you into intellectual topics.

Bing search answers the same questions. (Copied the first line of first result with text.) I think Bing's winning the reality contest.

>>what year did tom hanks land on the moon

>I n the movie Apollo 13 (1995), Tom Hanks (and some others) was wearing the Omega Speedmaster.

>>what rocket launched the statue of liberty into the ocean

> Once fully assembled, NASA said the SLS rocket will stand taller than the Statue of Liberty

>>which apollo 13 astronaut exploded

>Two days into the Apollo 13 mission in April 1970, an oxygen tank exploded.

>>when did canada become part of the united states

>America invades Canada, 1775. The invitation was officially made on March 1, 1781,

>>when did the simpsons first learn to control fire

>It's a not-so-merry Christmas for the Simpsons. Mr. Burns plans to cut Christmas bonuses,

When I Google "Tom Hanks", result #5 is USA Today debunking a QAnon theory that pedophilia is considered a disability in Greece, and somehow this is related to the false claim that Hanks is a Greek citizen.

I wonder what Google has to say about Eric Schmidt and his actual Cypriot citizenship.

Is this the game where we use Google Search terms to imply the holocaust didn't happen and get faux outraged when the results include sites saying it didn't, or the Deep Dream game where we find beauty in what AI returns when you feed it illogical data?

I think there's a legitimate distinction between Google returning a list of links in response to "Find me all the holocaust denial stuff" and google presenting a single authoritative answer to questions. Google isn't saying "here is what everyone thinks, sorted by relevance", Google is saying "Here is what everyone thinks, but first let me tell you what I think, and I think X".

Yes, but all the answers are correct -

"What year did Tom Hanks land on the moon" "1970" is a legitimate answer.

Nonsense followed by nonsense as per first year Philosophy/Propositional Logic is a correct statement.

You can't try and trick Google by inputting rubbish and claim you are smarter because really you wanted Google to say "This makes no sense", when really you didn't.

Randall got exactly what he wanted from Google, funny answers to funny questions.

> "What year did Tom Hanks land on the moon" "1970" is a legitimate answer.

I'm just going to leave this here: https://www.nasa.gov/mission_pages/apollo/apollo11.html

I'm amused by the implication that if Google has even one piece of incorrect information in its knowledge corpus, it's justified in returning absolutely any answer to any question. You can deduce anything from a contradiction, after all.

Reply on news.ycombinator.com